Methods: This was a cross-sectional study conducted in rural and urban Bengali school going girls. SES, 24 hours dietary information and anthropometries were collected. Upper arm fat (UAFA) and muscle areas (UAMS), fat free mass (FFM) and fat mass (FM) were estimated. Mann-Whitney U test for median test, Chi-squire or Fisher exact for frequency test and Spearman rank correlation were performed to find association between maternal variables and menarcheal age.

Result: A total 463 girls (232 of rural area and 231 of urban area) participated in this study. The menarcheal age of rural girls was significantly higher than urban girls (12.7 vs 12.2 years; P< 0.001). Consumption of protein, fat and energy were higher and carbohydrate was lower in urban girls compared to rural girls (both P< 0.05). Fat mass and UAFA were positively associated with age at menarche only in urban girls (P< 0.05). In urban girls, 10% increment of MUAC and fat percent caused an increase of 0.34% and 0.21% menarche age respectively (both P< 0.0001). Energy was inversely related with menarche age and 10% increment of energy intake menarcheal age decreased by 2.42% (P< 0.0001).

Conclusion: This study indicated that urban girls have had early menarcheal age compared to rural girls. Nutrition and body composition are associated with age at menarche only in urban area.