This study was approved by the Research Ethics Committee of the Medical School (057/2004). All participants were informed about the purpose of the study and provided informed consent before starting it. This was a cross-sectional, population-based study with simple random sampling and stratified by age groups, which was carried out from March 2004 to November 2005 in the urban adult population (≥ 18 years) of São José do Rio Preto, São Paulo, Brazil.12 The sampling was stratified according to age groups, based on the data provided by the Brazilian Institute of Geography and Statistics (Instituto Brasileiro de Geografia e Estatística, IBGE). At the time the survey, the city had a population of 370,000 inhabitants, predominantly white (82.8%) and with a balanced distribution between men (48.4%) and women (51.6%).13 The groups were divided by age ranges: 18-39, 40-49, 50-59, 60-69 and ≥ 70 years. The parameters used to calculate the stratum sample sizes were number of inhabitants, expected prevalence of hypertension for each age group, a confidence interval of 95% and a maximum error of 3%.14

The city was divided into census sections according to IBGE. In each sector, the number of individuals was determined according to the proportionality of the population. For each region, streets, homes, an adult living for more than six months in a house, who met the inclusion criteria, was randomly chosen. After the visit to the first residence, houses located on alternating sides of the street, after skipping two households, were visited. If the selected individual did not agree to participate, the next-door neighbor was randomly chosen. Exclusion criteria included pregnancy, severe degenerative diseases, incapacitating mental disorders, severe psychiatric diseases or mental disability and bedridden patients.12

Interviewers were previously trained and monitored by a field coordinator. The participants answered a questionnaire that included their personal data, income and assets, in order to assess their socioeconomic status, formal education (number of years of schooling), personal and family medical history, lifestyle, awareness of hypertension and diabetes medication being used. After that, body mass index (BMI), pulse rate and blood pressure (BP) were assessed.

The BP measurement technique was that standardized by the VII Joint National Committee:15 1) measurements were taken with a recently calibrated aneroid sphygmomanometer (Welch Allyn / Tycos) known to be accurate; 2) the cuff was placed so that the lower edge was 3 cm above the elbow crease and the bladder was centered over the brachial artery; 3) a standard, a large, and a small bladder were available for thicker and thinner arms, respectively; 4) the arm was bare and supported, with the blood pressure cuff positioned at the heart level 6) phase I and V (disappearance) Korotkoff sounds were used to identify systolic and diastolic BP, respectively; 7) the pressure was rapidly increased to 30 mmHg above the level at which the radial pulse was extinguished; 8) a cuff deflation rate of 2 mmHg per beat was used; 9) a minimum of 1-minute intervals were recommended between readings to prevent venous congestion; 10) BP was measured in both arms to detect possible differences due to peripheral vascular disease; in this case, the higher value was taken. Arterial hypertension was defined as systolic blood pressure (SBP) ≥ 140 mmHg and/or diastolic blood pressure (DBP) ≥ 90 mmHg or with the use of antihypertensive medications. For individuals with borderline BP values, a new measurement was taken on a different day at the same hour of the last measurement.

The socioeconomic status was divided into classes A, B, C, D and E, and later into groups AB, C and DE (based on family income and assets). Groups AB had a monthly income higher than 10 minimum wages; Group C, between 3 and 5 minimum wages; and Groups DE, lower than 3 minimum wages.12^,16 Formal education was defined by the number of years of study, considering two levels: Level 1: 0-11 school years; and Level 2: > 11 school years, including university degrees.12

The short version of the International Physical Activity Questionnaire was used to assess the overall physical activity (OPA), including work-related physical activity (WPA), transport-related physical activity (TPA), domestic activities (DA) and leisure time physical activity (LTPA). The level of physical activity was classified as physically active individuals (PA), who performed more than 150 minutes of physical activity per week (including manual labor jobs, walking, running, swimming and cycling) and physically inactive individuals (PI), who performed < 150 minutes per week.

