<!DOCTYPE article PUBLIC "-//NLM//DTD JATS (Z39.96) Journal Publishing DTD v1.1 20151215//EN" "https://jats.nlm.nih.gov/publishing/1.1/JATS-journalpublishing1.dtd">
<article xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:mml="http://www.w3.org/1998/Math/MathML" dtd-version="1.1" specific-use="sps-1.9" article-type="research-article" xml:lang="pt">
  <front>
    <journal-meta>
      <journal-id journal-id-type="nlm-ta">Arq Bras Cardiol</journal-id>
      <journal-id journal-id-type="publisher-id">abc</journal-id>
      <journal-title-group>
        <journal-title>Arquivos Brasileiros de Cardiologia</journal-title>
        <abbrev-journal-title abbrev-type="publisher">Arq. Bras. Cardiol.</abbrev-journal-title>
      </journal-title-group>
      <issn pub-type="ppub">0066-782X</issn>
      <issn pub-type="epub">1678-4170</issn>
      <publisher>
        <publisher-name>Sociedade Brasileira de Cardiologia - SBC</publisher-name>
      </publisher>
    </journal-meta>
    <article-meta>
      <article-id specific-use="scielo-v3" pub-id-type="publisher-id">VBPZd8vMmxsZFXrhHy4BqXp</article-id>
      <article-id specific-use="scielo-v2" pub-id-type="publisher-id">S0066-782X2021000600879</article-id>
      <article-id pub-id-type="publisher-id">abc.20200231</article-id>
      <article-id pub-id-type="doi">10.36660/abc.20200231</article-id>
      <article-categories>
        <subj-group subj-group-type="heading">
          <subject>Artigo Original</subject>
        </subj-group>
      </article-categories>
      <title-group>
        <article-title>O Efeito Direto do &#205;ndice de Massa Corporal nos Resultados Cardiovasculares entre Participantes sem Obesidade Central pela Estimativa por M&#225;xima Verossimilhan&#231;a Direcionada</article-title>
      </title-group>
      <contrib-group>
        <contrib contrib-type="author">
          <contrib-id contrib-id-type="orcid">0000-0002-5992-3710</contrib-id>
          <name>
            <surname>Saadati</surname>
            <given-names>Hossein Mozafar</given-names>
          </name>
          <xref ref-type="aff" rid="aff1">
            <sup>1</sup>
          </xref>
        </contrib>
        <contrib contrib-type="author">
          <contrib-id contrib-id-type="orcid">0000-0002-1928-992X</contrib-id>
          <name>
            <surname>Sabour</surname>
            <given-names>Siamak</given-names>
          </name>
          <xref ref-type="aff" rid="aff1">
            <sup>1</sup>
          </xref>
        </contrib>
        <contrib contrib-type="author">
          <name>
            <surname>Mansournia</surname>
            <given-names>Mohammad Ali</given-names>
          </name>
          <xref ref-type="aff" rid="aff2">
            <sup>2</sup>
          </xref>
        </contrib>
        <contrib contrib-type="author">
          <name>
            <surname>Mehrabi</surname>
            <given-names>Yadollah</given-names>
          </name>
          <xref ref-type="aff" rid="aff1">
            <sup>1</sup>
          </xref>
        </contrib>
        <contrib contrib-type="author">
          <name>
            <surname>Nazari</surname>
            <given-names>Seyed Saeed Hashemi</given-names>
          </name>
          <xref ref-type="aff" rid="aff3">
            <sup>3</sup>
          </xref>
          <xref ref-type="corresp" rid="c1"/>
        </contrib>
        <aff id="aff1">
          <label>1</label>
          <institution content-type="orgname">Shahid Beheshti University of Medical Sciences</institution>
          <institution content-type="orgdiv1">Faculdade de Sa&#250;de P&#250;blica e Seguran&#231;a</institution>
          <institution content-type="orgdiv2">Departamento de Epidemiologia</institution>
          <addr-line>
            <named-content content-type="city">Tehran</named-content>
          </addr-line>
          <country country="IR">Ir&#227;</country>
          <institution content-type="original">Departamento de Epidemiologia, Faculdade de Sa&#250;de P&#250;blica e Seguran&#231;a, Shahid Beheshti University of Medical Sciences, Tehran - Ir&#227;.</institution>
        </aff>
        <aff id="aff2">
          <label>2</label>
          <institution content-type="orgname">Tehran University of Medical Sciences</institution>
          <institution content-type="orgdiv1">Faculdade de Sa&#250;de P&#250;blica</institution>
          <institution content-type="orgdiv2">Departamento de Epidemiologia e Bioestat&#237;stica</institution>
          <addr-line>
            <named-content content-type="city">Tehran</named-content>
          </addr-line>
          <country country="IR">Ir&#227;</country>
          <institution content-type="original">Departamento de Epidemiologia e Bioestat&#237;stica, Faculdade de Sa&#250;de P&#250;blica, Tehran University of Medical Sciences, Tehran - Ir&#227;.</institution>
        </aff>
        <aff id="aff3">
          <label>3</label>
          <institution content-type="orgname">Shahid Beheshti University of Medical Sciences</institution>
          <institution content-type="orgdiv1">Faculdade de Sa&#250;de P&#250;blica e Seguran&#231;a</institution>
          <institution content-type="orgdiv2">Centro de Pesquisa de Preven&#231;&#227;o de Doen&#231;as Cardiovasculares</institution>
          <addr-line>
            <named-content content-type="city">Tehran</named-content>
          </addr-line>
          <country country="IR">Ir&#227;</country>
          <institution content-type="original">Centro de Pesquisa de Preven&#231;&#227;o de Doen&#231;as Cardiovasculares, Faculdade de Sa&#250;de P&#250;blica e Seguran&#231;a, Shahid Beheshti University of Medical Sciences, Tehran - Ir&#227;.</institution>
        </aff>
      </contrib-group>
      <author-notes>
        <corresp id="c1"><bold>Correspond&#234;ncia: Seyed Saeed Hashemi Nazari</bold> &#8226; Associate Professor, Prevention of Cardiovascular Disease Research Center, School of Public Health and Safety, Shahid Beheshti University of Medical Sciences, Tehran - Ir&#227;. E-mail: <email>saeedh_1999@yahoo.com</email></corresp>
        <fn fn-type="conflict">
          <p>
            <bold>Potencial conflito de interesses</bold>
          </p>
          <p>Os autores declaram n&#227;o haver conflito de interesses pertinentes.</p>
        </fn>
        <fn fn-type="con">
          <p>
            <bold>Contribui&#231;&#227;o dos autores</bold>
          </p>
          <p>Concep&#231;&#227;o e desenho da pesquisa, An&#225;lise e interpreta&#231;&#227;o dos dados e An&#225;lise estat&#237;stica: Saadati HM, Sabour S, Mansournia MA, Mehrabi Y, Nazari SSH; Obten&#231;&#227;o de dados e Reda&#231;&#227;o do manuscrito: Saadati HM, Nazari SSH; Revis&#227;o cr&#237;tica do manuscrito quanto ao conte&#250;do intelectual importante: Sabour S, Mansournia MA, Mehrabi Y, Nazari SSH.</p>
        </fn>
      </author-notes>
      <pub-date publication-format="electronic" date-type="pub">
        <day>17</day>
        <month>05</month>
        <year>2021</year>
      </pub-date>
      <pub-date publication-format="electronic" date-type="collection">
        <month>05</month>
        <year>2021</year>
      </pub-date>
      <volume>116</volume>
      <issue>5</issue>
      <fpage>879</fpage>
      <lpage>886</lpage>
      <history>
        <date date-type="received">
          <day>23</day>
          <month>03</month>
          <year>2020</year>
        </date>
        <date date-type="rev-recd">
          <day>03</day>
          <month>08</month>
          <year>2020</year>
        </date>
        <date date-type="accepted">
          <day>09</day>
          <month>09</month>
          <year>2020</year>
        </date>
      </history>
      <permissions>
        <license license-type="open-access" xml:lang="en" xlink:href="https://creativecommons.org/licenses/by/4.0/">
          <license-p>This is an open-access article distributed under the terms of the Creative Commons Attribution License</license-p>
        </license>
      </permissions>
      <abstract>
        <title>Resumo</title>
        <sec>
          <title>Fundamento:</title>
          <p>O &#237;ndice de massa corporal (IMC) &#233; o &#237;ndice mais usado para categorizar uma pessoa como obesa ou n&#227;o-obesa, e est&#225; sujeito a limita&#231;&#245;es importantes.</p>
        </sec>
        <sec>
          <title>Objetivo:</title>
          <p>Avaliar o efeito direto do IMC nos desfechos cardiovasculares em participantes sem obesidade central.</p>
        </sec>
        <sec>
          <title>M&#233;todos:</title>
          <p>Esta an&#225;lise incluiu 14.983 homens e mulheres com idades entre 45-75 anos do Estudo de Risco de Aterosclerose em Comunidades (ARIC). O IMC foi medido como obesidade geral e a circunfer&#234;ncia da cintura (CC), a rela&#231;&#227;o cintura-quadril (RCQ) e circunfer&#234;ncia do quadril como obesidade central. A estimativa de m&#225;xima verossimilhan&#231;a direcionada (TMLE, no acr&#244;nimo em ingl&#234;s) foi usada para estimar os efeitos totais (TEs) e os efeitos diretos controlados (CDEs). A propor&#231;&#227;o de ET que seria eliminada se todos os participantes fossem n&#227;o obesos em rela&#231;&#227;o &#224; obesidade central foi calculada usando o &#237;ndice de propor&#231;&#227;o eliminada (PE). P&lt;0,05 foi considerado estatisticamente significativo. As an&#225;lises foram realizadas no pacote TMLE R.</p>
        </sec>
        <sec>
          <title>Resultados:</title>
          <p>O risco de desfechos cardiovasculares atribu&#237;dos ao IMC foi significativamente revertido com a elimina&#231;&#227;o da obesidade na RCQ (p &lt;0,001). A propor&#231;&#227;o eliminada dos efeitos do IMC foi mais tang&#237;vel para participantes n&#227;o obesos em rela&#231;&#227;o &#224; CC (PE = 127%; IC95% (126,128)) e RCQ (PE = 97%; IC95% (96,98)) para doen&#231;a arterial coronariana (DAC), e RCQ (PE = 92%; IC95% (91,94)) para acidente vascular cerebral, respectivamente. Com rela&#231;&#227;o ao sexo, a propor&#231;&#227;o eliminada dos efeitos do IMC foi mais tang&#237;vel para participantes n&#227;o obesos em rela&#231;&#227;o a RCQ (PE = 428%; IC95% (408.439)) para DAC em homens e CC (PE = 99%; IC95% (89,111)) para acidente vascular cerebral em mulheres, respectivamente.</p>
        </sec>
        <sec>
          <title>Conclus&#227;o:</title>
          <p>Esses resultados indicam diferentes efeitos potenciais da elimina&#231;&#227;o da obesidade central na associa&#231;&#227;o entre IMC e desfechos cardiovasculares em homens e mulheres. (Arq Bras Cardiol. 2021; 116(5):879-886)</p>
        </sec>
      </abstract>
      <kwd-group xml:lang="pt">
        <title>Palavras-chave</title>
        <kwd>Obesidade</kwd>
        <kwd>Doen&#231;as Cardiovasculares</kwd>
        <kwd>&#205;ndice de Massa Corporal</kwd>
        <kwd>Circunfer&#234;ncia Abdominal</kwd>
        <kwd>Circunfer&#234;ncia da Cintura</kwd>
        <kwd>Aterosclerose; Fatores de Risco</kwd>
      </kwd-group>
      <counts>
        <fig-count count="0"/>
        <table-count count="6"/>
        <equation-count count="2"/>
        <ref-count count="28"/>
        <page-count count="8"/>
      </counts>
    </article-meta>
  </front>
  <body>
    <sec sec-type="intro">
      <title>Introdu&#231;&#227;o</title>
      <p>A obesidade, como fator preditor de doen&#231;a cardiovascular, possui v&#225;rias defini&#231;&#245;es e crit&#233;rios. O &#237;ndice de massa corporal (IMC) &#233; o mais usado para classificar uma pessoa como obesa ou n&#227;o obesa.<xref ref-type="bibr" rid="B1"><sup>1</sup></xref> No entanto, esse &#237;ndice est&#225; sujeito a limita&#231;&#245;es importantes,<xref ref-type="bibr" rid="B1"><sup>1</sup></xref><sup>,</sup><xref ref-type="bibr" rid="B2"><sup>2</sup></xref> pois n&#227;o fornece informa&#231;&#245;es sobre distribui&#231;&#227;o de gordura, e tamb&#233;m n&#227;o pode discriminar entre as diferentes massas corporais (m&#250;sculos, ossos e gordura). Essas limita&#231;&#245;es podem levar a uma classifica&#231;&#227;o incorreta dos n&#237;veis de obesidade.<xref ref-type="bibr" rid="B3"><sup>3</sup></xref><sup>,</sup><xref ref-type="bibr" rid="B4"><sup>4</sup></xref> Por outro lado, os &#237;ndices de obesidade - central, como circunfer&#234;ncia da cintura (CC) e rela&#231;&#227;o cintura-quadril (RCQ), como medidas simples e alternativas de obesidade, medem diretamente a massa de gordura central, o que fornece informa&#231;&#245;es importantes sobre os desfechos de sa&#250;de.<xref ref-type="bibr" rid="B5"><sup>5</sup></xref> Em um estudo de coorte, verificou-se que a CC pode nem sempre estar alinhada com o IMC, e foi proposto que uma combina&#231;&#227;o de IMC e CC poderia estimar melhor as doen&#231;as relacionadas &#224; obesidade.<xref ref-type="bibr" rid="B6"><sup>6</sup></xref> Al&#233;m disso, o IMC &#233; um &#237;ndice geral de obesidade e fornece evid&#234;ncias contradit&#243;rias entre adultos e pessoas com 65 anos ou mais. Esse fen&#244;meno &#233; conhecido como &#8220;paradoxo da obesidade&#8221;.<xref ref-type="bibr" rid="B7"><sup>7</sup></xref><sup>,</sup><xref ref-type="bibr" rid="B8"><sup>8</sup></xref></p>
      <p>Para revelar uma rela&#231;&#227;o causal, precisamos controlar ao m&#225;ximo os potenciais confundidores e as suposi&#231;&#245;es causais. A esse respeito, dois m&#233;todos causais &#8212; Pondera&#231;&#227;o pela Probabilidade Inversa (IPW &#8211; do ingl&#234;s, Inverse Probability Weighting) e f&#243;rmula G &#8212; foram introduzidos. Eles s&#227;o baseados em modelos de exposi&#231;&#227;o e resultados, respectivamente. Em rela&#231;&#227;o ao nosso t&#243;pico, se o modelo ajustado for especificado incorretamente, os resultados ser&#227;o enviesados. Os m&#233;todos duplamente robustos t&#234;m a vantagem de usar simultaneamente modelos de exposi&#231;&#227;o e de resultado e, se apenas um deles for mal especificado, o resultado ainda &#233; v&#225;lido.<xref ref-type="bibr" rid="B9"><sup>9</sup></xref><sup>,</sup><xref ref-type="bibr" rid="B10"><sup>10</sup></xref> Considerando as limita&#231;&#245;es do IMC e as restri&#231;&#245;es dos estudos observacionais, usamos a estimativa de m&#225;xima verossimilhan&#231;a direcionada (TMLE, Targeted Maximum Likelihood Estimation) como um estimador duplamente robusto para reduzir o vi&#233;s dos par&#226;metros-alvo se a exposi&#231;&#227;o ou os mecanismos de resultado fossem estimados de forma consistente,<xref ref-type="bibr" rid="B10"><sup>10</sup></xref> para estimar os efeitos totais (TEs) e os efeitos diretos controlados (CDEs) do IMC. Portanto, este estudo teve como objetivo determinar os TEs e CDEs do IMC nos desfechos cardiovasculares para demonstrar a import&#226;ncia do efeito total do IMC nos desfechos cardiovasculares e quanto desse efeito seria eliminado se todos os participantes fossem n&#227;o obesos em rela&#231;&#227;o &#224; obesidade central (CDE).</p>
    </sec>
    <sec sec-type="methods">
      <title>M&#233;todo</title>
      <sec>
        <title>Participantes</title>
        <p>O estudo Risco de Aterosclerose em Comunidades (ARIC) foi um estudo de coorte prospectivo que come&#231;ou em 1987 em quatro condados nos EUA (Washington County, Maryland; Jackson, Mississippi; Forsyth County, Carolina do Norte; e os sub&#250;rbios de Minneapolis, Minnesota). Os investigadores recrutaram 15.792 participantes com idades entre 45-64 anos. Mais detalhes s&#227;o descritos em outros estudos.<xref ref-type="bibr" rid="B11"><sup>11</sup></xref> Analisamos todos os dados desde a visita um (1987-1989) e a ocorr&#234;ncia do desfecho at&#233; 2014. Para o presente estudo, os participantes com informa&#231;&#245;es ausentes ou hist&#243;rico de qualquer doen&#231;a cardiovascular foram exclu&#237;dos. Os conselhos de revis&#227;o institucional de cada centro aprovaram o protocolo do estudo ARIC e um termo de consentimento livre e esclarecido foi obtido de todos os participantes em cada visita do estudo.</p>
      </sec>
      <sec>
        <title>Medidas</title>
        <sec>
          <title>Exposi&#231;&#227;o: obesidade por defini&#231;&#227;o de &#237;ndice de massa corporal</title>
          <p>Neste estudo, a principal exposi&#231;&#227;o de interesse &#233; a obesidade com defini&#231;&#227;o de IMC. O IMC foi calculado como peso em quilogramas dividido a altura em metros ao quadrado. Obesidade geral foi definida como IMC &#8805;30 kg/m<sup>2</sup>.</p>
        </sec>
        <sec>
          <title>Mediadores: &#237;ndices de obesidade central definidos por circunfer&#234;ncia da cintura, rela&#231;&#227;o cintura-quadril e circunfer&#234;ncia do quadril</title>
          <p>Para avaliar o efeito direto controlado do IMC mediado pela obesidade central (massa gorda central), consideramos tr&#234;s defini&#231;&#245;es para obesidade central, incluindo CC, RCQ e circunfer&#234;ncia do quadril. A CC foi categorizada no ponto de corte &#8805;102 cm nos homens e no ponto de corte &#8805;88 cm nas mulheres. O valor de corte da RCQ foi estabelecido em &#8805;0,9 para homens e &#8805;0,85 para mulheres, de acordo com a Organiza&#231;&#227;o Mundial da Sa&#250;de (OMS).<xref ref-type="bibr" rid="B12"><sup>12</sup></xref> Uma vez que n&#227;o existe um acordo universal quanto ao valor de corte da circunfer&#234;ncia do quadril, ele foi avaliado com base no melhor valor de limiar em uma curva ROC (Receiver Operating Character).</p>
        </sec>
      </sec>
      <sec>
        <title>Doen&#231;a coronariana, insufici&#234;ncia card&#237;aca, acidente vascular cerebral e mortalidade por todas as causas como resultados</title>
        <p>Os resultados deste estudo inclu&#237;ram eventos de doen&#231;a arterial coron&#225;ria (DAC) e insufici&#234;ncia card&#237;aca (IC) registrados at&#233; 31 de dezembro de 2014. De acordo com os crit&#233;rios do estudo ARIC, os resultados de DAC s&#227;o definidos como infarto do mioc&#225;rdio definitivo ou prov&#225;vel ou DAC fatal. Os resultados da IC s&#227;o definidos com base nos crit&#233;rios CID-9 e CID-10. Incidente de IC foi definido como hospitaliza&#231;&#227;o que incluiu o c&#243;digo para IC come&#231;ando com &#8220;428&#8221; (ou seja, 428,0 a 428,9) em qualquer posi&#231;&#227;o, ou uma certid&#227;o de &#243;bito com c&#243;digo CID-9 come&#231;ando com &#8220;428&#8221; ou c&#243;digo CID-10 &#8220;I50&#8221; (IC ou I50.0 a I50.9) em qualquer posi&#231;&#227;o. Os eventos de acidente vascular cerebral (AVC) foram identificados por acompanhamento anual, c&#243;digos hospitalares CID-9 430 a 436 (listados como um c&#243;digo de diagn&#243;stico de alta em qualquer posi&#231;&#227;o) ou em certid&#245;es de &#243;bito. A mortalidade por causa espec&#237;fica foi classificada com base nos atestados de &#243;bito: mortalidade cardiovascular (c&#243;digos CID-9 390-459, c&#243;digos CID-10 I00-I99), mortalidade por c&#226;ncer (c&#243;digos CID-9 140-239, c&#243;digos CID-10 C00-D49), e todas as outras causas de morte.</p>
      </sec>
      <sec>
        <title>Covari&#225;veis e fatores confundidores</title>
        <p>Os dados covari&#225;veis foram considerados fatores de confus&#227;o na associa&#231;&#227;o de desfecho de exposi&#231;&#227;o, mediador de exposi&#231;&#227;o e associa&#231;&#227;o mediador-desfecho. Idade, sexo (masculino e feminino), ra&#231;a (preto e branco), n&#237;vel de educa&#231;&#227;o (b&#225;sico, intermedi&#225;rio, avan&#231;ado), local (condado de Washington, condado de Forsyth, cidade de Jackson, sub&#250;rbios do noroeste de Minneapolis selecionados), tabagismo (definido como atual, ex-fumante e nunca fumante), consume de bebidas alco&#243;licas (definido como consumo atual, ex-alco&#243;latra e nunca consumiu bebidas alco&#243;licas) e pontua&#231;&#227;o total de atividade f&#237;sica (em tr&#234;s dimens&#245;es: no trabalho, no lazer e nos esportes) foram baseados em question&#225;rios de autorrelato. Outras covari&#225;veis inclu&#237;ram ingest&#227;o total de calorias (kcal), hipertens&#227;o (press&#227;o arterial sist&#243;lica &#8805;140 ou press&#227;o arterial diast&#243;lica &#8805;90 mmHg, ou uso de qualquer medicamento para press&#227;o alta), diabetes mellitus (glicemia &#8805;200 e glicemia em jejum &#8805;126 mg/dl, ou uso de qualquer medicamento para diabetes), lip&#237;dios plasm&#225;ticos (mg/dl) e hist&#243;rico de AVC no in&#237;cio do estudo. Os lip&#237;dios plasm&#225;ticos inclu&#237;am colesterol, colesterol de lipoprote&#237;na de alta densidade e triglicer&#237;deos. As covari&#225;veis biol&#243;gicas foram exclu&#237;das das an&#225;lises de TE e inclu&#237;das como potenciais confundidores da associa&#231;&#227;o mediador-desfecho nas an&#225;lises de CDE.</p>
      </sec>
      <sec>
        <title>An&#225;lise estat&#237;stica</title>
