ABC | Volume 114, Nº5, May 2020

Original Article Ribeiro Junior & Fernandes Effect of Activity on Hypertension Arq Bras Cardiol. 2020; 114(5):755-761 Method Study design and population A cross-sectional study with Footwear Industry workers (n=446) and Urban Cleaning workers (n=624), totaling 1,070 subjects. In both survey clusters, the response rate was 97%. The sampling plan in the Footwear Industry was a stratified random sample proportional to the gender and number of workers from each of the two factories in the countryside of Bahia; in Urban Cleaning, a census was carried out at the municipal company in Salvador. Data collection was carried out in 2012 and 2010. As an inclusion criterion, all workers interviewed should be employed during data collection, which was conducted by a trained team of health and safety professionals and students of the physical therapy program. Everyone was aware of the need to clarify any aspects about the questions in order to obtain the most reliable answers possible. The interviews were carried out in each participating company, during regular working days, in a reserved place, ensuring the workers’ privacy. Survey The survey carried out by the interviewer included que s t i on s abou t s oc i odemog r aph i c f ac t o r s , j ob characteristics and life habits. The sociodemographic factors asked were gender, age, marital status, schooling, color or race, and if the interviewed worker had any children. Job characteristics: physical demands, evaluated through the handling of loads (lifting, pushing and pulling); psychosocial aspects of work, measured using the Job Content Questionnaire (JCQ), an instrument that measures the forms of the Demand-Control Model; 15 working hours and time of service in the company studied. Life habits: current or previous smoker and frequency of alcohol use. Direct measurements of Blood Pressure (BP), weight, height and waist circumference were performed. Dependent variable The dependent variable is HBP, defined by two measures of BP: the first at the beginning of the interview, with subjects sitting for five minutes before the start of the survey, and the second upon completion of the survey, as carried out by the interviewer, with a mean interval of 20 minutes between these measurements. A duly calibrated aneroid sphygmomanometer and stethoscope were used. The measurements were performed according to the recommendations of the Sixth Brazilian Guidelines on High Blood Pressure. Cases of hypertension were considered as those with systolic blood pressure ≥ 140 and/or diastolic blood pressure ≥ 90, and subjects undergoing regular treatment for hypertension. The information was divided into two groups, classifying subjects as those who did and those who did not show a hypertensive condition. Main independent variable The main independent variable is PA in three forms: occupational (PA-O), domestic (PA-D) and leisure (PA-L). The level of PA-O was obtained through the worker’s self-report on his physical demands of load handling (lifting, pushing and pulling), using a 0-5 scale, where 0 meant no exposure (never) and 5 meant maximum exposure (all the time). The three variables were summed, with the resulting value divided by three, to calculate the mean of the scores. Those workers who reached score ≤ 2 (less active) were considered exposed, whereas the non-exposed ones had scores ranging from 3 to 5 (more active). For the PA-L, the following question was asked, “What do you do when you’re not working at home or at work?”, with four answer options: (I) Talk to family or friends, read the newspaper or a magazine, watch TV, go to church; (II) Hike, fish, maintain the garden or the yard; (III) Run, exercise, swim, play a ball game, ride a bicycle; (IV) Train for a competition. Individuals with “I” and “II” responses were considered, therefore, to have lower PA-L levels, i.e., the exposed ones; and those who responded positively to “III” and “IV” were considered active in leisure, i.e., the non-exposed ones. In the PA-D domain, workers were questioned about weekly hours devoted to domestic work. They were classified as exposed and non-exposed, absence and presence of hours dedicated to housework, respectively. In the multivariate analysis, the main variable PA was used, subdividing the three forms into three strata. Workers who had high PA levels in the three forms, i.e., active at leisure, at work and at home, were considered non-exposed, and two strata of exposure were used: the first one included active individuals in two forms; and the second, individuals who were inactive or active in only one form. Covariables For the age variable, those workers who were older than the median age were considered to be exposed. Color or race was stratified into “non-black”, which includedWhite, Brown, Asian and Indigenous, classified as non-exposed, whereas Black individuals were classified as exposed. As for schooling, the exposed ones were those who had not completed High School. For marital status, those who had a stable partner were exposed, even if they were not formally married. The variable “type of workday” consists of rotating shifts (exposed groups) and business hours (non-exposed groups). For the psychosocial aspects of work, psychological demand and job control questions, from the JCQ, were used. Exposed individuals were those who had high demand and low control over their work. For the time of service variable, those exposed were the ones with values above the median. An individual who smokes or used to smoke was considered exposed. For “alcohol use frequency”, those who consumed alcohol more than once a week were classified as exposed. The anthropometric indicator used to verify the association with HBP in the multivariate analysis was Body Mass Index (BMI), calculated from subjects’ direct height and weight measurements. Excess weight was considered present above the cut-off point of ≥ 25 kg/m 2 , and those were the exposed ones, including, therefore, those who were overweight and obese. 16 756