The BMI was obtained by the weight/height2 ratio (kg/m²). A calibrated portable scale was used for weight measurements. Individuals were classified as normal weight (18.5 to 24.9), overweight (25 to 29.9) and obese (≥ 30 kg/m²).17

The diagnosis of type T2DM was established based on the patient’s medical history, on hypoglycemic medication and blood glucose measurements.18

The levels of blood glucose, total cholesterol (TC), high-density lipoprotein cholesterol (HDL-c) and triglycerides (TG) were analyzed after 12 hours of fasting by the colorimetric method. An RXL analyzer and the Dade Behringer reagent were used for the analysis. The low-density lipoprotein cholesterol (LDL-c) levels were calculated using Friedewald formula: LDL-c = TC - (HDL-c + TG /5) for TG < 400mg/dL.19

Metabolic syndrome (MetS) was classified according to the following diagnostic criteria: waist circumference (men ≥ 102 cm and women ≥ 88 cm), TG ≥ 150 mg/dL, HDL-c [< 40 mg/dL (for male) and < 50 mg/dL (for female)], BP (Systolic BP ≥ 130 or Diastolic BP ≥ 85 mmHg), fasting blood glucose (≥ 100 mg/dL or T2DM) or specific treatment for these conditions.20

Statistical analysis was carried out using the Minitab programs version 12.22 and R 2.4.1. Estimated frequency of physical activity according to factors (education, age, gender and socioeconomic status), characteristics (BMI, HT, MetS), and metabolic parameters (cholesterol, HDL-c, LDL-c, TG and blood glucose) was performed by estimation and population association testing by means of the method of weighted least squares. Correction for the weight of the population was made according to the age group: 55.33% for 18-39 years, 18.45% for 40-49 years, 12.29% for 50-59 years, 8.20% for 60-69 years and 5.73% for 70 years or more. The bootstrap statistical method (simulation method of convex combinations with the same weights used for the analysis of frequencies, where 1,000 bootstrap samples are generated for each comparison) was used to assess the physical activity, associated with serum lipid and glucose levels. This method (resampling technique) attempts to accomplish what would be desirable to do in practice if it were possible: to repeat the experiment. The observations are chosen at random and the estimates recalculated for the purpose of improving the final estimate.21 The level of significance was 5%.

A sample of 1,717 urban adults (≥ 18 years) was evaluated in this study. Table 1 shows the distribution of the studied sample and the prevalence of PI individuals adjusted for the population according to age groups (in years) and the expected number of PI in the population. In the general population, the prevalence of PI individuals was 65.8% (95% CI: 62.2%-69.5%) and 34.2% (95% CI: 30.5%-37.8%) for the PA group. Regarding gender and age range, there was a higher prevalence of PI among women in both age groups 18 to 39 years and ≥ 70 years (women 73.9% and men 56.3%; p = 0.006 and women 83.1% and men 72.7%; p = 0.03, respectively). No differences were found between genders in the other age groups.

Table 2 demonstrates the levels of physical activity estimated for the population and related to demographic data (age, gender, education and socioeconomic status), as well as risk factors (HT, obesity and MetS). Table 2 also shows the prevalence ratios among the studied variables in the PI individuals.

The prevalence of PI individuals was higher among women as compared with men, with a prevalence ratio of 1.19 (p = 0.003). There was a lower prevalence of PI in class AB, in comparison with class C (p = 0.03). The prevalence rate of PI among those with lower educational levels when compared with those with higher educational level was 1.15 (95% CI: 1.02 to 1.31%) (p = 0.02). There was no difference regarding PI prevalence ratios in the groups of evaluated BMI (table 2).

The corrected prevalence of HT in the studied population was 25.4% (95% CI: 22.8%-27.9%). The PI/PA prevalence ratio was 1.28 (95% CI: 1.01-1.64) (p = 0.04) showing a lower prevalence of HT in active subjects (table 2).

The corrected prevalence of MetS for the studied population was 22.7%. The sedentary/active prevalence ratio was 1.56 (95% CI:1.10-2.23; p = 0.007) (table 2).