        <p>Estat&#237;sticas descritivas foram utilizadas para descrever os participantes (m&#233;dia &#177; desvio-padr&#227;o (DP) para vari&#225;veis cont&#237;nuas e n&#250;mero e porcentagem para vari&#225;veis categ&#243;ricas). Uma an&#225;lise de teste t independente foi usada para examinar as diferen&#231;as estat&#237;sticas em covari&#225;veis cont&#237;nuas entre dois n&#237;veis de exposi&#231;&#227;o de interesse (IMC). Al&#233;m disso, o teste <italic>&#967;</italic>2 foi usado para examinar as associa&#231;&#245;es de vari&#225;veis categ&#243;ricas com a exposi&#231;&#227;o. A normalidade dos dados foi avaliada pela curva normal (assimetria e desvio-padr&#227;o da assimetria) e pelo teste de Kolmogorov-Smirnov. Para calcular o efeito direto controlado do IMC sobre os desfechos (CC, IC, AVC e mortalidade por todas as causas) mediados pela obesidade central, foi utilizado o modelo TMLE. O TMLE, como estimador duplamente robusto, usa modelos de resultado e exposi&#231;&#227;o. A implementa&#231;&#227;o do TMLE segue as seguintes etapas: na primeira, geramos estimadores para o modelo de desfecho da exposi&#231;&#227;o e todos os confundidores listados. Em seguida, geramos estimadores para o modelo de tratamento (exposi&#231;&#227;o e desfecho ausente) em todos os fatores de confus&#227;o listados. Na terceira etapa, calculamos a covari&#225;vel inteligente, H, com base no modelo de tratamento (exposi&#231;&#227;o e desfecho ausente) para ambos expostos &#8220;H = 1/PS&#8221; (PS como escore de propens&#227;o, a probabilidade de exposi&#231;&#227;o) e grupos n&#227;o expostos &#8220;H = -1 / (1-PS)&#8221;.<xref ref-type="bibr" rid="B10"><sup>10</sup></xref><sup>,</sup><xref ref-type="bibr" rid="B13"><sup>13</sup></xref></p>
        <p>O mecanismo ausente &#233; definido como a ocorr&#234;ncia de um evento competitivo (mortalidade total por todas as outras causas, AVC, DAC e IC, para cada desfecho interessado) ou perda de acompanhamento antes da ocorr&#234;ncia do desfecho de interesse, onde &#8220;ausente = 1&#8221; Indica que o resultado &#233; observado e &#8220;ausente = 0&#8221; indica que o resultado est&#225; ausente. Usamos uma defini&#231;&#227;o dicot&#244;mica de exposi&#231;&#227;o (IMC); os valores acima do ponto de corte definido foram classificados como &#8220;obesos&#8221; e os abaixo do ponto de corte como &#8220;n&#227;o obesos&#8221;. Para a vari&#225;vel mediadora, foram utilizados os tr&#234;s &#237;ndices de obesidade central. Dessa forma, fixamos os valores do mediador em zero (n&#227;o obesos de acordo com a obesidade central), de acordo com o modelo causal contrafactual, e avaliamos o efeito direto controlado. TE, na abordagem de infer&#234;ncia causal, &#233; normalmente definido como a diferen&#231;a entre o resultado de interesse de um indiv&#237;duo ou grupo, se exposto a uma exposi&#231;&#227;o espec&#237;fica, e o resultado do mesmo indiv&#237;duo ou grupo, se n&#227;o exposto. O CDE &#233; geralmente definido como a diferen&#231;a entre o resultado de interesse de um indiv&#237;duo ou grupo, se exposto a uma exposi&#231;&#227;o espec&#237;fica, e o resultado do mesmo indiv&#237;duo ou grupo, se n&#227;o exposto, ao fixar o valor dos mediadores. Em nosso estudo, o CDE do IMC foi definido como o efeito do IMC ap&#243;s o controle dos &#237;ndices de CC, RCQ e circunfer&#234;ncia do quadril.<xref ref-type="bibr" rid="B14"><sup>14</sup></xref><sup>,</sup><xref ref-type="bibr" rid="B15"><sup>15</sup></xref></p>
        <p>Para controlar os fatores confundidores e as poss&#237;veis intera&#231;&#245;es, usamos um algoritmo de aprendizagem de m&#225;quina de superaprendizado, que modela diferentes combina&#231;&#245;es de fatores de confus&#227;o e que interagem em diferentes modelos, sendo que as estimativas finais s&#227;o a m&#233;dia ponderada de diferentes estimativas de modelos.</p>
        <p>Ajustamos os algoritmos (modelo linear generalizado, MLG stepwise e MLG de intera&#231;&#227;o) para cada um dos modelos de exposi&#231;&#227;o e resultado, inserindo todas as covari&#225;veis listadas como preditoras e o IMC como exposi&#231;&#227;o bin&#225;ria.</p>
        <p>Em seguida, calculamos o efeito de tratamento aditivo (ETA) como diferen&#231;a de risco para TEs e CDEs, e os intervalos de confian&#231;a correspondentes. A estimativa de vari&#226;ncia baseada na curva de influ&#234;ncia foi usada para estimar os intervalos de confian&#231;a. A valida&#231;&#227;o interna foi realizada no modelo superaprendizado como valida&#231;&#227;o cruzada. A propor&#231;&#227;o eliminada foi calculada de acordo com a seguinte f&#243;rmula:<xref ref-type="bibr" rid="B16"><sup>16</sup></xref></p>
        <disp-formula id="ueq1">
          <mml:math display="block" id="m1">
            <mml:mrow>
              <mml:mi>P</mml:mi>
              <mml:mi>E</mml:mi>
              <mml:mrow>
                <mml:mo>(</mml:mo>
                <mml:mi>m</mml:mi>
                <mml:mo>)</mml:mo>
              </mml:mrow>
              <mml:mo>=</mml:mo>
              <mml:mfrac>
                <mml:mrow>
                  <mml:mi>T</mml:mi>
                  <mml:mi>E</mml:mi>
                  <mml:mo>&#8722;</mml:mo>
                  <mml:mi>C</mml:mi>
                  <mml:mi>D</mml:mi>
                  <mml:mi>E</mml:mi>
                  <mml:mrow>
                    <mml:mo>(</mml:mo>
                    <mml:mi>m</mml:mi>
                    <mml:mo>)</mml:mo>
                  </mml:mrow>
                </mml:mrow>
                <mml:mrow>
                  <mml:mi>T</mml:mi>
                  <mml:mi>E</mml:mi>
                </mml:mrow>
              </mml:mfrac>
            </mml:mrow>
          </mml:math>
        </disp-formula>
        <p>Onde PE &#233; a propor&#231;&#227;o eliminada, TE &#233; o efeito total, CDE &#233; o efeito direto controlado e m &#233; a fixa&#231;&#227;o do n&#237;vel do mediador em zero (n&#227;o obeso). Intervalos de confian&#231;a (95%) para PE foram avaliados usando o m&#233;todo <italic>bootstrap</italic>. O valor de P&lt;0,05 foi considerado estatisticamente significativo. A an&#225;lise foi realizada no pacote TMLE R vers&#227;o 3.5.3.</p>
      </sec>
    </sec>
    <sec sec-type="results">
      <title>Resultados</title>
      <sec>
        <title>Caracter&#237;sticas dos participantes</title>
        <p>Dos 14.983 participantes na linha de base, inclu&#237;mos 12.085, 12.085, 12.725 e 12.936 participantes nesta an&#225;lise depois de excluir todos os indiv&#237;duos com hist&#243;rico de qualquer doen&#231;a cardiovascular e dados ausentes na linha de base para DAC, IC, AVC e mortalidade por todas as causas, respectivamente. Para mortalidade por todas as causas, inclu&#237;mos todos os participantes com hist&#243;rico de qualquer doen&#231;a cardiovascular. Durante uma mediana de 27 anos de acompanhamento, 1.616 (13,37%), 2.229 (18,44%), 1.078 (8,47%) e 5.364 (41,47%) participantes tiveram DAC, IC, AVC e mortalidade por todas as causas, respectivamente. Nesse per&#237;odo, 3.416 (22,8%) e 1.035 (6,91%) participantes tiveram perda de acompanhamento e risco competitivo, respectivamente. Em rela&#231;&#227;o aos participantes com obesidade com base no IMC, durante uma mediana de 27 anos de acompanhamento, 500 (16,43%), 848 (27,86%), 357 (10,67%) e 1.676 (49,08%) participantes experimentaram DAC, IC, AVC e mortalidade por todas as causas, respectivamente. As caracter&#237;sticas da linha de base (m&#233;dia e desvio-padr&#227;o para vari&#225;veis cont&#237;nuas e n&#250;mero e porcentagem para vari&#225;veis categ&#243;ricas para participantes com e sem obesidade, com base no IMC) s&#227;o fornecidas na <xref ref-type="table" rid="t1">Tabela 1</xref> e nas Tabelas complementares 1-4. Os participantes obesos, por defini&#231;&#227;o de IMC, eram mais propensos a serem mulheres, de cor parda, ter n&#237;vel educacional mais baixo e renda familiar anual mais baixa, e menos propensos a terem plano de sa&#250;de em compara&#231;&#227;o com indiv&#237;duos n&#227;o obesos. Em rela&#231;&#227;o &#224;s vari&#225;veis mediadoras, participantes obesos tinham maior probabilidade de ser obesos com base nos &#237;ndices de CC, circunfer&#234;ncia do quadril e RCQ, respectivamente.</p>
        <table-wrap id="t1">
          <label>Tabela 1</label>
          <caption>
            <title>Caracter&#205;sticas de base dos participantes do Estudo ARIC por IMC, 1987-2014</title>
          </caption>
          <table frame="hsides" rules="rows">
            <colgroup width="20%">
              <col/>
              <col/>
              <col/>
              <col/>
              <col/>
            </colgroup>
            <thead style="border-top: thin solid; border-bottom: thin solid; border-color: #000000">
              <tr>
                <th valign="top" align="left" rowspan="2">Caracter&#237;sticas</th>
                <th valign="top" align="center" rowspan="2"/>
                <th valign="top" align="center" colspan="2">&#205;ndice de massa corporal</th>
                <th valign="top" align="center" rowspan="3">Valor de p*</th>
              </tr>
              <tr>
                <th valign="top" align="center">Obeso</th>
                <th valign="top" align="center">N&#227;o obeso</th>
              </tr>
              <tr>
                <th valign="top" align="left">Confundidores categ&#243;ricos</th>
                <th valign="top" align="center"/>
                <th valign="top" align="center" colspan="2">No. %</th>
              </tr>
            </thead>
            <tbody style="border-bottom: thin solid; border-color: #000000">
              <tr>
                <td valign="top" align="left">Sexo</td>
                <td valign="top" align="center">feminino</td>
                <td valign="top" align="center">2.484 (60,85)</td>
                <td valign="top" align="center">5.686 (52,16)</td>
                <td valign="top" align="center">&lt;0,001</td>
              </tr>
              <tr>
                <td valign="top" align="center"/>
                <td valign="top" align="center">masculino</td>
                <td valign="top" align="center">1.598 (39,15)</td>
                <td valign="top" align="center">5.215 (47,84)</td>
                <td valign="top" align="center"/>
              </tr>
              <tr>
                <td valign="top" align="left">Ra&#231;a</td>
                <td valign="top" align="center">branca</td>
                <td valign="top" align="center">2.514 (61,59)</td>
                <td valign="top" align="center">8.613 (79,01)</td>
                <td valign="top" align="center">&lt;0,001</td>
              </tr>
              <tr>
                <td valign="top" align="center"/>
                <td valign="top" align="center">negra</td>
                <td valign="top" align="center">1.568 (38,41)</td>
                <td valign="top" align="center">2.288 (20,99)</td>
                <td valign="top" align="center"/>
              </tr>
              <tr>
                <td valign="top" align="left">Educa&#231;&#227;o</td>
                <td valign="top" align="center">B&#225;sica</td>
                <td valign="top" align="center">1.237 (30,39)</td>
                <td valign="top" align="center">2.300 (21,12)</td>
                <td valign="top" align="center">&lt;0,001</td>
              </tr>
              <tr>
                <td valign="top" align="center"/>
                <td valign="top" align="center">Intermedi&#225;ria</td>
                <td valign="top" align="center">1.633 (40,11)</td>
                <td valign="top" align="center">4.494 (41,27)</td>
                <td valign="top" align="center"/>
              </tr>
              <tr>
                <td valign="top" align="center"/>
                <td valign="top" align="center">Avan&#231;ada</td>
                <td valign="top" align="center">1.201 (29,50)</td>
                <td valign="top" align="center">4.094 (37,60)</td>
                <td valign="top" align="center"/>
              </tr>
              <tr>
                <td valign="top" align="left">Family Income (per year) Renda familiar (por ano)</td>
                <td valign="top" align="center">menos de $16,000</td>
                <td valign="top" align="center">1.208 (31,71)</td>
                <td valign="top" align="center">2.011 (19,48)</td>
                <td valign="top" align="center">&lt;0,001</td>
              </tr>
              <tr>
                <td valign="top" align="center"/>
                <td valign="top" align="center">$16,000 &#8211;$50,000</td>
                <td valign="top" align="center">1.948 (51,13)</td>
                <td valign="top" align="center">5.437 (52,66)</td>
                <td valign="top" align="center"/>
              </tr>
              <tr>
                <td valign="top" align="center"/>
                <td valign="top" align="center">mais de $50,000</td>
                <td valign="top" align="center">654 (17,17)</td>
                <td valign="top" align="center">2.876 (27,86)</td>
                <td valign="top" align="center"/>
              </tr>
              <tr>
                <td valign="top" align="left">Ingest&#227;o de bebida alco&#243;lica</td>
                <td valign="top" align="center">Bebe atualmente</td>
                <td valign="top" align="center">1.805 (44,61)</td>
                <td valign="top" align="center">6.591 (60,64)</td>
                <td valign="top" align="center">&lt;0,001</td>
              </tr>
              <tr>
                <td valign="top" align="center"/>
                <td valign="top" align="center">Bebia no passado</td>
                <td valign="top" align="center">912 (22,54)</td>
                <td valign="top" align="center">1.931 (17,77)</td>
                <td valign="top" align="center"/>
              </tr>
              <tr>
                <td valign="top" align="center"/>
                <td valign="top" align="center">Nunca bebeu</td>
                <td valign="top" align="center">1.329 (32,85)</td>
                <td valign="top" align="center">2.347 (21,59)</td>
                <td valign="top" align="center"/>
              </tr>
              <tr>
                <td valign="top" align="left">Tabagismo</td>
                <td valign="top" align="center">Fumante</td>
                <td valign="top" align="center">793 (19,44)</td>
                <td valign="top" align="center">3.158 (28,99)</td>
                <td valign="top" align="center">&lt;0,001</td>
              </tr>
              <tr>
                <td valign="top" align="center"/>
                <td valign="top" align="center">Ex-fumante</td>
                <td valign="top" align="center">1.369 (33,56)</td>
                <td valign="top" align="center">3.487 (32,01)</td>
                <td valign="top" align="center"/>
              </tr>
              <tr>
                <td valign="top" align="center"/>
                <td valign="top" align="center">Nunca fumou</td>
                <td valign="top" align="center">1.917 (47,00)</td>
                <td valign="top" align="center">4.248 (39,00)</td>
                <td valign="top" align="center"/>
              </tr>
              <tr>
                <td valign="top" align="left">Plano de sa&#250;de</td>
                <td valign="top" align="center">N&#227;o</td>
                <td valign="top" align="center">562 (13,82)</td>
                <td valign="top" align="center">901 (8,27)</td>
                <td valign="top" align="center">&lt;0,001</td>
              </tr>
              <tr>
                <td valign="top" align="center"/>
                <td valign="top" align="center">Sim</td>
                <td valign="top" align="center">3,506 (86,18)</td>
                <td valign="top" align="center">9,991(91,73)</td>
                <td valign="top" align="center"/>
              </tr>
              <tr>
                <td valign="top" align="left">Hist&#243;rico familiar de CAD</td>
                <td valign="top" align="center">N&#227;o</td>
                <td valign="top" align="center">1,719 (42,70)</td>
                <td valign="top" align="center">4.556 (42,42)</td>
                <td valign="top" align="center">0,76</td>
              </tr>
              <tr>
                <td valign="top" align="center"/>
                <td valign="top" align="center">Sim</td>
                <td valign="top" align="center">2.307 (57,30)</td>
                <td valign="top" align="center">6.183 (57,58)</td>
                <td valign="top" align="center"/>
              </tr>
              <tr>
                <td valign="top" align="left">Hipertens&#227;o</td>
                <td valign="top" align="center">N&#227;o</td>
                <td valign="top" align="center">2.229 (54,97)</td>
                <td valign="top" align="center">8.258 (76,16)</td>
                <td valign="top" align="center">&lt;0,001</td>
              </tr>
              <tr>
                <td valign="top" align="center"/>
                <td valign="top" align="center">Sim</td>
                <td valign="top" align="center">1.826 (45,03)</td>
                <td valign="top" align="center">2.585 (23,84)</td>
                <td valign="top" align="center"/>
              </tr>
              <tr>
                <td valign="top" align="left">Rem&#233;dios para hipertens&#227;o</td>
                <td valign="top" align="center">N&#227;o</td>
                <td valign="top" align="center">2.208 (54,12)</td>
                <td valign="top" align="center">8.156 (74,85)</td>
                <td valign="top" align="center">&lt;0,001</td>
              </tr>
              <tr>
                <td valign="top" align="center"/>
                <td valign="top" align="center">Sim</td>
                <td valign="top" align="center">1.872 (45,88)</td>
                <td valign="top" align="center">2.740 (25,15)</td>
                <td valign="top" align="center"/>
              </tr>
              <tr>
                <td valign="top" align="left">Diabetes mellitus</td>
                <td valign="top" align="center">N&#227;o</td>
                <td valign="top" align="center">3.234 (80,31)</td>
                <td valign="top" align="center">10.114 (93,38)</td>
                <td valign="top" align="center">&lt;0,001</td>
              </tr>
              <tr>
                <td valign="top" align="center"/>
                <td valign="top" align="center">Sim</td>
                <td valign="top" align="center">793 (19,69)</td>
                <td valign="top" align="center">717 (6,62)</td>
                <td valign="top" align="center"/>
              </tr>
              <tr>
                <td valign="top" align="left">
                  <bold>Confundidores cont&#237;nuos</bold>
                </td>
                <td valign="top" align="center"/>
                <td valign="top" align="center" colspan="2">
                  <bold>M&#233;dia (DP)</bold>
                </td>
                <td valign="top" align="center"/>
              </tr>
              <tr>
                <td valign="top" align="left">Idade, anos</td>
                <td valign="top" align="center"/>
                <td valign="top" align="center">54,09 (5,70)</td>
                <td valign="top" align="center">54,30 (5,78)</td>
                <td valign="top" align="center">0,04</td>
              </tr>
              <tr>
                <td valign="top" align="left">Atividade f&#237;sica (trabalho)</td>
                <td valign="top" align="center"/>
                <td valign="top" align="center">2,18 (0,99)</td>
                <td valign="top" align="center">2,17 (0,93)</td>
                <td valign="top" align="center">0,67</td>
              </tr>
              <tr>
                <td valign="top" align="left">Atividade f&#237;sica (esporte)</td>
                <td valign="top" align="center"/>
                <td valign="top" align="center">2,27 (0,72)</td>
                <td valign="top" align="center">2,49 (0,81)</td>
                <td valign="top" align="center">&lt;0,001</td>
              </tr>
              <tr>
                <td valign="top" align="left">Atividade f&#237;sica (lazer)</td>
                <td valign="top" align="center"/>
                <td valign="top" align="center">2,26 (0,57)</td>
                <td valign="top" align="center">2,39 (0,57)</td>
                <td valign="top" align="center">&lt;0,001</td>
              </tr>
              <tr>
                <td valign="top" align="left">Ingest&#227;o total de energia (Kcal)</td>
                <td valign="top" align="center"/>
                <td valign="top" align="center">1632,4 (702,3)</td>
                <td valign="top" align="center">1637,2 (703,1)</td>
                <td valign="top" align="center">0,72</td>
              </tr>
              <tr>
                <td valign="top" align="left">&#193;cidos graxos saturados (% Kcal)</td>
                <td valign="top" align="center"/>
                <td valign="top" align="center">12,23 (2,93)</td>
                <td valign="top" align="center">11,93 (3,02)</td>
                <td valign="top" align="center">&lt;0,001</td>
              </tr>
              <tr>
                <td valign="top" align="left">Colesterol total mg/dl</td>
                <td valign="top" align="center"/>
                <td valign="top" align="center">5,62 (1,12)</td>
                <td valign="top" align="center">5,54 (1,07)</td>
                <td valign="top" align="center">&lt;0,001</td>
              </tr>
              <tr>
                <td valign="top" align="left">Triglicer&#237;deos mg/dl</td>
                <td valign="top" align="center"/>
                <td valign="top" align="center">1,76 (1,28)</td>
                <td valign="top" align="center">1,40 (0,89)</td>
                <td valign="top" align="center">&lt;0,001</td>
              </tr>
              <tr>
                <td valign="top" align="left">Colesterol HDL mg/dl</td>
                <td valign="top" align="center"/>
                <td valign="top" align="center">1,20 (0,36)</td>
