The determinants for a child with group A β-haemolytic streptococcal throat infection (BHS infection) to develop rheumatic fever (RF) remain unclear. In this case-control study, we deal with BHS infected children to examine whether nutritional factors are related to RF.

In a RF hospital in Dhaka (Bangladesh) we examined 218 consecutive out-patients who had antecedent BHS infection. Sixty of them met the updated Jones criteria for RF (cases) while 104 did not (controls). Fifty-four possible RF patients were excluded. We used anthropometric measurements and a food frequency questionnaire to assess nutritional factors.

Higher risk of RF was observed for low height for age (odds ratio 3.82, 95% confidence interval 1.73–8.42); low weight for age (2.41, 1.12–5.57); low upper arm circumference for age (3.76, 1.87–7.89); and low consumption of eggs (3.81 , 1.95–7.63), milk (2.60, 1.36–5.08), chicken (2.62, 1.35–5.21), pulses (1.98, 1.03–3.84), fruits (2.29, 1.20–4.45), and ruti (home-made bread) (3.15, 1.61–6.34). Reduced risk was observed for soybean oil consumption (0.28, 0.12–0.62). The significant association of upper arm circumference and eggs persisted after adjustment for multiple sociodemographic confounders. The association of ruti and soybean oil appeared to be suggestive (0.05<p<0.1).

Protein-energy malnutrition is likely to be associated with RF. The protective effect of moderate consumption of eggs and soybean oil may support other published work which suggests that the antiinflammatory substance present in these food items may prevent maturation of the rheumatic process.