Biochemical data (TC, LDL-c, HDL-c, TG, glucose) and the presence of MetS were evaluated in 1,369 subjects who completed the examinations.

The bootstrap statistical method was used to assess the physical activity, associated with serum lipids and blood glucose levels. The levels of total cholesterol, LDL-c and TG were higher in the physically inactive than in physically active individuals (189.1 ± 1.8 vs 183.3 ± 3.1 mg/dL; 116.5 ± 1.4 vs 111.6 ± 2.6 mg/dL and 134.8 ± 4.2 vs 126.5 ± 0.9 mg/dL, respectively). The images demonstrate a rightward shift of the curve, with a tendency for higher plasma levels of these lipoproteins in physically inactive subjects (Figure 1, panels A, C and D for TC, LDL-c and TG, respectively). For HDL-c, the curves are practically superimposed (46.3 ± 0.7 mg/dL in the physically inactive and 46.9 ± 1.0 mg/dL in the active individuals) (Figure 1, Panel B). Concerning blood glucose, sedentary individuals showed a rightward shift of the curve, which demonstrated a clear tendency for higher levels of glucose in the physically inactive group (84.1 ± 1.2 mg/dL for PI and 79.4 ± 0.9 mg/dL for PA) (Figure 2).

Figure 1

Figure 2

In the present study, there was a 28% higher probability of HT in PI individuals. In turn, studies have shown a reduced prevalence of HT related to a higher rate of physical activity.33^,34 Physical activity decreases total peripheral resistance by improving endothelium-dependent relaxation, mainly mediated by a significant increase in vascular nitric oxide (NO) production and/or decrease in NO scavenging by reactive oxygen species (ROS). Moreover, exercise has also been shown to release several cytokines and anti-inflammatory peptides, which in turn increase NO bioavailability by decreasing ROS production. Exercise-related vasodilation was also associated with the growth of new arterioles and the reduction of nervous sympathetic activity.34

It was observed a 56% higher probability of MetS in PI individuals. Several studies have found that the practice of LTPA was associated with a reduction in MetS components.35^-37 Longer duration of daily physical activities leads to a lower prevalence of MetS, especially for the components of central obesity and HDL-c levels.37 Moreover, a recent study showed that middle-aged individuals with MetS who usually perform physical activity had lower arterial stiffness and more favorable cerebral white matter integrity than their sedentary peers.38

The analysis of bootstrap in the PI group showed a tendency for higher levels of lipoproteins (TC, LDL-c and TG), when compared with PA subjects. Physical exercise has shown beneficial effects on plasma lipoproteins, such as decrease in TC, LDL-c, and TG levels.39^,40 Spending less time in sedentary behaviors, and having medium levels of intense physical activity may be associated with a more favorable blood lipid profile.40

The bootstrap method curve showed a rightward shift in the PI group, and, consequently, lower glucose levels in the PA group. Data have shown an inverse association between physical activity and blood glucose.34^,35 Studies have demonstrated that improved blood glucose control is due to increased insulin sensitivity and glucose metabolism promoted by physical exercise.37

Some limitations of this study should be mentioned. First, the guidelines of the American Diabetes Association18 recommend the confirmation of hyperglycemia through a second blood glucose measurement, which was not performed in this study. However, epidemiological studies, including NHANES, used almost exclusively a single blood glucose measurement for the diagnosis of T2DM. Second, the interference of antilipemic drugs on lipid values determination was not assessed. Third, the association between physical activity and its benefits in preventing CVD risk and T2DM could be better evaluated in a cohort study rather than a cross-sectional one. Nevertheless, this fact does not invalidate the present study since that its main objective was to evaluate the prevalence of physical inactivity in the urban population. The benefits of physical activity could be better assessed through a cohort study, which is not the goal of this study. On the other hand, this population-based, age-stratified study, is unique as it gathers different demographic, epidemiologic and risk factors involved in the association between physical activity, hypertension and CVD in a single sample with a population assessment calculation, which might be extrapolated to other populations.