                <td valign="top" align="center">1,37 (0,46)</td>
                <td valign="top" align="center">&lt;0,001</td>
              </tr>
              <tr>
                <td valign="top" align="left">
                  <bold>Mediadores</bold>
                </td>
                <td valign="top" align="center"/>
                <td valign="top" align="center" colspan="2">
                  <bold>No. %</bold>
                </td>
                <td valign="top" align="center"/>
              </tr>
              <tr>
                <td valign="top" align="left">Circunfer&#234;ncia da cintura</td>
                <td valign="top" align="center">N&#227;o obeso</td>
                <td valign="top" align="center">108 (2,65)</td>
                <td valign="top" align="center">6.893 (63,23)</td>
                <td valign="top" align="center">&lt;0,001</td>
              </tr>
              <tr>
                <td valign="top" align="left"/>
                <td valign="top" align="center">Obeso</td>
                <td valign="top" align="center">3.974 (97,35)</td>
                <td valign="top" align="center">4.008 (36,77)</td>
                <td valign="top" align="center"/>
              </tr>
              <tr>
                <td valign="top" align="left">Rela&#231;&#227;o cintura-quadril</td>
                <td valign="top" align="center">N&#227;o obeso</td>
                <td valign="top" align="center">275 (6,74)</td>
                <td valign="top" align="center">2.927 (26,85)</td>
                <td valign="top" align="center">&lt;0,001</td>
              </tr>
              <tr>
                <td valign="top" align="left"/>
                <td valign="top" align="center">Obeso</td>
                <td valign="top" align="center">3.807 (93,26)</td>
                <td valign="top" align="center">7.974 (73,15)</td>
                <td valign="top" align="center"/>
              </tr>
              <tr>
                <td valign="top" align="left">Circunfer&#234;ncia do quadril</td>
                <td valign="top" align="center">N&#227;o obeso</td>
                <td valign="top" align="center">989 (24,23)</td>
                <td valign="top" align="center">10.246 (93,99)</td>
                <td valign="top" align="center">&lt;0,001</td>
              </tr>
              <tr>
                <td valign="top" align="left"/>
                <td valign="top" align="center">Obeso</td>
                <td valign="top" align="center">3.093 (75,77)</td>
                <td valign="top" align="center">655 (6,01)</td>
                <td valign="top" align="center"/>
              </tr>
            </tbody>
          </table>
          <table-wrap-foot>
            <fn id="TFN1">
              <p><italic>Valor de p baseado no teste &#967;2 e teste t independente para vari&#225;veis categ&#243;ricas e cont&#237;nuas, respectivamente; ARIC: Estudo de Risco de Aterosclerose em Comunidades; M&#233;dia e desvio-padr&#227;o das vari&#225;veis cont&#237;nuas em cada grupo de &#237;ndice de massa corporal; N&#250;mero e porcentagem de vari&#225;veis categ&#243;ricas em cada grupo de &#237;ndice de massa corporal</italic>.</p>
            </fn>
          </table-wrap-foot>
        </table-wrap>
      </sec>
      <sec>
        <title>Efeitos totais e efeitos diretos controlados</title>
        <p>Os TEs e CDEs do IMC, para todos os desfechos de interesse como efeito de tratamento aditivo (diferen&#231;a de risco) com intervalos de confian&#231;a de 95%, estimados por TMLE para todos os participantes e por sexo, est&#227;o demonstrados nas <xref ref-type="table" rid="t2">Tabelas 2</xref> e <xref ref-type="table" rid="t3">3</xref> e na Figura 1 complementar. Sobre os TEs, os resultados mostram uma associa&#231;&#227;o forte e significativa entre o IMC e todos os resultados. Os mais fortes s&#227;o estimados para IC, mortalidade por todas as causas, DAC e AVC, respectivamente. Em rela&#231;&#227;o a CDEs, grandes CDEs para IC e mortalidade por todas as causas, ap&#243;s o controle de todos os tr&#234;s &#237;ndices de obesidade central separadamente (ETA entre = 4,27 e 7,95), sugerem que, mesmo se a obesidade central fosse eliminada, permaneceria um forte efeito para o IMC. Por outro lado, pequenos efeitos diretos controlados para DAC e AVC, ap&#243;s ajuste para todos os tr&#234;s &#237;ndices de obesidade central separadamente, especialmente CC em casos de CAD e circunfer&#234;ncia do quadril para AVC (ETA entre = -2,81 e 3,06), sugerem que, se a obesidade central fosse eliminada, seria eliminado um forte efeito do IMC.</p>
        <table-wrap id="t2">
          <label>Tabela 2</label>
          <caption>
            <title>Estimativa do efeito direto controlado do &#237;ndice de massa corporal para DAC, IC, AVC e mortalidade por todas as causas, de acordo com Obesidade Central (n&#227;o obeso), participantes do Estudo ARIC, 1987-2014 (caso completo)</title>
          </caption>
          <table frame="hsides" rules="rows">
            <colgroup width="20%">
              <col/>
              <col/>
              <col/>
              <col/>
              <col/>
            </colgroup>
            <thead style="border-top: thin solid; border-bottom: thin solid; border-color: #000000">
              <tr>
                <th valign="top" align="left" rowspan="2">Desfechos</th>
                <th valign="top" align="center" rowspan="2">Mediador (&#205;ndice de obesidade central, n&#227;o obeso)</th>
                <th valign="top" align="center">Efeito direto controlado</th>
                <th valign="top" align="center">Propor&#231;&#227;o eliminada</th>
                <th valign="top" align="center" rowspan="2">Efeito total (ETA) (IC95%)</th>
              </tr>
              <tr>
                <th valign="top" align="center">ETA (IC95%)</th>
                <th valign="top" align="center">PE % (IC95%)</th>
              </tr>
            </thead>
            <tbody style="border-bottom: thin solid; border-color: #000000">
              <tr>
                <td valign="top" align="left" rowspan="3">DAC</td>
                <td valign="top" align="center">CC</td>
                <td valign="top" align="center">-2,81 (-5,01, -0,61)</td>
                <td valign="top" align="center">127 (109, 135)</td>
                <td valign="top" align="center" rowspan="3">10,47 (7,76, 13,18)</td>
              </tr>
              <tr>
                <td valign="top" align="center">RCQ</td>
                <td valign="top" align="center">0,62 (-2,58, 3,82)</td>
                <td valign="top" align="center">94 (79, 99)</td>
              </tr>
              <tr>
                <td valign="top" align="center">Quadril</td>
                <td valign="top" align="center">3,06 (-1,20, 7,33)</td>
                <td valign="top" align="center">71 (63, 75)</td>
              </tr>
              <tr>
                <td valign="top" align="left" rowspan="3">IC</td>
                <td valign="top" align="center">CC</td>
                <td valign="top" align="center">6,41 (4,09, 8,72)</td>
                <td valign="top" align="center">60 (55,75)</td>
                <td valign="top" align="center" rowspan="3">15,92 (13,45, 18,39)</td>
              </tr>
              <tr>
                <td valign="top" align="center">RCQ</td>
                <td valign="top" align="center">7,95 (4,54, 11,37)</td>
                <td valign="top" align="center">50 (36,52)</td>
              </tr>
              <tr>
                <td valign="top" align="center">Quadril</td>
                <td valign="top" align="center">7,23 (3,33,11,13)</td>
                <td valign="top" align="center">54 (57,64)</td>
              </tr>
              <tr>
                <td valign="top" align="left" rowspan="3">AVC</td>
                <td valign="top" align="center">CC</td>
                <td valign="top" align="center">2,11 (-0,06, 4,29)</td>
                <td valign="top" align="center">76 (74,101)</td>
                <td valign="top" align="center" rowspan="3">8,32 (6,01, 10,63)</td>
              </tr>
              <tr>
                <td valign="top" align="center">RCQ</td>
                <td valign="top" align="center">0,69 (-2,34, 3,73)</td>
                <td valign="top" align="center">92 (81,106)</td>
              </tr>
              <tr>
                <td valign="top" align="center">Quadril</td>
                <td valign="top" align="center">0,05 (-3,82,3,92)</td>
                <td valign="top" align="center">99 (93,108)</td>
              </tr>
              <tr>
                <td valign="top" align="left" rowspan="3">Mortalidade (todas as causas)</td>
                <td valign="top" align="center">CC</td>
                <td valign="top" align="center">4,88 (2,56, 7,20)</td>
                <td valign="top" align="center">56 (49,73)</td>
                <td valign="top" align="center" rowspan="3">11,05 (9,21, 12,88)</td>
              </tr>
              <tr>
                <td valign="top" align="center">RCQ</td>
                <td valign="top" align="center">4,27 (0,36, 8,17)</td>
                <td valign="top" align="center">61 (52,69)</td>
              </tr>
              <tr>
                <td valign="top" align="center">Quadril</td>
                <td valign="top" align="center">5,19 (2,01,8,36)</td>
                <td valign="top" align="center">53 (50,59)</td>
              </tr>
            </tbody>
          </table>
          <table-wrap-foot>
            <fn id="TFN2">
              <p><italic>ETA: efeito de tratamento aditivo; PE: propor&#231;&#227;o eliminada; DAC: doen&#231;a arterial coron&#225;ria; IC: insufici&#234;ncia card&#237;aca; AVC: acidente vascular cerebral; CC: circunfer&#234;ncia da cintura; RCQ: rela&#231;&#227;o cintura-quadril; Quadril: circunfer&#234;ncia do quadril</italic>.</p>
            </fn>
          </table-wrap-foot>
        </table-wrap>
        <table-wrap id="t3">
          <label>Tabela 3</label>
          <caption>
            <title>Estimativa do efeito direto controlado do IMC na DAC, IC, AVC e mortalidade por todas as causas, de acordo com Obesidade Central (n&#227;o obeso), em homens e mulheres no Estudo ARIC, 1987-2014 (caso completo)</title>
          </caption>
          <table frame="hsides" rules="rows">
            <colgroup width="16%">
              <col/>
              <col/>
              <col/>
              <col/>
              <col/>
              <col/>
            </colgroup>
            <thead style="border-top: thin solid; border-bottom: thin solid; border-color: #000000">
              <tr>
                <th valign="top" align="left" rowspan="2">Sexo</th>
                <th valign="top" align="center" rowspan="2">Desfechos</th>
                <th valign="top" align="center" rowspan="2">Mediador (&#205;ndice de obesidade central)</th>
                <th valign="top" align="center">Efeito direto controlado</th>
                <th valign="top" align="center">Propor&#231;&#227;o eliminada</th>
                <th valign="top" align="center" rowspan="2">Efeito total (ETA) (IC95%)</th>
              </tr>
              <tr>
                <th valign="top" align="center">ETA (IC95%)</th>
                <th valign="top" align="center">PE % (IC95%)</th>
              </tr>
            </thead>
            <tbody style="border-bottom: thin solid; border-color: #000000">
              <tr>
                <td valign="middle" align="left" rowspan="12">Masculino</td>
                <td valign="top" align="center" rowspan="3">DAC</td>
                <td valign="top" align="center">CC</td>
                <td valign="top" align="center">-4,32 (-7,69, -0,96)</td>
                <td valign="top" align="center">144 (124,161)</td>
                <td valign="top" align="center" rowspan="3">9,83 (5,74, 13,92)</td>
              </tr>
              <tr>
                <td valign="top" align="center">RCQ</td>
                <td valign="top" align="center">-32,28 (-36,07, -28,48)</td>
                <td valign="top" align="center">428 (408,439)</td>
              </tr>
              <tr>
                <td valign="top" align="center">Quadril</td>
                <td valign="top" align="center">2,86 (-2,54, 8,26)</td>
                <td valign="top" align="center">71 (62,81)</td>
              </tr>
              <tr>
                <td valign="top" align="center" rowspan="3">IC</td>
                <td valign="top" align="center">CC</td>
                <td valign="top" align="center">4,94 (1,31, 8,56)</td>
                <td valign="top" align="center">69 (57,79)</td>
                <td valign="top" align="center" rowspan="3">15,76 (12,04, 19,49)</td>
              </tr>
              <tr>
                <td valign="top" align="center">RCQ</td>
                <td valign="top" align="center">-11,20 (-14,98, -7,42)</td>
                <td valign="top" align="center">171 (158,186)</td>
              </tr>
              <tr>
                <td valign="top" align="center">Quadril</td>
                <td valign="top" align="center">15,06 (10,32, 19,79)</td>
                <td valign="top" align="center">4 (0,07,11)</td>
              </tr>
              <tr>
                <td valign="top" align="center" rowspan="3">AVC</td>
                <td valign="top" align="center">CC</td>
                <td valign="top" align="center">1,14 (-1,63, 3,91)</td>
                <td valign="top" align="center">86 (75,107)</td>
                <td valign="top" align="center" rowspan="3">8,10 (3,84, 12,37)</td>
              </tr>
              <tr>
                <td valign="top" align="center">RCQ</td>
                <td valign="top" align="center">-6,57 (-10,63, -2,50)</td>
                <td valign="top" align="center">181 (174,201)</td>
              </tr>
              <tr>
                <td valign="top" align="center">Quadril</td>
                <td valign="top" align="center">1,13 (-4,42, 6,69)</td>
                <td valign="top" align="center">86 (76,95)</td>
              </tr>
              <tr>
                <td valign="top" align="center" rowspan="3">Mortalidade (todas as causas)</td>
                <td valign="top" align="center">CC</td>
                <td valign="top" align="center">5,57 (1,38, 9,77)</td>
                <td valign="top" align="center">49 (37,68)</td>
                <td valign="top" align="center" rowspan="3">10,89 (8,12, 13,65)</td>
              </tr>
              <tr>
                <td valign="top" align="center">RCQ</td>
                <td valign="top" align="center">-23,53 (-27,02, -20,04)</td>
                <td valign="top" align="center">316 (301,329)</td>
              </tr>
              <tr>
                <td valign="top" align="center">Quadril</td>
                <td valign="top" align="center">6,00 (1,54, 10,46)</td>
                <td valign="top" align="center">45 (36,57)</td>
              </tr>
              <tr>
                <td valign="middle" align="left" rowspan="12">Feminino</td>
                <td valign="top" align="center" rowspan="3">DAC</td>
                <td valign="top" align="center">CC</td>
                <td valign="top" align="center">5,02 (3,11, 6,93)</td>
                <td valign="top" align="center">57 (43,69)</td>
                <td valign="top" align="center" rowspan="3">11,78 (8,70, 14,86)</td>
              </tr>
              <tr>
                <td valign="top" align="center">RCQ</td>
                <td valign="top" align="center">1,13 (-2,67, 4,92)</td>
                <td valign="top" align="center">90 (79,103)</td>
              </tr>
              <tr>
                <td valign="top" align="center">Quadril</td>
                <td valign="top" align="center">4,68 (-0,25, 9,62)</td>
                <td valign="top" align="center">60 (52,67)</td>
              </tr>
              <tr>
                <td valign="top" align="center" rowspan="3">IC</td>
                <td valign="top" align="center">CC</td>
                <td valign="top" align="center">11,57 (9,18, 13,96)</td>
                <td valign="top" align="center">30 (15,39)</td>
                <td valign="top" align="center" rowspan="3">16,66 (13,79,19,53)</td>
              </tr>
              <tr>
                <td valign="top" align="center">RCQ</td>
                <td valign="top" align="center">9,36 (4,01, 14,70)</td>
                <td valign="top" align="center">44 (31,52)</td>
              </tr>
              <tr>
                <td valign="top" align="center">Quadril</td>
                <td valign="top" align="center">8,28 (3,66, 12,89)</td>
                <td valign="top" align="center">50 (37,61)</td>
              </tr>
              <tr>
                <td valign="top" align="center" rowspan="3">AVC</td>
                <td valign="top" align="center">CC</td>
                <td valign="top" align="center">0,06 (-1,92, 2,05)</td>
                <td valign="top" align="center">99 (89,111)</td>
                <td valign="top" align="center" rowspan="3">7,70 (4,73, 10,68)</td>
              </tr>
              <tr>
                <td valign="top" align="center">RCQ</td>
                <td valign="top" align="center">0,23 (-4,09, 4,55)</td>
                <td valign="top" align="center">97 (87,113)</td>
              </tr>
              <tr>
                <td valign="top" align="center">Quadril</td>
                <td valign="top" align="center">4,04 (0,90, 7,19)</td>
                <td valign="top" align="center">47 (34,63)</td>
              </tr>
              <tr>
                <td valign="top" align="center" rowspan="3">Mortalidade (todas as causas)</td>
                <td valign="top" align="center">CC</td>
                <td valign="top" align="center">5,24 (2,91, 7,57)</td>
                <td valign="top" align="center">53 (41,66)</td>
                <td valign="top" align="center" rowspan="3">11,24 (8,76, 13,73)</td>
              </tr>
              <tr>
                <td valign="top" align="center">RCQ</td>
                <td valign="top" align="center">5,22 (-0,72, 11,16)</td>
                <td valign="top" align="center">53 (42,64)</td>
              </tr>
              <tr>
                <td valign="top" align="center">Quadril</td>
                <td valign="top" align="center">1,67 (-3,43, 6,77)</td>
                <td valign="top" align="center">85 (77,98)</td>
              </tr>
            </tbody>
          </table>
          <table-wrap-foot>
            <fn id="TFN3">
              <p><italic>ETA: efeito de tratamento aditivo; PE, propor&#231;&#227;o eliminada; DAC: doen&#231;a arterial coronariana; IC: insufici&#234;ncia card&#237;aca; AVC: acidente vascular cerebral; CC: circunfer&#234;ncia da cintura; RCQ: rela&#231;&#227;o cintura-quadril; Quadril: circunfer&#234;ncia do quadril</italic>.</p>
            </fn>
          </table-wrap-foot>
        </table-wrap>
        <p>Em rela&#231;&#227;o ao sexo, os resultados mostram associa&#231;&#227;o forte e significativa entre o IMC e todos os desfechos, tanto para homens quanto para mulheres. Um forte efeito direto controlado para IC em homens e mulheres, exceto para o RCQ em homens nos &#237;ndices de obesidade central (ETA entre = 4,94 e 15,06), sugere que, mesmo se a obesidade central fosse eliminada, um grande efeito permaneceria para o IMC. Por outro lado, um leve efeito direto controlado para AVC em homens e mulheres, exceto para a circunfer&#234;ncia do quadril em mulheres (ETA entre = -6,27 e 1,14) e para DAC em homens (ETA entre = -32,28 a 2,86) nos &#237;ndices de obesidade central, sugere que, se a obesidade central fosse eliminada, seria eliminado um forte efeito do IMC e, em alguns casos, o efeito do IMC seria revertido (protetor).</p>
      </sec>
      <sec>
        <title>Propor&#231;&#227;o Eliminada</title>
        <p>O &#237;ndice PE para TEs do IMC, para todos os resultados de interesse com intervalos de confian&#231;a de 95% para todos os participantes e sexos, est&#225; listado nas <xref ref-type="table" rid="t2">Tabelas 2</xref> e <xref ref-type="table" rid="t3">3</xref>. A associa&#231;&#227;o total de IMC com DAC pode ser completamente eliminada pela elimina&#231;&#227;o do papel da CC, em 127%. Este efeito pode ser reduzido em 94% e 71%, eliminando o papel da RCQ e da circunfer&#234;ncia do quadril, respectivamente. Em rela&#231;&#227;o ao AVC, o efeito do IMC poderia ser eliminado ao retirar o papel da circunfer&#234;ncia do quadril, em 99%. Com rela&#231;&#227;o &#224; IC e &#224; mortalidade por todas as causas, o papel dos &#237;ndices de obesidade central na elimina&#231;&#227;o do efeito do IMC foi um pouco semelhante, entre 50% e 61%. Com rela&#231;&#227;o ao sexo, a associa&#231;&#227;o total do IMC com CC, IC, AVC e mortalidade por todas as causas em homens poderia ser completamente eliminada removendo-se o papel da RCQ, em 428%, 171%, 181% e 316%, respectivamente. Por outro lado, em mulheres, a associa&#231;&#227;o total do IMC com CC, IC, AVC e mortalidade por todas as causas n&#227;o p&#244;de ser completamente eliminada ao remover o papel de quaisquer &#237;ndices de obesidade central (entre 30% para &#237;ndice de CC em IC e 99% para &#237;ndice de CC em AVC).</p>
      </sec>
    </sec>
    <sec sec-type="discussion">
      <title>Discuss&#227;o</title>
