We report here the results of the first nationally representative survey on the prevalence and risk factors for alcohol consumption in Burkina Faso.
The overall prevalence of people with abusive alcohol consumption is 2.7% [2.0 – 3.7]. This consumption is lower compared to other African countries  and the rest of the world . This prevalence might be underestimated because even though Burkina Faso is a laic country, the majority of its inhabitants have a religious faith that prohibits alcohol consumption that could prevent some people from declaring their alcohol intake, this might be particularly truth in the Sahel region. In addition, alcohol consumption is blamed by society and could also lead to an under-declaration (positive social perception bias) .
We also found differences in the different regions of the country with the highest consumption found in the Sud-Ouest region. In this region almost each household produces local beer ('dolo') this might explain this observation. The income disparities and the alcohol availability between the different regions could explain those differences . From a public health perspective this region also have the highest prevalence of Hepatitis B and C . The inhabitants of these regions seem to cumulate the hepatotoxic risk of Hepatitis and alcohol consumption. We found a null prevalence in the Sahel region, this is an entirely Muslim region and the desirability bias may be more important compared to other region.
We also found gender differences with an increased risk in men compared to women for abusive alcohol consumption. It is important to specify here that, following the WHO guidelines, the thresholds used to define the different levels of alcohol consumption are different for men and women (33 to 50% lower for women compared to men).
In most countries, the prevalence is higher in men compared to women [17, 34], but when adjusting for multiple factors such as social supports and financial aspects, this effect seems less important. Compared to men, more women are lifetime abstainers, drink less, and are less likely to engage in problem drinking, develop alcohol-related disorders or alcohol withdrawal symptoms. However, women drinking excessively develop more medical problems. Biological (sex-related) factors, including differences in alcohol pharmacokinetics as well as its effect on brain function and the levels of sex hormones, may contribute to some of those differences . Since pregnant women are particularly vulnerable, we performed sub-group analysis. A subgroup of 299 out of the 2,449 (12.2%) women were pregnant during the survey, amongst them, 6 (3.9%) reported alcohol consumption, which is statistically lower compared to non-pregnant women (12.6%, p = .001).
Interestingly we observed that the age was significantly associated with alcohol consumption with a gradient effect of age on alcohol consumption but the risk of abusive consumption is not influenced by age. The influence of age on alcohol consumption is still unclear and not well documented in the literature, except for binge drinking, where young adults are the most at risk . However, this result should be interpreted carefully, considering our study design. On the other hand, the results could be the results of preventive campaigns as reflected by a cohort effect A large Australian study including 7 cross-sectional waves showed indeed that male cohorts born between the 1965 and 1974 and female cohorts born between 1955 and 1974 reported higher rates of drinking participation (P < 0.05), while the most recent cohorts (born in the 1990s) had lower rates of participation (P < 0.01) .
Concerning the risk of switching to an abusive consumption among drinkers only two factors have been identified: the risk in increased in male (as for the general consumption) and decreased with the occupational status, probably due to financial constraints.
The association with tobacco consumption is probably one of the most important from a public health view because of the comorbidities and the double burden it presents for the population. In a previous study, we identified people most at risk of tobacco consumption in Burkina Faso: tobacco smoking among men was significantly associated with increased age and alcohol consumption. Analysis of risk factors for other tobacco use stratified by gender show that age, education, residence and alcohol consumption were significantly associated with consumption for women, age and alcohol consumption for men .
As recommended by the WHO, in this paper we presented risk factor associated to abusive alcohol consumption because of the proven negative effects. It is, however, interesting to note that from a medical point of view a limited and reasonable alcohol consumption could have some health related benefits: for example the relation between dementia and cognitive disorders is not linear and limited alcohol consumption has a protective role for dementia [38–40] or the protective effect of alcohol on cardio vascular risk, previously known as 'the French paradox' [41–43]. However, these results must be interpreted with caution, especially since it is known that even in low doses, alcohol consumption transiently increases the risk of cardiovascular accidents . Another important point is that considering the prevention side experiences showed that adopting a too strong position by prohibiting any consumption or behavior will lead to poor results [45, 46]. Considering these two aspects low alcohol consumption could be considered as acceptable.
The main limitation of this study is that alcohol (and tobacco) consumptions were obtained during interviews and is therefore dependent on the faith of the participants. There is therefore, both a risk of memory bias and social desirability, probably more marked during pregnancy. It can thus be estimated that the numbers and prevalence obtained in this survey underestimate the actual consumption. Another potential limitation is that some well-known risk factors for alcohol consumption were not included in the study because data on these variables have not been collected during the STEPS survey. Part of such variables is socioeconomic status. It is a transversal study; therefore there is a risk of survival bias indicating that elder participants with high consumption may die prematurely due to this consumption .
Despite these limitations, given the study design (cluster sampling design) and the sample size, the results of this study can be extended to the whole of Burkina Faso.