Background: There is no study available concerning specifically the role of underweight in PAD prevalence. Patients and methods: Individuals ≥ 65 years living in urban and rural areas of two countries in Central Africa (Central African Republic and the Republic of Congo) were invited. Demographic, clinical and biological data were collected, and ankle-brachial index measured. BMI was calculated as weight/height2 and participants were categorized according to the World Health Organization as with underweight (< 18.5 kg/m2), normal weight (18.5-24.9 kg/m2), overweight (25-29.9 kg/m2), and obese (≥ 30 kg/m2). Results: Among the 1815 participants (age 73.0 years, 61.8 % females), the prevalence of underweight was 34.1 %, higher in subjects with PAD than in PAD free subjects (37.1 % vs. 33.5 %, p = 0.0333). The overall prevalence of PAD was 14.3 %. Underweight and obesity were still significantly associated with PAD after adjustment to all potential confounding factors (OR: 2.09, p = 0.0009 respectively OR: 1.90, p = 0.0336) while overweight was no more significantly associated with PAD after multivariate analysis. Conclusions: While obesity is a well-known PAD associated marker, low BMI provides novel independent and incremental information on African subject's susceptibility to present PAD, suggesting a "U-shaped" relationship between BMI and PAD in this population.