      <p>Neste grande estudo de coorte baseado em comunidades, os TEs e CDEs de IMC relacionados ao risco de DAC, IC, AVC e mortalidade por todas as causas em participantes sem obesidade central foram avaliados usando o m&#233;todo TMLE. &#201; importante mencionar que consideramos duas limita&#231;&#245;es comuns do IMC e dos estimadores convencionais, incluindo a capacidade limitada do IMC de distinguir entre massa gorda e massa livre de gordura, o que resulta em classifica&#231;&#227;o incorreta e especifica&#231;&#227;o incorreta do modelo, uma fonte comum de vi&#233;s em estimativas convencionais.</p>
      <p>Em resumo, em compara&#231;&#227;o com os TEs de IMC, os CDEs de IMC entre os participantes sem obesidade central para todos os desfechos de interesse foram atenuados e pr&#243;ximos a nulo. Esses resultados s&#227;o mais destacados para DAC e AVC. Esse achado destaca a capacidade dos &#237;ndices de obesidade central de prever o risco de doen&#231;as cardiovasculares e mortalidade por todas as causas. Al&#233;m disso, em rela&#231;&#227;o aos tr&#234;s &#237;ndices de obesidade central para todos os participantes, a propor&#231;&#227;o eliminada dos efeitos do IMC n&#227;o foi consistente para todos os desfechos. A propor&#231;&#227;o eliminada dos efeitos do IMC foi mais tang&#237;vel para o &#237;ndice RCQ em homens, enquanto os resultados n&#227;o foram consistentes para mulheres. Em geral, para a maioria dos desfechos, os resultados mostraram que, com a redu&#231;&#227;o ou elimina&#231;&#227;o da obesidade central com base no &#237;ndice de RCQ, o efeito do IMC foi completamente ou quase todo removido.</p>
      <p>Al&#233;m disso, esses achados destacam as limita&#231;&#245;es do IMC em prever o risco cardiovascular como um todo ou com base no sexo. Essa discord&#226;ncia do IMC em rela&#231;&#227;o aos desfechos cardiovasculares foi considerada &#8220;o paradoxo da obesidade&#8221;. V&#225;rias explica&#231;&#245;es para o paradoxo da obesidade relacionado &#224; associa&#231;&#227;o do IMC com doen&#231;as cardiovasculares t&#234;m sido relatadas. Uma das mais importantes refere-se ao erro de classifica&#231;&#227;o dos n&#237;veis de obesidade com base na defini&#231;&#227;o do IMC<xref ref-type="bibr" rid="B4"><sup>4</sup></xref>. Considerando que o IMC &#233; incapaz de discriminar entre massa gorda, massa muscular e superf&#237;cie corporal, o efeito do IMC &#233; uma combina&#231;&#227;o desses tr&#234;s tipos de massa.<xref ref-type="bibr" rid="B17"><sup>17</sup></xref> Portanto, um IMC mais elevado &#233; um indicador n&#227;o s&#243; de maior quantidade de gordura central e visceral, mas tamb&#233;m de maior massa muscular ou perif&#233;rica (gordura ou osso).</p>
      <p>Nas &#250;ltimas d&#233;cadas, muitos estudos avaliaram a associa&#231;&#227;o de diferentes distribui&#231;&#245;es de gordura com doen&#231;as cardiovasculares, demonstrando que a gordura corporal, principalmente o excesso de gordura central, independentemente da gordura corporal total, &#233; um importante fator de risco para esses desfechos.<xref ref-type="bibr" rid="B18"><sup>18</sup></xref><sup>&#8211;</sup><xref ref-type="bibr" rid="B22"><sup>22</sup></xref> A esse respeito, estudos anteriores mostraram que o excesso de gordura em homens &#233; comumente armazenado nas partes viscerais, enquanto nas mulheres &#233; armazenado em partes subcut&#226;neas perif&#233;ricas.<xref ref-type="bibr" rid="B23"><sup>23</sup></xref><sup>,</sup><xref ref-type="bibr" rid="B24"><sup>24</sup></xref> Os resultados do presente estudo confirmam os achados anteriores e refor&#231;am a import&#226;ncia da distribui&#231;&#227;o de gordura para homens e mulheres separadamente.</p>
      <p>Em rela&#231;&#227;o ao m&#233;todo estat&#237;stico em uso, estudos metodol&#243;gicos anteriores e originais confirmam a superioridade do m&#233;todo TMLE sobre outras abordagens regularmente utilizadas em estudos observacionais para medir causalidade. Nesse sentido, o m&#233;todo de pondera&#231;&#227;o pela probabilidade inversa (IPW, Inverse Probability Weighting) resulta em estimativas inst&#225;veis na presen&#231;a de pesos extremos e viola&#231;&#245;es da suposi&#231;&#227;o de positividade. Por outro lado, comparado ao m&#233;todo TMLE, o m&#233;todo G s&#243; &#233; realizado com base no modelo de resultado e, se especificado incorretamente, traz estimativas enviesadas. TMLE &#233; um estimador duplamente robusto que permanece consistente se a exposi&#231;&#227;o ou os mecanismos de resultado forem estimados de forma consistente.<xref ref-type="bibr" rid="B9"><sup>9</sup></xref><sup>,</sup><xref ref-type="bibr" rid="B10"><sup>10</sup></xref><sup>,</sup><xref ref-type="bibr" rid="B25"><sup>25</sup></xref></p>
      <p>Trabalhos anteriores confirmam a utilidade do efeito direto controlado, especialmente na avalia&#231;&#227;o de pol&#237;ticas.<xref ref-type="bibr" rid="B26"><sup>26</sup></xref><sup>,</sup><xref ref-type="bibr" rid="B27"><sup>27</sup></xref> No entanto, o uso desse conceito necessita de outras premissas que n&#227;o as TEs.<xref ref-type="bibr" rid="B16"><sup>16</sup></xref><sup>,</sup><xref ref-type="bibr" rid="B26"><sup>26</sup></xref> Na an&#225;lise direta controlada, deve-se considerar a suposi&#231;&#227;o para a associa&#231;&#227;o entre mediador e desfecho, bem como a associa&#231;&#227;o entre exposi&#231;&#227;o e desfecho.<xref ref-type="bibr" rid="B28"><sup>28</sup></xref> Al&#233;m disso, a intera&#231;&#227;o entre a exposi&#231;&#227;o e os mediadores &#233; uma quest&#227;o importante nesta an&#225;lise.<xref ref-type="bibr" rid="B16"><sup>16</sup></xref><sup>,</sup><xref ref-type="bibr" rid="B28"><sup>28</sup></xref> Com rela&#231;&#227;o a essa quest&#227;o, n&#227;o podemos usar a diferen&#231;a entre TE e CDE para estimar os efeitos diretos e indiretos.</p>
      <p>Em resumo, com base em estudos metodol&#243;gicos anteriores, no que diz respeito &#224; limita&#231;&#227;o do efeito direto controlado e &#224; necessidade de premissas mais fortes, isso n&#227;o pode ser usado como estimativa v&#225;lida de media&#231;&#227;o; mas se tivermos um efeito mediado controlado diferente de zero, pode ser sugestivo da presen&#231;a de efeito mediador.<xref ref-type="bibr" rid="B28"><sup>28</sup></xref></p>
      <p>Antes de interpretar os resultados e chegar a qualquer conclus&#227;o, os pontos fortes e as limita&#231;&#245;es deste estudo devem ser considerados. Os pontos fortes incluem a aplica&#231;&#227;o de um m&#233;todo duplamente robusto que estima consistentemente o par&#226;metro em um modelo semiparam&#233;trico quando um dos dois modelos (exposi&#231;&#227;o e resultado) &#233; especificado corretamente, independentemente de qual. Al&#233;m disso, consideramos o mecanismo ausente para minimizar o impacto de um risco competitivo e perda de acompanhamento para estimar melhor os efeitos reais. No entanto, devido ao tamanho pequeno da amostra do resultado de interesse e ao vi&#233;s de dados esparsos, n&#227;o foi poss&#237;vel avaliar essas estimativas de acordo com faixa et&#225;ria. Este estudo &#233; limitado, ainda, pelo fato de que n&#227;o consideramos a varia&#231;&#227;o dos fatores confundidores vari&#225;veis no tempo.</p>
    </sec>
    <sec sec-type="conclusions">
      <title>Conclus&#227;o</title>
      <p>Neste estudo, o efeito direto controlado do IMC caiu para quase nulo em participantes sem obesidade central. Esses resultados destacam a import&#226;ncia de considerar a distribui&#231;&#227;o das massas de gordura ao estimar a associa&#231;&#227;o entre obesidade e um desfecho de interesse, para homens e mulheres separadamente.</p>
    </sec>
  </body>
  <back>
    <fn-group>
      <fn fn-type="supported-by" id="fn1">
        <p>
          <bold>Fontes de financiamento</bold>
        </p>
        <p>O presente estudo n&#227;o contou com fontes de financiamento externas.</p>
      </fn>
      <fn fn-type="other" id="fn2">
        <p>
          <bold>Vincula&#231;&#227;o acad&#234;mica</bold>
        </p>
        <p>Este artigo &#233; parte de tese de doutorado de Hossein Mozafar Saadati pela Shahid Beheshti University of Medical Sciences.</p>
      </fn>
      <fn fn-type="other" id="fn3">
        <p>
          <bold>Aprova&#231;&#227;o &#233;tica e consentimento informado</bold>
        </p>
        <p>Este estudo foi aprovado pelo Comit&#234; de &#201;tica do vicechancellor in research affair-Shahid Beheshti University of Medical Sciences sob o n&#250;mero de protocolo IR.SBMU.RETECH.REC.1399.763. Todos os procedimentos envolvidos nesse estudo est&#227;o de acordo com a Declara&#231;&#227;o de Helsinki de 1975, atualizada em 2013. O consentimento informado foi obtido de todos os participantes inclu&#237;dos no estudo.</p>
      </fn>
      <fn fn-type="other" id="fn4">
        <p>
          <bold>*Material suplementar</bold>
        </p>
        <p>Para informa&#231;&#227;o adicional, por favor, clique aqui.</p>
      </fn>
    </fn-group>
    <ack>
      <title>Agradecimentos</title>
      <p>Este estudo est&#225; relacionado ao projeto n&#186; IR.SBMU.PHNS.REC.1396.152 da Shahid Beheshti University of Medical Sciences (SBUMS), Teer&#227;, Ir&#227;. Os conselhos de revis&#227;o institucional de cada local aprovaram o protocolo do estudo ARIC e um termo de consentimento livre e informado foi obtido dos participantes em todas as visitas do estudo. Os autores gostariam de agradecer &#224; equipe e aos participantes do estudo ARIC por suas importantes contribui&#231;&#245;es. Para acessar os dados, assinamos um RMDA (NHLBI, Research Materials Distribution Agreement), dispon&#237;vel mediante solicita&#231;&#227;o.</p>
    </ack>
    <ref-list>
      <title>Refer&#234;ncias</title>
      <ref id="B1">
        <label>1</label>
        <element-citation publication-type="journal">
          <person-group person-group-type="author">
            <name>
              <surname>Nuttall</surname>
              <given-names>FQ</given-names>
            </name>
          </person-group>
          <article-title>Body mass index: obesity, BMI, and health: a critical review</article-title>
          <source>Nutrition today</source>
          <year>2015</year>
          <volume>50</volume>
          <issue>3</issue>
          <fpage>117</fpage>
          <lpage>117</lpage>
        </element-citation>
        <mixed-citation>1. Nuttall FQ. Body mass index: obesity, BMI, and health: a critical review. Nutrition today. 2015;50(3):117.</mixed-citation>
      </ref>
      <ref id="B2">
        <label>2</label>
        <element-citation publication-type="journal">
          <person-group person-group-type="author">
            <name>
              <surname>Buss</surname>
              <given-names>J</given-names>
            </name>
          </person-group>
          <article-title>Limitations of body mass index to assess body fat</article-title>
          <source>Workplace Health &amp; Safety</source>
          <year>2014</year>
          <volume>62</volume>
          <issue>6</issue>
          <fpage>264</fpage>
          <lpage>264</lpage>
        </element-citation>
        <mixed-citation>2. Buss J. Limitations of body mass index to assess body fat. Workplace Health &amp; Safety. 2014;62(6):264</mixed-citation>
      </ref>
      <ref id="B3">
        <label>3</label>
        <element-citation publication-type="journal">
          <person-group person-group-type="author">
            <name>
              <surname>Frankenfield</surname>
              <given-names>DC</given-names>
            </name>
            <name>
              <surname>Rowe</surname>
              <given-names>WA</given-names>
            </name>
            <name>
              <surname>Cooney</surname>
              <given-names>RN</given-names>
            </name>
            <name>
              <surname>Smith</surname>
              <given-names>JS</given-names>
            </name>
            <name>
              <surname>Becker</surname>
              <given-names>D</given-names>
            </name>
          </person-group>
          <article-title>Limits of body mass index to detect obesity and predict body composition</article-title>
          <source>Nutrition</source>
          <year>2001</year>
          <volume>17</volume>
          <issue>1</issue>
          <fpage>26</fpage>
          <lpage>30</lpage>
        </element-citation>
        <mixed-citation>3. Frankenfield DC, Rowe WA, Cooney RN, Smith JS, Becker D. Limits of body mass index to detect obesity and predict body composition. Nutrition. 2001;17(1):26-30.</mixed-citation>
      </ref>
      <ref id="B4">
        <label>4</label>
        <element-citation publication-type="journal">
          <person-group person-group-type="author">
            <name>
              <surname>Romero-Corral</surname>
              <given-names>A</given-names>
            </name>
            <name>
              <surname>Somers</surname>
              <given-names>VK</given-names>
            </name>
            <name>
              <surname>Sierra-Johnson</surname>
              <given-names>J</given-names>
            </name>
            <name>
              <surname>Thomas</surname>
              <given-names>RJ</given-names>
            </name>
            <name>
              <surname>Collazo-Clavell</surname>
              <given-names>M</given-names>
            </name>
            <name>
              <surname>Korinek</surname>
              <given-names>Jec</given-names>
            </name>
            <etal/>
          </person-group>
          <article-title>Accuracy of body mass index in diagnosing obesity in the adult general population</article-title>
          <source>Int J Obes</source>
          <year>2008</year>
          <volume>32</volume>
          <issue>6</issue>
          <fpage>959</fpage>
          <lpage>966</lpage>
        </element-citation>
        <mixed-citation>4. Romero-Corral A, Somers VK, Sierra-Johnson J, Thomas RJ, Collazo-Clavell M, Korinek Jec, et al. Accuracy of body mass index in diagnosing obesity in the adult general population. Int J Obes.. 2008;32(6):959-66.</mixed-citation>
      </ref>
      <ref id="B5">
        <label>5</label>
        <element-citation publication-type="journal">
          <person-group person-group-type="author">
            <name>
              <surname>Lee</surname>
              <given-names>CMY</given-names>
            </name>
            <name>
              <surname>Huxley</surname>
              <given-names>RR</given-names>
            </name>
            <name>
              <surname>Wildman</surname>
              <given-names>RP</given-names>
            </name>
            <name>
              <surname>Woodward</surname>
              <given-names>M</given-names>
            </name>
          </person-group>
          <article-title>Indices of abdominal obesity are better discriminators of cardiovascular risk factors than BMI: a meta-analysis</article-title>
          <source>J Clin Epidemiol</source>
          <year>2008</year>
          <volume>61</volume>
          <issue>7</issue>
          <fpage>646</fpage>
          <lpage>653</lpage>
        </element-citation>
        <mixed-citation>5. Lee CMY, Huxley RR, Wildman RP, Woodward M. Indices of abdominal obesity are better discriminators of cardiovascular risk factors than BMI: a meta-analysis.J Clin Epidemiol. 2008;61(7):646-53.</mixed-citation>
      </ref>
      <ref id="B6">
        <label>6</label>
        <element-citation publication-type="journal">
          <person-group person-group-type="author">
            <name>
              <surname>Li</surname>
              <given-names>C</given-names>
            </name>
            <name>
              <surname>Ford</surname>
              <given-names>ES</given-names>
            </name>
            <name>
              <surname>McGuire</surname>
              <given-names>LC</given-names>
            </name>
            <name>
              <surname>Mokdad</surname>
              <given-names>AH</given-names>
            </name>
          </person-group>
          <article-title>Increasing trends in waist circumference and abdominal obesity among US adults</article-title>
          <source>Obesity</source>
          <year>2007</year>
          <volume>15</volume>
          <issue>1</issue>
          <fpage>216</fpage>
          <lpage>216</lpage>
        </element-citation>
        <mixed-citation>6. Li C, Ford ES, McGuire LC, Mokdad AH. Increasing trends in waist circumference and abdominal obesity among US adults. Obesity. 2007;15(1):216</mixed-citation>
      </ref>
      <ref id="B7">
        <label>7</label>
        <element-citation publication-type="journal">
          <person-group person-group-type="author">
            <name>
              <surname>Chrysant</surname>
              <given-names>SG</given-names>
            </name>
            <name>
              <surname>Chrysant</surname>
              <given-names>GS</given-names>
            </name>
          </person-group>
          <article-title>The single use of body mass index for the obesity paradox is misleading and should be used in conjunction with other obesity indices</article-title>
          <source>Postgr Med</source>
          <year>2019</year>
          <volume>131</volume>
          <issue>2</issue>
          <fpage>96</fpage>
          <lpage>102</lpage>
        </element-citation>
        <mixed-citation>7. Chrysant SG, Chrysant GS. The single use of body mass index for the obesity paradox is misleading and should be used in conjunction with other obesity indices. Postgr Med. 2019;131(2):96-102.</mixed-citation>
      </ref>
      <ref id="B8">
        <label>8</label>
        <element-citation publication-type="journal">
          <person-group person-group-type="author">
            <name>
              <surname>Iliodromiti</surname>
              <given-names>S</given-names>
            </name>
            <name>
              <surname>Celis-Morales</surname>
              <given-names>CA</given-names>
            </name>
            <name>
              <surname>Lyall</surname>
              <given-names>DM</given-names>
            </name>
            <name>
              <surname>Anderson</surname>
              <given-names>J</given-names>
            </name>
            <name>
              <surname>Gray</surname>
              <given-names>SR</given-names>
            </name>
            <name>
              <surname>Mackay</surname>
              <given-names>DF</given-names>
            </name>
            <etal/>
          </person-group>
          <article-title>The impact of confounding on the associations of different adiposity measures with the incidence of cardiovascular disease: a cohort study of 296 535 adults of white European descent</article-title>
          <source>Eur Heart J</source>
          <year>2018</year>
          <volume>39</volume>
          <issue>17</issue>
          <fpage>1514</fpage>
          <lpage>1520</lpage>
        </element-citation>
        <mixed-citation>8. Iliodromiti S, Celis-Morales CA, Lyall DM, Anderson J, Gray SR, Mackay DF, et al. The impact of confounding on the associations of different adiposity measures with the incidence of cardiovascular disease: a cohort study of 296 535 adults of white European descent. Eur Heart J.2018;39(17):1514-20.</mixed-citation>
      </ref>
      <ref id="B9">
        <label>9</label>
        <element-citation publication-type="journal">
          <person-group person-group-type="author">
            <name>
              <surname>Mansournia</surname>
              <given-names>MA</given-names>
            </name>
            <name>
              <surname>Etminan</surname>
              <given-names>M</given-names>
            </name>
            <name>
              <surname>Danaei</surname>
              <given-names>G</given-names>
            </name>
            <name>
              <surname>Kaufman</surname>
              <given-names>JS</given-names>
            </name>
            <name>
              <surname>Collins</surname>
              <given-names>G</given-names>
            </name>
          </person-group>
          <article-title>Handling time varying confounding in observational research</article-title>
          <source>BMJ</source>
          <year>2017</year>
          <volume>359</volume>
          <fpage>j4587</fpage>
          <lpage>j4587</lpage>
        </element-citation>
        <mixed-citation>9. Mansournia MA, Etminan M, Danaei G, Kaufman JS, Collins G. Handling time varying confounding in observational research.BMJ. 2017;359:j4587.</mixed-citation>
      </ref>
      <ref id="B10">
        <label>10</label>
        <element-citation publication-type="journal">
          <person-group person-group-type="author">
            <name>
              <surname>Van der Laan</surname>
              <given-names>MJ</given-names>
            </name>
            <name>
              <surname>Rose</surname>
              <given-names>S</given-names>
            </name>
          </person-group>
          <article-title>Targeted learning: causal inference for observational and experimental data</article-title>
          <source>Philadelphia:Springer Science &amp; Business Media</source>
          <year>2011</year>
        </element-citation>
        <mixed-citation>10. Van der Laan MJ, Rose S. Targeted learning: causal inference for observational and experimental data. Philadelphia:Springer Science &amp; Business Media; 2011.</mixed-citation>
      </ref>
      <ref id="B11">
        <label>11</label>
        <element-citation publication-type="journal">
