Description of the Burkina Faso STEPS Survey
The WHO STEPS surveys use a standardized tool for data collection which includes specific sections on behavioural risk factors (tobacco and alcohol consumption, oral hygiene practices, FV intake, physical activity); anthropometric [body mass index (BMI)] and blood pressure measurements; and blood biochemistry [9].
The first WHO STEPS survey in Burkina Faso, conducted from 3 September to 24 October 2013, was nationally-representative and covered all the country’s 13 administrative regions. It involved interviews on behavioural or lifestyle factors as well as anthropometric and blood pressure measurements [9]. The survey enrolled adults aged 25–64 years, based on a calculated sample size, large enough to allow sub-group comparisons. The sample was weighted by sex, age group, and rural/urban residence. Face-to-face interviews were conducted in a language spoken by the participant and data captured using personal digital assistants pre-loaded with eSTEPS software.
All methods were carried out in accordance with relevant guidelines and regulations. The protocol of the STEPS survey was approved by the Ethics Committee for Health Research of the Ministry of Health of Burkina Faso (deliberation No: 2012-12092; December 05, 2012). Written informed consent was systematically obtained from each participant in the STEPS survey.
Study variables
Sociodemographic data collected included living environment, sex, age, marital status, education level and occupation. Self-reported data on the modifiable lifestyle factors were also collected: alcohol and/or tobacco use, oral hygiene practices, FV consumption, and physical inactivity. The anthropometric measurements of weight, height, as well as blood pressure were taken. Current alcohol consumption was defined as alcohol intake in the past one month while current tobacco use was defined as ever use of smoked or smokeless tobacco in the past 12 months. The oral hygiene practices were categorized based on the frequency of cleaning teeth per day, with, at least twice daily cleaning being recommended [10]. Daily FV intake was derived from the number of servings of FV consumed per day during a typical week. Five or more daily FV servings is recommended [11]. Physical activity was investigated via the amount of time being physically active in three domains; transport, at work and during leisure time and participants were asked about the frequency, intensity and duration of their work-, travel- and leisure-related physical activity (vigorous or moderate), in a typical week [12]. We considered participants who reported no vigorous- or moderate physical activity during a typical week as being physically-inactive. BMI, calculated as a subject’s weight divided by height², in kg/m², was characterized as underweight (BMI<18.5 kg/m²), normal (BMI=18.5 – 24.9 kg/m²) overweight (BMI=25 – 29.9 kg/m²) or obesity (BMI ≥30 kg/m²) states [13].
We defined persons living with hypertension as those with higher than or equal to 140 mmHg and/or diastolic blood pressure higher than or equal to 90 mmHg or those who reported current antihypertensive therapy use [1, 14]. Of those with hypertension, those who reported having been told by a doctor or a health professional as having raised blood pressure or hypertension were considered as being aware of their condition.
Sample size: Of a sample of 4800 individuals enrolled, 105 were not eligible; 10 had invalid data on sociodemographic variables. The number with missing or implausible data on lifestyle factors was as follows: 1 for tobacco, 6 oral hygiene practice; 279 for FV intake, 205 for BMI, 7 for blood pressure The 3413 normotensive individuals were excluded and finally, only the 774 hypertensive individuals with complete data were included in our secondary data analyses.
Statistical analyses
We used StataCorp Stata Statistical Software for Windows (Version 14.0, College Station, Texas, US) to analyse the data. The quantitative variables were expressed as the means ± standard deviations, and the qualitative variables expressed as percentages (%) with 95% confidence intervals (CIs). Student’s t test was used to compare quantitative variables, and the chi-square or the Fishers exact tests were used to compare categorical variables.
In the stepwise logistic regression models, we dichotomized the outcome variable (yes/no) as being aware or unaware of one’s hypertension status, while the lifestyle factors were the explanatory variables, with adjustment on sociodemographic factors (sex, age, urban-rural residence, marital status, education and occupation). The Hosmer-Lemeshow test was performed to determine the goodness-of-fit of the logistic regression models. A p-value greater than 0.05 in the Hosmer-Lemeshow chi-square test was considered significant. Excluding the Hosmer-Lemeshow test, for all analyses, a p-value below 0.05 was considered statistically significant.