          <person-group person-group-type="author">
            <collab>The Atherosclerosis Risk in Communities (ARIC) Study</collab>
          </person-group>
          <article-title>design and objectives. The ARIC investigators</article-title>
          <source>Am J Epidemiol</source>
          <year>1989</year>
          <volume>129</volume>
          <issue>4</issue>
          <fpage>687</fpage>
          <lpage>702</lpage>
        </element-citation>
        <mixed-citation>11. The Atherosclerosis Risk in Communities (ARIC) Study: design and objectives. The ARIC investigators. Am J Epidemiol. 1989;129(4):687-702.</mixed-citation>
      </ref>
      <ref id="B12">
        <label>12</label>
        <element-citation publication-type="journal">
          <person-group person-group-type="author">
            <collab>World Health Organization</collab>
          </person-group>
          <article-title>Waist circumference and waist-hip ratio Geneva 2011</article-title>
          <source>Report of a WHO Expert Consultation</source>
          <month>12</month>
          <year>2008</year>
        </element-citation>
        <mixed-citation>12. World Health Organization. Waist circumference and waist-hip ratio Geneva; 2011. (Report of a WHO Expert Consultation Dec.2008)</mixed-citation>
      </ref>
      <ref id="B13">
        <label>13</label>
        <element-citation publication-type="journal">
          <person-group person-group-type="author">
            <name>
              <surname>Almasi-Hashiani</surname>
              <given-names>A</given-names>
            </name>
            <name>
              <surname>Nedjat</surname>
              <given-names>S</given-names>
            </name>
            <name>
              <surname>Mansournia</surname>
              <given-names>MA</given-names>
            </name>
          </person-group>
          <article-title>Causal Methods for Observational Research: A Primer</article-title>
          <source>Arch Iran Med. (AIM)</source>
          <year>2018</year>
          <volume>21</volume>
          <issue>4</issue>
          <fpage>164</fpage>
          <lpage>169</lpage>
        </element-citation>
        <mixed-citation>13. Almasi-Hashiani A, Nedjat S, Mansournia MA. Causal Methods for Observational Research: A Primer. Arch Iran Med. (AIM). 2018;21(4):164-9.</mixed-citation>
      </ref>
      <ref id="B14">
        <label>14</label>
        <element-citation publication-type="journal">
          <person-group person-group-type="author">
            <name>
              <surname>Acharya</surname>
              <given-names>A</given-names>
            </name>
            <name>
              <surname>Blackwell</surname>
              <given-names>M</given-names>
            </name>
            <name>
              <surname>Sen</surname>
              <given-names>M</given-names>
            </name>
          </person-group>
          <article-title>Explaining causal findings without bias: Detecting and assessing direct effects</article-title>
          <source>Am Polit Scienc Rev</source>
          <year>2016</year>
          <volume>110</volume>
          <issue>3</issue>
          <fpage>512</fpage>
          <lpage>529</lpage>
        </element-citation>
        <mixed-citation>14. Acharya A, Blackwell M, Sen M. Explaining causal findings without bias: Detecting and assessing direct effects. Am Polit Scienc Rev. 2016;110(3):512-29.</mixed-citation>
      </ref>
      <ref id="B15">
        <label>15</label>
        <element-citation publication-type="journal">
          <person-group person-group-type="author">
            <name>
              <surname>Nguyen</surname>
              <given-names>TQ</given-names>
            </name>
            <name>
              <surname>Schmid</surname>
              <given-names>I</given-names>
            </name>
            <name>
              <surname>Stuart</surname>
              <given-names>EA</given-names>
            </name>
          </person-group>
          <article-title>Clarifying causal mediation analysis for the applied researcher: Defining effects based on what we want to learn</article-title>
          <source>arXiv preprint arXiv:190408515</source>
          <year>2019</year>
        </element-citation>
        <mixed-citation>15. Nguyen TQ, Schmid I, Stuart EA. Clarifying causal mediation analysis for the applied researcher: Defining effects based on what we want to learn. arXiv preprint arXiv:190408515. 2019.</mixed-citation>
      </ref>
      <ref id="B16">
        <label>16</label>
        <element-citation publication-type="book">
          <person-group person-group-type="author">
            <name>
              <surname>VanderWeele</surname>
              <given-names>T</given-names>
            </name>
          </person-group>
          <source>Explanation in causal inference: methods for mediation and interaction</source>
          <publisher-loc>Oxford</publisher-loc>
          <publisher-name>University Press</publisher-name>
          <year>2015</year>
        </element-citation>
        <mixed-citation>16. VanderWeele T. Explanation in causal inference: methods for mediation and interaction. Oxford: University Press; 2015.</mixed-citation>
      </ref>
      <ref id="B17">
        <label>17</label>
        <element-citation publication-type="journal">
          <person-group person-group-type="author">
            <name>
              <surname>Romero-Corral</surname>
              <given-names>A</given-names>
            </name>
            <name>
              <surname>Lopez-Jimenez</surname>
              <given-names>F</given-names>
            </name>
            <name>
              <surname>Sierra-Johnson</surname>
              <given-names>J</given-names>
            </name>
            <name>
              <surname>Somers</surname>
              <given-names>VK</given-names>
            </name>
          </person-group>
          <article-title>Differentiating between body fat and lean mass&#8212;how should we measure obesity?</article-title>
          <source>Nature Clin &amp; Metab</source>
          <year>2008</year>
          <volume>4</volume>
          <issue>6</issue>
          <fpage>322</fpage>
          <lpage>323</lpage>
        </element-citation>
        <mixed-citation>17. Romero-Corral A, Lopez-Jimenez F, Sierra-Johnson J, Somers VK. Differentiating between body fat and lean mass&#8212;how should we measure obesity? Nature Clin &amp; Metab. 2008;4(6):322-3.</mixed-citation>
      </ref>
      <ref id="B18">
        <label>18</label>
        <element-citation publication-type="journal">
          <person-group person-group-type="author">
            <name>
              <surname>Browning</surname>
              <given-names>LM</given-names>
            </name>
            <name>
              <surname>Hsieh</surname>
              <given-names>SD</given-names>
            </name>
            <name>
              <surname>Ashwell</surname>
              <given-names>M</given-names>
            </name>
          </person-group>
          <article-title>A systematic review of waist-to-height ratio as a screening tool for the prediction of cardiovascular disease and diabetes: 0&#183; 5 could be a suitable global boundary value</article-title>
          <source>Nutrit Res Rev</source>
          <year>2010</year>
          <volume>23</volume>
          <issue>2</issue>
          <fpage>247</fpage>
          <lpage>269</lpage>
        </element-citation>
        <mixed-citation>18. Browning LM, Hsieh SD, Ashwell M. A systematic review of waist-to-height ratio as a screening tool for the prediction of cardiovascular disease and diabetes: 0&#183; 5 could be a suitable global boundary value. Nutrit Res Rev. 2010;23(2):247-69.</mixed-citation>
      </ref>
      <ref id="B19">
        <label>19</label>
        <element-citation publication-type="journal">
          <person-group person-group-type="author">
            <name>
              <surname>Leitzmann</surname>
              <given-names>MF</given-names>
            </name>
            <name>
              <surname>Moore</surname>
              <given-names>SC</given-names>
            </name>
            <name>
              <surname>Koster</surname>
              <given-names>A</given-names>
            </name>
            <name>
              <surname>Harris</surname>
              <given-names>TB</given-names>
            </name>
            <name>
              <surname>Park</surname>
              <given-names>Y</given-names>
            </name>
            <name>
              <surname>Hollenbeck</surname>
              <given-names>A</given-names>
            </name>
            <etal/>
          </person-group>
          <article-title>Waist circumference as compared with body-mass index in predicting mortality from specific causes</article-title>
          <source>PloS one</source>
          <year>2011</year>
          <volume>6</volume>
          <issue>4</issue>
          <fpage>18</fpage>
          <lpage>582</lpage>
        </element-citation>
        <mixed-citation>19. Leitzmann MF, Moore SC, Koster A, Harris TB, Park Y, Hollenbeck A, et al. Waist circumference as compared with body-mass index in predicting mortality from specific causes. PloS one. 2011;6(4):18.582.</mixed-citation>
      </ref>
      <ref id="B20">
        <label>20</label>
        <element-citation publication-type="journal">
          <person-group person-group-type="author">
            <name>
              <surname>Shields</surname>
              <given-names>M</given-names>
            </name>
            <name>
              <surname>Tremblay</surname>
              <given-names>MS</given-names>
            </name>
            <name>
              <surname>Connor Gorber</surname>
              <given-names>S</given-names>
            </name>
            <name>
              <surname>Janssen</surname>
              <given-names>I</given-names>
            </name>
          </person-group>
          <article-title>Abdominal obesity and cardiovascular disease risk factors within body mass index categories</article-title>
          <source>Health Rep</source>
          <year>2012</year>
          <volume>23</volume>
          <issue>2</issue>
          <fpage>7</fpage>
          <lpage>15</lpage>
        </element-citation>
        <mixed-citation>20. Shields M, Tremblay MS, Connor Gorber S, Janssen I. Abdominal obesity and cardiovascular disease risk factors within body mass index categories. Health Rep. 2012;23(2):7-15.</mixed-citation>
      </ref>
      <ref id="B21">
        <label>21</label>
        <element-citation publication-type="journal">
          <person-group person-group-type="author">
            <name>
              <surname>Staiano</surname>
              <given-names>A</given-names>
            </name>
            <name>
              <surname>Reeder</surname>
              <given-names>B</given-names>
            </name>
            <name>
              <surname>Elliott</surname>
              <given-names>S</given-names>
            </name>
            <name>
              <surname>Joffres</surname>
              <given-names>M</given-names>
            </name>
            <name>
              <surname>Pahwa</surname>
              <given-names>P</given-names>
            </name>
            <name>
              <surname>Kirkland</surname>
              <given-names>S</given-names>
            </name>
            <etal/>
          </person-group>
          <article-title>Body mass index versus waist circumference as predictors of mortality in Canadian adults</article-title>
          <source>Int J Obes</source>
          <year>2012</year>
          <volume>36</volume>
          <issue>11</issue>
          <fpage>1450</fpage>
          <lpage>1454</lpage>
        </element-citation>
        <mixed-citation>21. Staiano A, Reeder B, Elliott S, Joffres M, Pahwa P, Kirkland S, et al. Body mass index versus waist circumference as predictors of mortality in Canadian adults. Int J Obes. 2012;36(11):1450-4.</mixed-citation>
      </ref>
      <ref id="B22">
        <label>22</label>
        <element-citation publication-type="journal">
          <person-group person-group-type="author">
            <name>
              <surname>Van Dijk</surname>
              <given-names>S</given-names>
            </name>
            <name>
              <surname>Takken</surname>
              <given-names>T</given-names>
            </name>
            <name>
              <surname>Prinsen</surname>
              <given-names>E</given-names>
            </name>
            <name>
              <surname>Wittink</surname>
              <given-names>H</given-names>
            </name>
          </person-group>
          <article-title>Different anthropometric adiposity measures and their association with cardiovascular disease risk factors: a meta-analysis</article-title>
          <source>Netherl Heart J</source>
          <year>2012</year>
          <volume>20</volume>
          <issue>5</issue>
          <fpage>208</fpage>
          <lpage>218</lpage>
        </element-citation>
        <mixed-citation>22. Van Dijk S, Takken T, Prinsen E, Wittink H. Different anthropometric adiposity measures and their association with cardiovascular disease risk factors: a meta-analysis. Netherl Heart J. 2012;20(5):208-18.</mixed-citation>
      </ref>
      <ref id="B23">
        <label>23</label>
        <element-citation publication-type="journal">
          <person-group person-group-type="author">
            <name>
              <surname>Clark</surname>
              <given-names>AL</given-names>
            </name>
            <name>
              <surname>Chyu</surname>
              <given-names>J</given-names>
            </name>
            <name>
              <surname>Horwich</surname>
              <given-names>TB</given-names>
            </name>
          </person-group>
          <article-title>The obesity paradox in men versus women with systolic heart failure</article-title>
          <source>Am J Cardiol</source>
          <year>2012</year>
          <volume>110</volume>
          <issue>1</issue>
          <fpage>77</fpage>
          <lpage>82</lpage>
        </element-citation>
        <mixed-citation>23. Clark AL, Chyu J, Horwich TB. The obesity paradox in men versus women with systolic heart failure. Am J Cardiol. 2012;110(1):77-82.</mixed-citation>
      </ref>
      <ref id="B24">
        <label>24</label>
        <element-citation publication-type="journal">
          <person-group person-group-type="author">
            <name>
              <surname>Lemieux</surname>
              <given-names>S</given-names>
            </name>
            <name>
              <surname>Prud&#8217;homme</surname>
              <given-names>D</given-names>
            </name>
            <name>
              <surname>Bouchard</surname>
              <given-names>C</given-names>
            </name>
            <name>
              <surname>Tremblay</surname>
              <given-names>A</given-names>
            </name>
            <name>
              <surname>Despr&#233;s</surname>
              <given-names>J-P</given-names>
            </name>
          </person-group>
          <article-title>Sex differences in the relation of visceral adipose tissue accumulation to total body fatness</article-title>
          <source>Am J Clin Nutr</source>
          <year>1993</year>
          <volume>58</volume>
          <issue>4</issue>
          <fpage>463</fpage>
          <lpage>467</lpage>
        </element-citation>
        <mixed-citation>24. Lemieux S, Prud&#8217;homme D, Bouchard C, Tremblay A, Despr&#233;s J-P. Sex differences in the relation of visceral adipose tissue accumulation to total body fatness. Am J Clin Nutr. 1993;58(4):463-7.</mixed-citation>
      </ref>
      <ref id="B25">
        <label>25</label>
        <element-citation publication-type="journal">
          <person-group person-group-type="author">
            <name>
              <surname>Pang</surname>
              <given-names>M</given-names>
            </name>
            <name>
              <surname>Schuster</surname>
              <given-names>T</given-names>
            </name>
            <name>
              <surname>Filion</surname>
              <given-names>KB</given-names>
            </name>
            <name>
              <surname>Schnitzer</surname>
              <given-names>ME</given-names>
            </name>
            <name>
              <surname>Eberg</surname>
              <given-names>M</given-names>
            </name>
            <name>
              <surname>Platt</surname>
              <given-names>RW</given-names>
            </name>
          </person-group>
          <article-title>Effect Estimation in Point-Exposure Studies with Binary Outcomes and High-Dimensional Covariate Data&#8211;A Comparison of Targeted Maximum Likelihood Estimation and Inverse Probability of Treatment Weighting</article-title>
          <source>Int J Biostat</source>
          <year>2016</year>
          <volume>12</volume>
          <issue>2</issue>
          <pub-id pub-id-type="doi">10.1515/ijb-2015-0034</pub-id>
        </element-citation>
        <mixed-citation>25. Pang M, Schuster T, Filion KB, Schnitzer ME, Eberg M, Platt RW. Effect Estimation in Point-Exposure Studies with Binary Outcomes and High-Dimensional Covariate Data&#8211;A Comparison of Targeted Maximum Likelihood Estimation and Inverse Probability of Treatment Weighting. Int J Biostat. 2016;12(2).Doi:10.1515/ijb-2015-0034</mixed-citation>
      </ref>
      <ref id="B26">
        <label>26</label>
        <element-citation publication-type="book">
          <person-group person-group-type="author">
            <name>
              <surname>Pearl</surname>
              <given-names>J</given-names>
            </name>
          </person-group>
          <chapter-title>Direct and indirect effects</chapter-title>
          <source>Proceedings of the Seventeenth Conference on Uncertainty and Artificial Intelligence; San Francisco</source>
          <publisher-loc>Morgan Kaufmann</publisher-loc>
          <year>2001</year>
          <fpage>411</fpage>
          <lpage>420</lpage>
        </element-citation>
        <mixed-citation>26. Pearl J. Direct and indirect effects. In: Proceedings of the Seventeenth Conference on Uncertainty and Artificial Intelligence; San Francisco: Morgan Kaufmann; 2001;p.411-20.</mixed-citation>
      </ref>
      <ref id="B27">
        <label>27</label>
        <element-citation publication-type="journal">
          <person-group person-group-type="author">
            <name>
              <surname>Robins</surname>
              <given-names>JM</given-names>
            </name>
          </person-group>
          <article-title>Semantics of causal DAG models and the identification of direct and indirect effects</article-title>
          <source>Oxford Statistical Science Series</source>
          <year>2003</year>
          <fpage>70</fpage>
          <lpage>82</lpage>
        </element-citation>
        <mixed-citation>27. Robins JM. Semantics of causal DAG models and the identification of direct and indirect effects. Oxford Statistical Science Series. 2003:70-82.</mixed-citation>
      </ref>
      <ref id="B28">
        <label>28</label>
        <element-citation publication-type="journal">
          <person-group person-group-type="author">
            <name>
              <surname>VanderWeele</surname>
              <given-names>TJ</given-names>
            </name>
          </person-group>
          <article-title>Controlled direct and mediated effects: definition, identification and bounds</article-title>
          <source>Scand J Stat</source>
          <year>2011</year>
          <volume>38</volume>
          <issue>3</issue>
          <fpage>551</fpage>
          <lpage>563</lpage>
        </element-citation>
        <mixed-citation>28. VanderWeele TJ. Controlled direct and mediated effects: definition, identification and bounds. Scand J Stat. 2011;38(3):551-63.</mixed-citation>
      </ref>
    </ref-list>
  </back>
  <sub-article article-type="translation" id="S01" xml:lang="en">
    <front-stub>
      <article-categories>
        <subj-group subj-group-type="heading">
          <subject>Original Article</subject>
        </subj-group>
      </article-categories>
      <title-group>
        <article-title>The Direct Effect of Body Mass Index on Cardiovascular Outcomes among Participants Without Central Obesity by Targeted Maximum Likelihood Estimation</article-title>
      </title-group>
      <contrib-group>
        <contrib contrib-type="author">
          <name>
            <surname>Saadati</surname>
            <given-names>Hossein Mozafar</given-names>
          </name>
          <xref ref-type="aff" rid="aff4">
            <sup>1</sup>
          </xref>
        </contrib>
        <contrib contrib-type="author">
          <contrib-id contrib-id-type="orcid">0000-0002-1928-992X</contrib-id>
          <name>
            <surname>Sabour</surname>
            <given-names>Siamak</given-names>
          </name>
          <xref ref-type="aff" rid="aff4">
            <sup>1</sup>
          </xref>
        </contrib>
        <contrib contrib-type="author">
          <name>
            <surname>Mansournia</surname>
            <given-names>Mohammad Ali</given-names>
          </name>
          <xref ref-type="aff" rid="aff5">
            <sup>2</sup>
          </xref>
        </contrib>
        <contrib contrib-type="author">
          <name>
            <surname>Mehrabi</surname>
            <given-names>Yadollah</given-names>
          </name>
          <xref ref-type="aff" rid="aff4">
            <sup>1</sup>
          </xref>
        </contrib>
        <contrib contrib-type="author">
          <name>
            <surname>Nazari</surname>
            <given-names>Seyed Saeed Hashemi</given-names>
          </name>
          <xref ref-type="aff" rid="aff6">
            <sup>3</sup>
          </xref>
          <xref ref-type="corresp" rid="c2"/>
        </contrib>
        <aff id="aff4">
          <label>1</label>
          <addr-line>
            <named-content content-type="city">Tehran</named-content>
          </addr-line>
          <country country="IR">Iran</country>
          <institution content-type="original">Department of Epidemiology, School of Public Health and safety, Shahid Beheshti University of Medical Sciences, Tehran - Iran.</institution>
        </aff>
        <aff id="aff5">
          <label>2</label>
          <addr-line>
            <named-content content-type="city">Tehran</named-content>
          </addr-line>
          <country country="IR">Iran</country>
          <institution content-type="original">Department of Epidemiology and Biostatistics, School of Public Health, Tehran University of Medical Sciences, Tehran - Iran.</institution>
        </aff>
        <aff id="aff6">
          <label>3</label>
          <addr-line>
            <named-content content-type="city">Tehran</named-content>
          </addr-line>
          <country country="IR">Iran</country>
          <institution content-type="original">Prevention of Cardiovascular Disease Research Center, School of Public Health and Safety, Shahid Beheshti University of Medical Sciences, Tehran - Iran.</institution>
        </aff>
      </contrib-group>
      <author-notes>
        <corresp id="c2"><bold>Mailing Address: Seyed Saeed Hashemi Nazari</bold> &#8226; Associate Professor, Prevention of Cardiovascular Disease Research Center, School of Public Health and Safety, Shahid Beheshti University of Medical Sciences, Tehran - Iran. E-mail: <email>saeedh_1999@yahoo.com</email></corresp>
        <fn fn-type="conflict">
          <p>
            <bold>Potential Conflict of Interest</bold>
          </p>
          <p>The authors report no conflict of interest concerning the materials and methods used in this study or the findings specified in this paper.</p>
        </fn>
        <fn fn-type="con">
          <p>
            <bold>Author Contributions</bold>
          </p>
          <p>Conception and design of the research, Analysis and interpretation of the data and Statistical analysis: Saadati HM, Sabour S, Mansournia MA, Mehrabi Y, Nazari SSH; Data acquisition and Writing of the manuscript: Saadati HM, Nazari SSH; Critical revision of the manuscript for intellectual content: Sabour S, Mansournia MA, Mehrabi Y, Nazari SSH.</p>
        </fn>
      </author-notes>
      <abstract>
        <sec>
          <title>Background:</title>
          <p>Body mass index (BMI) is the most commonly used index to categorize a person as obese or non-obese, which is subject to important limitations.</p>
        </sec>
        <sec>
          <title>Objective:</title>
          <p>To evaluate the direct effect of BMI on cardiovascular outcomes among participants without central obesity.</p>
        </sec>
        <sec>
          <title>Methods:</title>
          <p>This analysis included 14,983 males and females aged 45-75 years from the Atherosclerosis Risk in Communities Study (ARIC). BMI was measured as general obesity, and waist circumference (WC), waist-to-hip ratio (WHR) and hip circumference as central obesity. Targeted maximum likelihood estimation (TMLE) was used to estimate the total effects (TEs) and the controlled direct effects (CDEs). The proportion of TE that would be eliminated if all participants were non-obese regarding central obesity was computed using the proportion eliminated (PE) index. P &lt;0.05 was considered statistically significant. Analyses were performed in the TMLE R package.</p>
        </sec>
        <sec>
          <title>Results:</title>
          <p>The risk of cardiovascular outcomes attributed to BMI was significantly reversed by eliminating WHR obesity (p&lt;0.001). The proportion eliminated of BMI effects was more tangible for non-obese participants regarding WC (PE=127%; 95%CI (126,128)) and WHR (PE=97%; 95%CI (96,98)) for coronary heart disease (CHD), and WHR (PE=92%; 95%CI (91,94)) for stroke, respectively. With respect to sex, the proportion eliminated of BMI effects was more tangible for non-obese participants regarding WHR (PE=428%; 95%CI (408,439)) for CHD in males, and WC (PE=99%; 95%CI (89,111)) for stroke in females, respectively.</p>
        </sec>
        <sec>
          <title>Conclusion:</title>
          <p>These results indicate different potential effects of eliminating central obesity on the association between BMI and cardiovascular outcomes for males and females. (Arq Bras Cardiol. 2021; 116(5):879-886)</p>
        </sec>
      </abstract>
      <kwd-group xml:lang="en">
        <title>Keywords:</title>
        <kwd>Body mass index</kwd>
        <kwd>central obesity</kwd>
        <kwd>cardiovascular</kwd>
        <kwd>controlled direct effect</kwd>
        <kwd>proportion eliminated</kwd>
      </kwd-group>
    </front-stub>
    <body>
      <sec sec-type="intro">
        <title>Introduction</title>
        <p>Obesity, as a predictor of cardiovascular disease, has several definitions and criteria. Body mass index (BMI) is the most commonly used index to categorize a person as obese or non-obese.<xref ref-type="bibr" rid="B1"><sup>1</sup></xref> However, this index is subject to important limitations,<xref ref-type="bibr" rid="B1"><sup>1</sup></xref><sup>,</sup><xref ref-type="bibr" rid="B2"><sup>2</sup></xref> as it gives no information regarding fat distribution, and also cannot discriminate between different body masses (muscles, bones and fat). These limitations may lead to a misclassification of obesity levels.<xref ref-type="bibr" rid="B3"><sup>3</sup></xref><sup>,</sup><xref ref-type="bibr" rid="B4"><sup>4</sup></xref> On the other hand, central obesity indices, such as waist circumference (WC) and waist-to-hip ratio (WHR), as simple and alternative measures of obesity, directly measure the central fat mass that gives important information on health outcomes.<xref ref-type="bibr" rid="B5"><sup>5</sup></xref> In a cohort study, it was found that WC may not always be aligned with BMI and it was proposed that a combination of BMI and WC could provide a better estimate of obesity-related diseases.<xref ref-type="bibr" rid="B6"><sup>6</sup></xref> In addition, BMI is a general obesity index and provides contradictory evidence among adults and people aged 65 years and older. This phenomenon is known as &#8220;the obesity paradox&#8221;.<xref ref-type="bibr" rid="B7"><sup>7</sup></xref><sup>,</sup><xref ref-type="bibr" rid="B8"><sup>8</sup></xref></p>
        <p>In order to reveal a causal relationship, we need to maximally control the potential confounders and causal assumptions. In this regard, two causal methods&#8212;inverse probability weighting (IPW) and G-formula&#8212;have been introduced. They are based on exposure and outcome models, respectively. Regarding this issue, if the fitted model is misspecified, the results would then be biased. Double-robust methods have the advantage of simultaneously using both exposure and outcome models, and, if only one of them is misspecified, the result is still valid.<xref ref-type="bibr" rid="B9"><sup>9</sup></xref><sup>,</sup><xref ref-type="bibr" rid="B10"><sup>10</sup></xref> Considering the limitations of BMI and the constraints of observational studies, we used the targeted maximum likelihood estimation (TMLE) as a double-robust estimator to reduce the bias of the target parameters if either exposure or outcome mechanisms are estimated consistently,<xref ref-type="bibr" rid="B10"><sup>10</sup></xref> aiming to estimate the total effects (TEs) and the controlled direct effects (CDEs) of BMI. Therefore, this study aimed to determine the TEs and CDEs of BMI on cardiovascular outcomes to demonstrate how important the total effect of BMI on cardiovascular outcomes is and how much of this effect would be eliminated if all participants were non-obese with regard to central obesity (CDE).</p>
      </sec>
      <sec sec-type="methods">
        <title>Method</title>
        <sec>
          <title>Participants</title>
          <p>The Atherosclerosis Risk in Communities (ARIC) study is a prospective cohort study which began in 1987 in four counties in the USA (Washington County, Maryland; Jackson, Mississippi; Forsyth County, North Carolina; and the suburbs of Minneapolis, Minnesota). Investigators recruited 15,792 participants aged 45-64 years. More details are described elsewhere.<xref ref-type="bibr" rid="B11"><sup>11</sup></xref> We analyzed all data of visit one (1987-1989) and outcome occurrence until 2014. For the present study, participants with missing information or history of any previous cardiovascular disease were excluded. The institutional review boards in each site approved the ARIC study protocol and an informed consent form was obtained from all participants in each study visit.</p>
        </sec>
        <sec>
          <title>Measurements</title>
          <sec>
            <title>Exposure: Obesity with Body Mass Index Definition</title>
            <p>In this study, the main exposure of interest is obesity with BMI definition. BMI was calculated as weight in kilograms divided by the square of height in meters. General obesity was defined as BMI &#8805;30 kg/m<sup>2</sup>.</p>
          </sec>
          <sec>
            <title>Mediators: Central Obesity Indices Defined by Waist Circumference, Waist-To-Hip Ratio and Hip Circumference</title>
            <p>For evaluating the controlled direct effect of BMI mediated by central obesity (central fat mass), we considered three definitions for central obesity, including WC, WHR and hip circumference. WC was categorized at the cut-off point of &#8805;102 cm in men and at the cut-off point of &#8805;88 cm in women. The WHR cut-off value was set at &#8805;0.9 in men and &#8805;0.85 in women, according to the World Health Organization (WHO).<xref ref-type="bibr" rid="B12"><sup>12</sup></xref> Since there is no universal agreement regarding the hip circumference cut-off value, it was evaluated based on the best threshold value in a receiver operating characteristic (ROC) curve.</p>
          </sec>
        </sec>
        <sec>
          <title>Coronary Heart Disease, Heart Failure, Stroke and All-Cause Mortality as Outcomes</title>
          <p>The outcomes of this study included coronary heart disease (CHD) and heart failure (HF) events registered by 31 December 2014. According to the ARIC study criteria, CHD outcomes are defined as definite or probable myocardial infarction or fatal CHD. The HF outcomes are defined based on ICD-9 and ICD-10 criteria. Incident HF was defined as hospitalization that included code for HF beginning with &#8220;428&#8221; (i.e., 428.0 to 428.9) in any position, or a death certificate ICD-9 code beginning with &#8220;428&#8221; or ICD-10 code &#8220;I50&#8221; (HF or I50.0 to I50.9) in any position. Stroke events were identified by annual follow-up, hospital ICD-9 codes 430 to 436 (listed as a discharge diagnostic code at any position), or in death certificates. Cause-specific mortality was classified based on death certificates: cardiovascular mortality (ICD-9 codes 390&#8211;459, ICD-10 codes I00&#8211;I99), cancer mortality (ICD-9 codes 140&#8211;239, ICD-10 codes C00&#8211;D49), and all other causes of death.</p>
        </sec>
        <sec>
          <title>Covariates and Confounders</title>
          <p>Covariate data were derived as confounders in exposure-outcome, exposure-mediator and mediator-outcome association. Age, gender (male and female), race (black and white), education level (basic, intermediate, advanced), center (Washington County, Forsyth County, the city of Jackson, selected northwestern suburbs of Minneapolis), cigarette smoking status (defined as current, former and never smoker), drinking status (defined as current, former and never drinking alcoholic beverages), and total physical activity score (in three dimensions: at work, during leisure time and sports) were based on self-reported questionnaires. Other covariates included total calorie intake (kcal), hypertension (systolic blood pressure &#8805;140 or diastolic blood pressure &#8805;90 mmHg, or use of any medication for high blood pressure), diabetes mellitus (blood glucose &#8805;200 and fasting blood glucose &#8805;126 mg/dl, or taking any medication for diabetes), plasma lipids (mg/dl) and history of stroke at the baseline. The plasma lipids included cholesterol, high-density lipoprotein cholesterol, and triglycerides. The biologic covariates were excluded from TE analyses and were included as potential confounders of the mediator-outcome association in CDE analyses.</p>
        </sec>
        <sec>
          <title>Statistical Analysis</title>
          <p>Descriptive statistics were used to describe the participants (mean &#177; standard deviation (SD) for continuous variables, and number and percentage for categorical variables). An independent t-test analysis was used to examine the statistical differences in continuous covariates between two levels of exposure of interest (BMI). In addition, the <italic>&#967;</italic>2 test was used to examine the associations of categorical variables with exposure. The normality of data was evaluated by the normal curve (skewness and standard deviation of skewness) and the Kolmogorov-Smirnov test. To calculate the controlled direct effect of BMI on the outcomes (CHD, HF, stroke and all-cause mortality) mediated through central obesity, the TMLE model was used. TMLE, as a double-robust estimator, uses both outcome and exposure models. The implementation of TMLE has the following steps: in the first step, we generated estimators for outcome model on exposure and all listed confounders. Then, we generated estimators for treatment model (both exposure and missing outcome) on all listed confounders. In the third step, we calculated the clever covariate, H, based on treatment model (both exposure and missing outcome) for both exposed &#8220;H = 1/PS&#8221; (PS as propensity score, the probability of exposure) and unexposed &#8220;H = -1/(1 - PS)&#8221; groups.<xref ref-type="bibr" rid="B10"><sup>10</sup></xref><sup>,</sup><xref ref-type="bibr" rid="B13"><sup>13</sup></xref></p>
          <p>The missing mechanism is defined as the occurrence of a competing event (total mortality of all other causes, stroke, CHD and HF, for each interested outcome) or loss to follow-up before occurrence of the outcome of interest, where &#8220;missing = 1&#8221; indicates the outcome is observed, and &#8220;missing = 0&#8221; indicates the outcome is missing. We used a dichotomous definition of exposure (BMI); values above the defined cut-off point were classified as &#8220;obese&#8221; and those below the cut-off as &#8220;non-obese&#8221;. For the mediator variable, the three central obesity indices were used. In this way, we fixed the mediator values to zero (non-obese according to central obesity), according to the counterfactual causal model, and evaluated the controlled direct effect. TE, in the causal inference approach, is often defined as the difference between the outcome of interest of an individual or a group if exposed to a specific exposure, and the outcome of the same individual or group if unexposed. The CDE is often defined as the difference between the outcome of interest of an individual or a group if exposed to a specific exposure, and the outcome of the same individual or group if unexposed while fixing the value of the mediators. In our study, CDE of BMI was defined as the effect of BMI after controlling for WC, WHR and hip circumference indices.<xref ref-type="bibr" rid="B14"><sup>14</sup></xref><sup>,</sup><xref ref-type="bibr" rid="B15"><sup>15</sup></xref></p>
          <p>To control the confounders and possible interactions, we used a super learner machine-learning algorithm, which models different combinations of confounders and interactors in different models, and the final estimates are the weighted average of different model estimates.</p>
          <p>We fitted the algorithms (generalized linear model, stepwise GLM, and interaction GLM) for each of the exposure and outcome models, inserting all listed covariates as predictors and BMI as binary exposure.</p>
          <p>We then calculated the additive treatment effect (ATE) as risk difference for TEs and CDEs, and the corresponding confidence intervals. Influence-curve-based variance estimation was used to estimate the confidence intervals. Internal validation was performed in the super learner model as cross-validation. The proportion eliminated was calculated according to the following formula:<xref ref-type="bibr" rid="B16"><sup>16</sup></xref></p>
          <disp-formula id="ueq2">
            <mml:math display="block" id="m2">
              <mml:mrow>
                <mml:mi>P</mml:mi>
                <mml:mi>E</mml:mi>
                <mml:mrow>
                  <mml:mo>(</mml:mo>
                  <mml:mi>m</mml:mi>
                  <mml:mo>)</mml:mo>
                </mml:mrow>
                <mml:mo>=</mml:mo>
                <mml:mfrac>
                  <mml:mrow>
                    <mml:mi>T</mml:mi>
                    <mml:mi>E</mml:mi>
                    <mml:mo>&#8722;</mml:mo>
                    <mml:mi>C</mml:mi>
                    <mml:mi>D</mml:mi>
                    <mml:mi>E</mml:mi>
                    <mml:mrow>
                      <mml:mo>(</mml:mo>
                      <mml:mi>m</mml:mi>
                      <mml:mo>)</mml:mo>
                    </mml:mrow>
                  </mml:mrow>
                  <mml:mrow>
                    <mml:mi>T</mml:mi>
                    <mml:mi>E</mml:mi>
                  </mml:mrow>
                </mml:mfrac>
              </mml:mrow>
            </mml:math>
          </disp-formula>
          <p>Where PE is proportion eliminated, TE is total effect, CDE is controlled direct effect and m is fixing the mediator level to zero (non-obese). Confidence intervals (95%) for PE were assessed using the bootstrap method. The value of P &lt; 0.05 was considered statistically significant. The analysis was performed in the TMLE R package version 3.5.3.</p>
        </sec>
      </sec>
      <sec sec-type="results">
        <title>Results</title>
        <sec>
          <title>Participants&#8217; Characteristics</title>
          <p>Out of the 14,983 participants at the baseline, we included 12,085, 12,085, 12,725 and 12,936 participants in this analysis after excluding all subjects with a history of any cardiovascular disorder and missing data at the baseline for CHD, HF, stroke and all-cause mortality, respectively. For all-cause mortality, we included all participants with a history of any cardiovascular disorder. During a median 27 years of follow-up, 1,616 (13.37%), 2,229 (18.44%), 1,078 (8.47%) and 5,364 (41.47%) participants experienced CHD, HF, stroke and all-cause mortality, respectively. Within this timeframe, 3,416 (22.8%) and 1,035 (6.91%) participants experienced administrative loss to follow-up and competing risk, respectively. Regarding the participants with obesity based on BMI, during a median 27 years of follow-up, 500 (16.43%), 848 (27.86%), 357 (10.67%) and 1,676 (49.08%) participants experienced CHD, HF, stroke and all-cause mortality, respectively. Baseline characteristics (mean and standard deviation for continuous variables, and number and percentage for categorical variables for participants with and without obesity based on body mass index) are provided in <xref ref-type="table" rid="t4">Table 1</xref> and supplementary <xref ref-type="table" rid="t4">Tables 1</xref>-4. Obese participants by BMI definition were more likely to be females, black-skinned, to have lower educational level, lower annual family income, and were less likely to have health insurance compared to non-obese individuals. Regarding the mediator variables, obese participants were more likely to be obese based on WC, hip circumference and WHR indices, respectively.</p>
          <table-wrap id="t4">
            <label>Table 1</label>
            <caption>
              <title>Baseline characteristics of participants in the ARIC Study by BMI, 1987-2014</title>
            </caption>
            <table frame="hsides" rules="rows">
              <colgroup width="20%">
                <col/>
                <col/>
                <col/>
                <col/>
                <col/>
              </colgroup>
              <thead style="border-top: thin solid; border-bottom: thin solid; border-color: #000000">
                <tr>
                  <th valign="top" align="center" rowspan="2">Characteristics</th>
                  <th valign="top" align="center" rowspan="2"/>
                  <th valign="top" align="center" colspan="2">Body mass index</th>
                  <th valign="top" align="center" rowspan="3">p-value*</th>
                </tr>
                <tr>
                  <th valign="top" align="center">Obese</th>
                  <th valign="top" align="center">Non-obese</th>
                </tr>
                <tr>
                  <th valign="top" align="left">Categorical confounders</th>
                  <th valign="top" align="center"/>
                  <th valign="top" align="center" colspan="2">No. %</th>
                </tr>
              </thead>
              <tbody style="border-bottom: thin solid; border-color: #000000">
                <tr>
                  <td valign="top" align="left">Sex</td>
                  <td valign="top" align="center">female</td>
                  <td valign="top" align="center">2,484 (60.85)</td>
                  <td valign="top" align="center">5,686 (52.16)</td>
                  <td valign="top" align="center">&lt;0.001</td>
                </tr>
                <tr>
                  <td valign="top" align="center"/>
                  <td valign="top" align="center">male</td>
                  <td valign="top" align="center">1,598 (39.15)</td>
                  <td valign="top" align="center">5,215 (47.84)</td>
                  <td valign="top" align="center"/>
                </tr>
                <tr>
                  <td valign="top" align="left">Race</td>
                  <td valign="top" align="center">white</td>
                  <td valign="top" align="center">2,514 (61.59)</td>
                  <td valign="top" align="center">8,613 (79.01)</td>
                  <td valign="top" align="center">&lt;0.001</td>
                </tr>
                <tr>
                  <td valign="top" align="center"/>
                  <td valign="top" align="center">black</td>
                  <td valign="top" align="center">1,568 (38.41)</td>
                  <td valign="top" align="center">2,288 (20.99)</td>
                  <td valign="top" align="center"/>
                </tr>
                <tr>
                  <td valign="top" align="left">Education</td>
                  <td valign="top" align="center">Basic</td>
                  <td valign="top" align="center">1,237 (30.39)</td>
                  <td valign="top" align="center">2,300 (21.12)</td>
                  <td valign="top" align="center">&lt;0.001</td>
                </tr>
                <tr>
                  <td valign="top" align="center"/>
                  <td valign="top" align="center">Intermediate</td>
                  <td valign="top" align="center">1,633 (40.11)</td>
                  <td valign="top" align="center">4,494 (41.27)</td>
                  <td valign="top" align="center"/>
                </tr>
                <tr>
                  <td valign="top" align="center"/>
                  <td valign="top" align="center">Advanced</td>
                  <td valign="top" align="center">1,201 (29.50)</td>
                  <td valign="top" align="center">4,094 (37.60)</td>
                  <td valign="top" align="center"/>
                </tr>
                <tr>
                  <td valign="top" align="left">Family Income (per year)</td>
                  <td valign="top" align="center">Less than $16,000</td>
                  <td valign="top" align="center">1,208 (31.71)</td>
                  <td valign="top" align="center">2,011 (19.48)</td>
                  <td valign="top" align="center">&lt;0.001</td>
                </tr>
                <tr>
                  <td valign="top" align="center"/>
                  <td valign="top" align="center">$16,000 &#8211;$50,000</td>
                  <td valign="top" align="center">1,948 (51.13)</td>
                  <td valign="top" align="center">5,437 (52.66)</td>
                  <td valign="top" align="center"/>
                </tr>
                <tr>
                  <td valign="top" align="center"/>
                  <td valign="top" align="center">More than $50,000</td>
                  <td valign="top" align="center">654 (17.17)</td>
                  <td valign="top" align="center">2,876 (27.86)</td>
                  <td valign="top" align="center"/>
                </tr>
                <tr>
                  <td valign="top" align="left">Drinking</td>
                  <td valign="top" align="center">Current drinker</td>
                  <td valign="top" align="center">1,805 (44.61)</td>
                  <td valign="top" align="center">6,591 (60.64)</td>
                  <td valign="top" align="center">&lt;0.001</td>
                </tr>
                <tr>
                  <td valign="top" align="center"/>
                  <td valign="top" align="center">Former drinker</td>
                  <td valign="top" align="center">912 (22.54)</td>
                  <td valign="top" align="center">1,931 (17.77)</td>
                  <td valign="top" align="center"/>
                </tr>
                <tr>
                  <td valign="top" align="center"/>
                  <td valign="top" align="center">Never been a drinker</td>
                  <td valign="top" align="center">1,329 (32.85)</td>
                  <td valign="top" align="center">2,347 (21.59)</td>
                  <td valign="top" align="center"/>
                </tr>
                <tr>
                  <td valign="top" align="left">Smoking</td>
                  <td valign="top" align="center">Current smoker</td>
                  <td valign="top" align="center">793 (19.44)</td>
                  <td valign="top" align="center">3,158 (28.99)</td>
                  <td valign="top" align="center">&lt;0.001</td>
                </tr>
                <tr>
                  <td valign="top" align="center"/>
                  <td valign="top" align="center">Former smoker</td>
                  <td valign="top" align="center">1,369 (33.56)</td>
                  <td valign="top" align="center">3,487 (32.01)</td>
                  <td valign="top" align="center"/>
                </tr>
                <tr>
                  <td valign="top" align="center"/>
                  <td valign="top" align="center">Never been a smoker</td>
                  <td valign="top" align="center">1,917 (47.00)</td>
                  <td valign="top" align="center">4,248 (39.00)</td>
                  <td valign="top" align="center"/>
                </tr>
                <tr>
                  <td valign="top" align="left">Health insurance</td>
                  <td valign="top" align="center">No</td>
                  <td valign="top" align="center">562 (13.82)</td>
                  <td valign="top" align="center">901 (8.27)</td>
                  <td valign="top" align="center">&lt;0.001</td>
                </tr>
                <tr>
                  <td valign="top" align="center"/>
                  <td valign="top" align="center">Yes</td>
                  <td valign="top" align="center">3,506 (86.18)</td>
                  <td valign="top" align="center">9,991(91.73)</td>
                  <td valign="top" align="center"/>
                </tr>
                <tr>
                  <td valign="top" align="left">Family history of CVD</td>
                  <td valign="top" align="center">No</td>
                  <td valign="top" align="center">1,719 (42.70)</td>
                  <td valign="top" align="center">4,556 (42.42)</td>
                  <td valign="top" align="center">0.76</td>
                </tr>
                <tr>
                  <td valign="top" align="center"/>
                  <td valign="top" align="center">Yes</td>
                  <td valign="top" align="center">2,307 (57.30)</td>
                  <td valign="top" align="center">6,183 (57.58)</td>
                  <td valign="top" align="center"/>
                </tr>
                <tr>
                  <td valign="top" align="left">Hypertension</td>
                  <td valign="top" align="center">No</td>
                  <td valign="top" align="center">2,229 (54.97)</td>
                  <td valign="top" align="center">8,258 (76.16)</td>
                  <td valign="top" align="center">&lt;0.001</td>
                </tr>
                <tr>
                  <td valign="top" align="center"/>
                  <td valign="top" align="center">Yes</td>
                  <td valign="top" align="center">1,826 (45.03)</td>
                  <td valign="top" align="center">2,585 (23.84)</td>
                  <td valign="top" align="center"/>
                </tr>
                <tr>
                  <td valign="top" align="left">Antihypertensive medicine</td>
                  <td valign="top" align="center">No</td>
                  <td valign="top" align="center">2,208 (54.12)</td>
                  <td valign="top" align="center">8,156 (74.85)</td>
                  <td valign="top" align="center">&lt;0.001</td>
                </tr>
                <tr>
                  <td valign="top" align="center"/>
                  <td valign="top" align="center">Yes</td>
                  <td valign="top" align="center">1,872 (45.88)</td>
                  <td valign="top" align="center">2,740 (25.15)</td>
                  <td valign="top" align="center"/>
                </tr>
                <tr>
                  <td valign="top" align="left">Diabetes mellitus</td>
                  <td valign="top" align="center">No</td>
                  <td valign="top" align="center">3,234 (80.31)</td>
                  <td valign="top" align="center">10,114 (93.38)</td>
                  <td valign="top" align="center">&lt;0.001</td>
                </tr>
                <tr>
                  <td valign="top" align="center"/>
                  <td valign="top" align="center">Yes</td>
                  <td valign="top" align="center">793 (19.69)</td>
                  <td valign="top" align="center">717 (6.62)</td>
                  <td valign="top" align="center"/>
                </tr>
                <tr>
                  <td valign="top" align="left">
                    <bold>Continual confounders</bold>
                  </td>
                  <td valign="top" align="center"/>
                  <td valign="top" align="center" colspan="2">
                    <bold>Mean (SD)</bold>
                  </td>
                  <td valign="top" align="center"/>
                </tr>
                <tr>
                  <td valign="top" align="left">Age, years</td>
                  <td valign="top" align="center"/>
                  <td valign="top" align="center">54.09 (5.70)</td>
                  <td valign="top" align="center">54.30 (5.78)</td>
                  <td valign="top" align="center">0.04</td>
                </tr>
                <tr>
                  <td valign="top" align="left">Physical activity (work)</td>
                  <td valign="top" align="center"/>
                  <td valign="top" align="center">2.18 (0.99)</td>
                  <td valign="top" align="center">2.17 (0.93)</td>
                  <td valign="top" align="center">0.67</td>
                </tr>
                <tr>
                  <td valign="top" align="left">Physical activity (sport)</td>
                  <td valign="top" align="center"/>
                  <td valign="top" align="center">2.27 (0.72)</td>
                  <td valign="top" align="center">2.49 (0.81)</td>
                  <td valign="top" align="center">&lt;0.001</td>
                </tr>
                <tr>
                  <td valign="top" align="left">Physical activity (leisure time)</td>
                  <td valign="top" align="center"/>
                  <td valign="top" align="center">2.26 (0.57)</td>
                  <td valign="top" align="center">2.39 (0.57)</td>
                  <td valign="top" align="center">&lt;0.001</td>
                </tr>
                <tr>
                  <td valign="top" align="left">Total energy intake (Kcal)</td>
                  <td valign="top" align="center"/>
                  <td valign="top" align="center">1632.4 (702.3)</td>
                  <td valign="top" align="center">1637.2 (703.1)</td>
                  <td valign="top" align="center">0.72</td>
                </tr>
                <tr>
                  <td valign="top" align="left">Saturated fatty acid (%Kcal)</td>
                  <td valign="top" align="center"/>
                  <td valign="top" align="center">12.23 (2.93)</td>
                  <td valign="top" align="center">11.93 (3.02)</td>
                  <td valign="top" align="center">&lt;0.001</td>
                </tr>
                <tr>
                  <td valign="top" align="left">Total cholesterol mg/dl</td>
                  <td valign="top" align="center"/>
                  <td valign="top" align="center">5.62 (1.12)</td>
                  <td valign="top" align="center">5.54 (1.07)</td>
                  <td valign="top" align="center">&lt;0.001</td>
                </tr>
                <tr>
                  <td valign="top" align="left">Triglycerides mg/dl</td>
                  <td valign="top" align="center"/>
                  <td valign="top" align="center">1.76 (1.28)</td>
                  <td valign="top" align="center">1.40 (0.89)</td>
                  <td valign="top" align="center">&lt;0.001</td>
                </tr>
                <tr>
                  <td valign="top" align="left">HDL cholesterol mg/dl</td>
                  <td valign="top" align="center"/>
                  <td valign="top" align="center">1.20 (0.36)</td>
                  <td valign="top" align="center">1.37 (0.46)</td>
                  <td valign="top" align="center">&lt;0.001</td>
                </tr>
                <tr>
                  <td valign="top" align="left">
                    <bold>Mediators</bold>
                  </td>
                  <td valign="top" align="center"/>
                  <td valign="top" align="center" colspan="2">
                    <bold>No. %</bold>
                  </td>
                  <td valign="top" align="center"/>
                </tr>
                <tr>
                  <td valign="top" align="left">Waist circumference</td>
                  <td valign="top" align="center">Non-obese</td>
                  <td valign="top" align="center">108 (2.65)</td>
                  <td valign="top" align="center">6,893 (63.23)</td>
                  <td valign="top" align="center">&lt;0.001</td>
                </tr>
                <tr>
                  <td valign="top" align="left"/>
                  <td valign="top" align="center">Obese</td>
                  <td valign="top" align="center">3,974 (97.35)</td>
                  <td valign="top" align="center">4,008 (36.77)</td>
                  <td valign="top" align="center"/>
                </tr>
                <tr>
                  <td valign="top" align="left">Waist-to-hip ratio</td>
                  <td valign="top" align="center">Non-obese</td>
                  <td valign="top" align="center">275 (6.74)</td>
                  <td valign="top" align="center">2,927 (26.85)</td>
                  <td valign="top" align="center">&lt;0.001</td>
                </tr>
                <tr>
                  <td valign="top" align="left"/>
                  <td valign="top" align="center">Obese</td>
                  <td valign="top" align="center">3,807 (93.26)</td>
                  <td valign="top" align="center">7,974 (73.15)</td>
                  <td valign="top" align="center"/>
                </tr>
                <tr>
                  <td valign="top" align="left">Hip circumference</td>
                  <td valign="top" align="center">Non-obese</td>
                  <td valign="top" align="center">989 (24.23)</td>
                  <td valign="top" align="center">10,246 (93.99)</td>
                  <td valign="top" align="center">&lt;0.001</td>
                </tr>
                <tr>
                  <td valign="top" align="left"/>
                  <td valign="top" align="center">Obese</td>
                  <td valign="top" align="center">3,093 (75.77)</td>
                  <td valign="top" align="center">655 (6.01)</td>
                  <td valign="top" align="center"/>
                </tr>
              </tbody>
            </table>
            <table-wrap-foot>
              <fn id="TFN4">
                <p><italic>P-value was based on the &#967;2 test and independent t-test for categorical and continues variables, respectively; ARIC: Atherosclerosis Risk in Communities Study; Mean and standard deviation of continuous variables in each group of body mass index; Number and percentage of categorical variables in each group of body mass index.</italic>.</p>
              </fn>
            </table-wrap-foot>
          </table-wrap>
        </sec>
        <sec>
          <title>Total Effects and Controlled Direct Effects</title>
          <p>The TEs and CDEs of BMI, for all outcomes of interest as additive treatment effect (risk difference) with 95% confidence intervals, estimated by TMLE for all participants and sex groups, are demonstrated in <xref ref-type="table" rid="t5">Tables 2</xref> and <xref ref-type="table" rid="t6">3</xref> and supplementary Figure 1. Regarding TEs, the results show a strong and significant association between BMI and all outcomes. The stronger results are estimated for HF, all-cause mortality, CHD and stroke, respectively. Regarding CDEs, large CDEs for HF and all-cause mortality, after controlling for all three central obesity indices separately (ATE between = 4.27 and 7.95), suggest that even if central obesities were eliminated, a large effect would remain for BMI. On the other hand, small controlled direct effects for CHD and stroke, after controlling for all three central obesity indices separately, especially for WC in CHD and hip circumference in stroke (ATE between = -2.81 and 3.06), suggest that if central obesities were eliminated, a large effect would be eliminated for BMI.</p>
          <table-wrap id="t5">
            <label>Table 2</label>
            <caption>
              <title>Estimated controlled direct effect of body mass index on CHD, HF, Stroke and all-cause mortality, by Central Obesity (non-obese), participants in the ARIC Study, 1987&#8211;2014 (complete case)</title>
            </caption>
            <table frame="hsides" rules="rows">
              <colgroup width="20%">
                <col/>
                <col/>
                <col/>
                <col/>
                <col/>
              </colgroup>
              <thead style="border-top: thin solid; border-bottom: thin solid; border-color: #000000">
                <tr>
                  <th valign="top" align="left" rowspan="2">Outcomes</th>
                  <th valign="top" align="center" rowspan="2">Mediator (Central Obesity index, non-obese)</th>
                  <th valign="top" align="center">Controlled direct effect</th>
                  <th valign="top" align="center">Proportion eliminated</th>
                  <th valign="top" align="center" rowspan="2">Total effect (ATE) (95%CI)</th>
                </tr>
                <tr>
                  <th valign="top" align="center">ATE (95% CI)</th>
                  <th valign="top" align="center">PE % (95% CI)</th>
                </tr>
              </thead>
              <tbody style="border-bottom: thin solid; border-color: #000000">
                <tr>
                  <td valign="top" align="left" rowspan="3">CHD</td>
                  <td valign="top" align="center">WC</td>
                  <td valign="top" align="center">-2.81 (-5.01, -0.61)</td>
                  <td valign="top" align="center">127 (109, 135)</td>
                  <td valign="top" align="center" rowspan="3">10.47 (7.76, 13.18)</td>
                </tr>
                <tr>
                  <td valign="top" align="center">WHR</td>
                  <td valign="top" align="center">0.62 (-2.58, 3.82)</td>
                  <td valign="top" align="center">94 (79, 99)</td>
                </tr>
                <tr>
                  <td valign="top" align="center">Hip</td>
                  <td valign="top" align="center">3.06 (-1.20, 7.33)</td>
                  <td valign="top" align="center">71 (63, 75)</td>
                </tr>
                <tr>
                  <td valign="top" align="left" rowspan="3">HF</td>
                  <td valign="top" align="center">WC</td>
                  <td valign="top" align="center">6.41 (4.09, 8.72)</td>
                  <td valign="top" align="center">60 (55,75)</td>
                  <td valign="top" align="center" rowspan="3">15.92 (13.45, 18.39)</td>
                </tr>
                <tr>
                  <td valign="top" align="center">WHR</td>
                  <td valign="top" align="center">7.95 (4.54, 11.37)</td>
                  <td valign="top" align="center">50 (36,52)</td>
                </tr>
                <tr>
                  <td valign="top" align="center">Hip</td>
                  <td valign="top" align="center">7.23 (3.33,11.13)</td>
                  <td valign="top" align="center">54 (57,64)</td>
                </tr>
                <tr>
                  <td valign="top" align="left" rowspan="3">Stroke</td>
                  <td valign="top" align="center">WC</td>
                  <td valign="top" align="center">2.11 (-0.06, 4.29)</td>
                  <td valign="top" align="center">76 (74,101)</td>
                  <td valign="top" align="center" rowspan="3">8.32 (6.01, 10.63)</td>
                </tr>
                <tr>
                  <td valign="top" align="center">WHR</td>
                  <td valign="top" align="center">0.69 (-2.34, 3.73)</td>
                  <td valign="top" align="center">92 (81,106)</td>
                </tr>
                <tr>
                  <td valign="top" align="center">Hip</td>
                  <td valign="top" align="center">0.05 (-3.82,3.92)</td>
                  <td valign="top" align="center">99 (93,108)</td>
                </tr>
                <tr>
                  <td valign="top" align="left" rowspan="3">Mortality (all-cause)</td>
                  <td valign="top" align="center">WC</td>
                  <td valign="top" align="center">4.88 (2.56, 7.20)</td>
                  <td valign="top" align="center">56 (49,73)</td>
                  <td valign="top" align="center" rowspan="3">11.05 (9.21, 12.88)</td>
                </tr>
                <tr>
                  <td valign="top" align="center">WHR</td>
                  <td valign="top" align="center">4.27 (0.36, 8.17)</td>
                  <td valign="top" align="center">61 (52,69)</td>
                </tr>
                <tr>
                  <td valign="top" align="center">Hip</td>
                  <td valign="top" align="center">5.19 (2.01,8.36)</td>
                  <td valign="top" align="center">53 (50,59)</td>
                </tr>
              </tbody>
            </table>
            <table-wrap-foot>
              <fn id="TFN5">
                <p><italic>ATE: additive treatment effect; PE: proportion eliminated; CHD: coronary heart disease; HF: heart failure; WC: waist circumference; WHR: waist-to-hip ratio; Hip: hip circumference.</italic>.</p>
              </fn>
            </table-wrap-foot>
          </table-wrap>
          <table-wrap id="t6">
            <label>Table 3</label>
            <caption>
              <title>Estimated controlled direct effect of BMI on CHD, HF, stroke and all-cause mortality, by Central Obesity (non-obese), in males and females in the ARIC Study, 1987&#8211;2014 (complete case)</title>
            </caption>
            <table frame="hsides" rules="rows">
              <colgroup width="16%">
                <col/>
                <col/>
                <col/>
                <col/>
                <col/>
                <col/>
              </colgroup>
              <thead style="border-top: thin solid; border-bottom: thin solid; border-color: #000000">
                <tr>
                  <th valign="top" align="left" rowspan="2">Sex</th>
                  <th valign="top" align="center" rowspan="2">Outcomes</th>
                  <th valign="top" align="center" rowspan="2">Mediator (Central Obesity index)</th>
                  <th valign="top" align="center">Controlled direct effect</th>
                  <th valign="top" align="center">Proportion eliminated</th>
                  <th valign="top" align="center" rowspan="2">Total effect (ATE) (95%CI)</th>
                </tr>
                <tr>
                  <th valign="top" align="center">ATE (95% CI)</th>
                  <th valign="top" align="center">PE % (95% CI)</th>
                </tr>
              </thead>
              <tbody style="border-bottom: thin solid; border-color: #000000">
                <tr>
                  <td valign="middle" align="left" rowspan="12">Males</td>
                  <td valign="top" align="center" rowspan="3">CHD</td>
                  <td valign="top" align="center">WC</td>
                  <td valign="top" align="center">-4.32 (-7.69, -0.96)</td>
                  <td valign="top" align="center">144 (124,161)</td>
                  <td valign="top" align="center" rowspan="3">9.83 (5.74, 13.92)</td>
                </tr>
                <tr>
                  <td valign="top" align="center">WHR</td>
                  <td valign="top" align="center">-32.28 (-36.07, -28.48)</td>
                  <td valign="top" align="center">428 (408,439)</td>
                </tr>
                <tr>
                  <td valign="top" align="center">Hip</td>
                  <td valign="top" align="center">2.86 (-2.54, 8.26)</td>
                  <td valign="top" align="center">71 (62,81)</td>
                </tr>
                <tr>
                  <td valign="top" align="center" rowspan="3">HF</td>
                  <td valign="top" align="center">WC</td>
                  <td valign="top" align="center">4.94 (1.31, 8.56)</td>
                  <td valign="top" align="center">69 (57,79)</td>
                  <td valign="top" align="center" rowspan="3">15.76 (12.04,19.49)</td>
                </tr>
                <tr>
                  <td valign="top" align="center">WHR</td>
                  <td valign="top" align="center">-11.20 (-14.98, -7.42)</td>
                  <td valign="top" align="center">171 (158,186)</td>
                </tr>
                <tr>
                  <td valign="top" align="center">Hip</td>
                  <td valign="top" align="center">15.06 (10.32, 19.79)</td>
                  <td valign="top" align="center">4 (0.07,11)</td>
                </tr>
                <tr>
                  <td valign="top" align="center" rowspan="3">Stroke</td>
                  <td valign="top" align="center">WC</td>
                  <td valign="top" align="center">1.14 (-1.63, 3.91)</td>
                  <td valign="top" align="center">86 (75,107)</td>
                  <td valign="top" align="center" rowspan="3">8.10 (3.84, 12.37)</td>
                </tr>
                <tr>
                  <td valign="top" align="center">WHR</td>
                  <td valign="top" align="center">-6.57 (-10.63, -2.50)</td>
                  <td valign="top" align="center">181 (174,201)</td>
                </tr>
                <tr>
                  <td valign="top" align="center">Hip</td>
                  <td valign="top" align="center">1.13 (-4.42, 6.69)</td>
                  <td valign="top" align="center">86 (76,95)</td>
                </tr>
                <tr>
                  <td valign="top" align="center" rowspan="3">Mortality (all-cause)</td>
                  <td valign="top" align="center">WC</td>
                  <td valign="top" align="center">5.57 (1.38, 9.77)</td>
                  <td valign="top" align="center">49 (37,68)</td>
                  <td valign="top" align="center" rowspan="3">10.89 (8.12, 13.65)</td>
                </tr>
                <tr>
                  <td valign="top" align="center">WHR</td>
                  <td valign="top" align="center">-23.53 (-27.02, -20.04)</td>
                  <td valign="top" align="center">316 (301,329)</td>
                </tr>
                <tr>
                  <td valign="top" align="center">Hip</td>
                  <td valign="top" align="center">6.00 (1.54, 10.46)</td>
                  <td valign="top" align="center">45 (36,57)</td>
                </tr>
                <tr>
                  <td valign="middle" align="left" rowspan="12">Females</td>
                  <td valign="top" align="center" rowspan="3">CHD</td>
                  <td valign="top" align="center">WC</td>
                  <td valign="top" align="center">5.02 (3.11, 6.93)</td>
                  <td valign="top" align="center">57 (43,69)</td>
                  <td valign="top" align="center" rowspan="3">11.78 (8.70, 14.86)</td>
                </tr>
                <tr>
                  <td valign="top" align="center">WHR</td>
                  <td valign="top" align="center">1.13 (-2.67, 4.92)</td>
                  <td valign="top" align="center">90 (79,103)</td>
                </tr>
                <tr>
                  <td valign="top" align="center">Hip</td>
                  <td valign="top" align="center">4.68 (-0.25, 9.62)</td>
                  <td valign="top" align="center">60 (52,67)</td>
                </tr>
                <tr>
                  <td valign="top" align="center" rowspan="3">HF</td>
                  <td valign="top" align="center">WC</td>
                  <td valign="top" align="center">11.57 (9.18, 13.96)</td>
                  <td valign="top" align="center">30 (15,39)</td>
                  <td valign="top" align="center" rowspan="3">16.66 (13.79,19.53)</td>
                </tr>
                <tr>
                  <td valign="top" align="center">WHR</td>
                  <td valign="top" align="center">9.36 (4.01, 14.70)</td>
                  <td valign="top" align="center">44 (31,52)</td>
                </tr>
                <tr>
                  <td valign="top" align="center">Hip</td>
                  <td valign="top" align="center">8.28 (3.66, 12.89)</td>
                  <td valign="top" align="center">50 (37,61)</td>
                </tr>
                <tr>
                  <td valign="top" align="center" rowspan="3">Stroke</td>
                  <td valign="top" align="center">WC</td>
                  <td valign="top" align="center">0.06 (-1.92, 2.05)</td>
                  <td valign="top" align="center">99 (89,111)</td>
                  <td valign="top" align="center" rowspan="3">7.70 (4.73, 10.68)</td>
                </tr>
                <tr>
                  <td valign="top" align="center">WHR</td>
                  <td valign="top" align="center">0.23 (-4.09, 4.55)</td>
                  <td valign="top" align="center">97 (87,113)</td>
                </tr>
                <tr>
                  <td valign="top" align="center">Hip</td>
                  <td valign="top" align="center">4.04 (0.90, 7.19)</td>
                  <td valign="top" align="center">47 (34,63)</td>
                </tr>
                <tr>
                  <td valign="top" align="center" rowspan="3">Mortality (all-cause)</td>
                  <td valign="top" align="center">WC</td>
                  <td valign="top" align="center">5.24 (2.91, 7.57)</td>
                  <td valign="top" align="center">53 (41,66)</td>
                  <td valign="top" align="center" rowspan="3">11.24 (8.76,13.73)</td>
                </tr>
                <tr>
                  <td valign="top" align="center">WHR</td>
                  <td valign="top" align="center">5.22 (-0.72, 11.16)</td>
                  <td valign="top" align="center">53 (42,64)</td>
                </tr>
                <tr>
                  <td valign="top" align="center">Hip</td>
                  <td valign="top" align="center">1.67 (-3.43, 6.77)</td>
                  <td valign="top" align="center">85 (77,98)</td>
                </tr>
              </tbody>
            </table>
            <table-wrap-foot>
              <fn id="TFN6">
                <p><italic>ATE: additive treatment effect; PE: proportion eliminated; CHD: coronary heart disease; HF: heart failure; WC: waist circumference; WHR: waist-to-hip ratio; Hip: hip circumference.</italic>.</p>
              </fn>
            </table-wrap-foot>
          </table-wrap>
          <p>Regarding sex, the results show strong and significant association between BMI and all outcomes for both males and females. Large controlled direct effect for HF in males and females, except for WHR index in males for central obesity indices (ATE between = 4.94 and 15.06), suggests that even if central obesities were eliminated, a large effect would remain for BMI. On the other hand, small controlled direct effect for stroke in males and females, except for the hip circumference in females (ATE between = -6.27 and 1.14) and for CHD in males (ATE between = -32.28 to 2.86) for central obesity indices, suggests that if central obesities were eliminated, a large effect would be eliminated for BMI, and in some cases the effect of BMI would be reversed (protective).</p>
        </sec>
        <sec>
          <title>Proportion Eliminated</title>
          <p>The PE index for TEs of BMI, for all outcomes of interest with 95% confidence intervals for all participants and sex groups, is listed in <xref ref-type="table" rid="t5">Tables 2</xref> and <xref ref-type="table" rid="t6">3</xref>. The total association of BMI with CHD could be completely eliminated by eliminating the role of WC, in 127%. This effect could be reduced by 94% and 71% by eliminating the role of WHR and hip circumference, respectively. Regarding stroke, the effect of BMI could be eliminated by eliminating the role of hip circumference, in 99%. With respect to HF and all-cause mortality, the role of central obesity indices in eliminating the effect of BMI was somewhat similar and between 50% and 61%. With respect to sex, the total association of BMI with CHD, HF, stroke and all-cause mortality in males could be completely eliminated by eliminating the role of WHR, in 428%, 171%, 181% and 316%, respectively. On the other hand, in females, the total association of BMI with CHD, HF, stroke and all-cause mortality could not be completely eliminated by eliminating the role of any central obesity indices (between 30% for WC index for HF and 99% for WC index for stroke).</p>
        </sec>
      </sec>
      <sec sec-type="discussion">
        <title>Discussion</title>
        <p>In this large, community-based cohort study, the TEs and CDEs of BMI related to the risk of CHD, HF, stroke and all-cause mortality in participants without central obesity were evaluated using the TMLE method. It is worth mentioning that we considered two common limitations of BMI and conventional estimators, including limited capacity of BMI to distinguish between fat mass and fat-free mass, which result in misclassification, and model misspecification, which is a common source of bias in conventional estimators.</p>
        <p>In brief, compared to TEs of BMI, the CDEs of BMI among participants without central obesity for all outcomes of interest were attenuated and close to null. Remarkably, these results are more highlighted for CHD and stroke. This finding highlights the capability of central obesity indices to predict the risk of cardiovascular disease and all-cause mortality. In addition, with regard to the three central obesity indices for all participants, the proportion eliminated of BMI effects did not have consistent results for all outcomes. The proportion eliminated of BMI effects was more tangible for WHR index in males, while results were not consistent for females. In general, for most of the outcomes, the results showed that, with reduction or elimination of central obesity based on WHR index, the effect of BMI was completely or mostly removed.</p>
        <p>Furthermore, these findings highlighted the limitations of BMI in predicting cardiovascular risk as a whole or based on sex. This disagreement of BMI in relation to cardiovascular outcomes was considered as &#8220;the obesity paradox&#8221;. Several explanations for the obesity paradox related to the association of BMI with cardiovascular disease have been reported. One of the most important explanations refers to the misclassification of obesity levels based on BMI definition<xref ref-type="bibr" rid="B4"><sup>4</sup></xref>. Considering that BMI is unable to discriminate between fat mass, muscle mass and body surface area, the effect of BMI refers to a combination of these three types of mass.<xref ref-type="bibr" rid="B17"><sup>17</sup></xref> Therefore, a higher BMI is an indicator not only of greater amount of central and visceral fat, but also of higher muscle or peripheral mass (fat or bone).</p>
        <p>In recent decades, many studies have evaluated the association of different fat distributions and cardiovascular disease, demonstrating that the distribution of body fat, especially excess central fat, independent from total body fat, is an important risk factor for these outcomes.<xref ref-type="bibr" rid="B18"><sup>18</sup></xref><sup>&#8211;</sup><xref ref-type="bibr" rid="B22"><sup>22</sup></xref> In this regard, previous studies have shown that excessive fat in males is commonly stored in visceral parts, while in females it is stored in peripheral subcutaneous parts.<xref ref-type="bibr" rid="B23"><sup>23</sup></xref><sup>,</sup><xref ref-type="bibr" rid="B24"><sup>24</sup></xref> The results of the present study confirm the previous findings and underline the importance of fat distribution for males and females separately.</p>
        <p>Regarding the statistical method in use, previous methodological and original studies confirm the superiority of the TMLE method over other approaches regularly used in observational studies to measure causality. In this regard, the inverse probability weighting (IPW) method result in unstable estimates in the presence of extreme weights and violations of positivity assumption. On the other hand, compared to the TMLE method, the G-formula method is only performed based on the outcome model and, if misspecified, it brings biased estimations. TMLE is a double-robust estimator that remains consistent if either exposure or outcome mechanisms are estimated consistently.<xref ref-type="bibr" rid="B9"><sup>9</sup></xref><sup>,</sup><xref ref-type="bibr" rid="B10"><sup>10</sup></xref><sup>,</sup><xref ref-type="bibr" rid="B25"><sup>25</sup></xref></p>
        <p>Previous papers confirm the usefulness of the controlled direct effect, especially in policy assessment.<xref ref-type="bibr" rid="B26"><sup>26</sup></xref><sup>,</sup><xref ref-type="bibr" rid="B27"><sup>27</sup></xref> However, the use of this concept needs assumptions other than TEs.<xref ref-type="bibr" rid="B16"><sup>16</sup></xref><sup>,</sup><xref ref-type="bibr" rid="B26"><sup>26</sup></xref> In controlled direct analysis, one must consider the assumption for the association between mediator and outcome as well as the association between exposure and outcome.<xref ref-type="bibr" rid="B28"><sup>28</sup></xref> In addition, the interaction between exposure and mediators is an important issue in this analysis.<xref ref-type="bibr" rid="B16"><sup>16</sup></xref><sup>,</sup><xref ref-type="bibr" rid="B28"><sup>28</sup></xref> Regarding this issue, we cannot use the difference between TE and CDE to estimate the direct and indirect effects.</p>
        <p>In summary, based on previous methodological studies, regarding the limitation of the controlled direct effect and the need for stronger assumptions, this cannot be used as a valid estimation of mediation; but if we have a non-zero controlled mediated effect, it can be suggestive of the presence of mediator effect.<xref ref-type="bibr" rid="B28"><sup>28</sup></xref></p>
        <p>Before interpreting the results and concluding anything, the strengths and limitations of this study should be addressed. The strengths of this study include the application of a double-robust method that consistently estimates the parameter under a semiparametric model when one of two (exposure and outcome) models is correctly specified, regardless of which. In addition, we consider the missing mechanism to minimize the impact of a competing risk and loss to follow-up for better estimation of true effects. However, due to the problem of small sample size of the outcome of interest and sparse-data bias, we could not evaluate these estimations in age groups. In addition, this study is limited by the fact that we did not consider the variation of time-varying confounders.</p>
      </sec>
      <sec sec-type="conclusions">
        <title>Conclusion</title>
        <p>In this study, the controlled direct effect of BMI decreased to almost null in participants without central obesity. These results highlight the importance of considering the distribution of fat masses when estimating the association between obesity and an outcome of interest, for males and females separately.</p>
      </sec>
    </body>
    <back>
      <fn-group>
        <fn fn-type="supported-by" id="fn5">
          <p>
            <bold>Sources of Funding</bold>
          </p>
          <p>There was no external funding source for this study.</p>
        </fn>
        <fn fn-type="other" id="fn6">
          <p>
            <bold>Study Association</bold>
          </p>
          <p>This article is part of the thesis of doctoral submitted by Hossein Mozafar Saadati, from Shahid Beheshti University of Medical Sciences.</p>
        </fn>
        <fn fn-type="other" id="fn7">
          <p>
            <bold>Ethics approval and consent to participate</bold>
          </p>
          <p>This study was approved by the Ethics Committee of the vicechancellor in research affair-Shahid Beheshti University of Medical Sciences under the protocol number IR.SBMU.RETECH.REC.1399.763. All the procedures in this study were in accordance with the 1975 Helsinki Declaration, updated in 2013. Informed consent was obtained from all participants included in the study.</p>
        </fn>
        <fn fn-type="other" id="fn8">
          <p>
            <bold>*Supplemental Materials</bold>
          </p>
          <p>For additional information, please click here.</p>
        </fn>
      </fn-group>
      <ack>
        <title>Acknowledgments</title>
        <p>This study is related to project No. IR.SBMU.PHNS.REC.1396.152 from Shahid Beheshti University of Medical Sciences, Tehran, Iran. The institutional review boards from each site approved the ARIC Study protocol and an informed consent form was obtained from participants in all study&#8217;s visits. The authors would like to thank the staff and participants of the ARIC Study for their important contributions. In order to access the data, we signed an RMDA (NHLBI Research Materials Distribution Agreement), which is available upon request.</p>
      </ack>
    </back>
  </sub-article>
</article>
