Nutritional Determinants of Pubertal Development of Girls in Rural Bangladesh

This doctoral dissertation investigated the usual patterns of pubertal development and nutritional determinants of timing of menarche of girls in rural South Asia. A prospective, population-based cohort study of adolescent girls was conducted in rural Gaibandha District in Bangladesh between 2015 and 2017. At an initial (baseline) home visit, 15,325 girls aged 9-15 years were recruited and assessed, of whom 14,061 girls (91.8%) were followed approximately one year later. Three aspects of pubertal development were assessed: the timing of menarche, serving as the primary sexual characteristic of focal research interest, supplemented by ages of appearance of the secondary sexual characteristics of breast development and pubic hair growth. For menarche, girls were initially asked to recall the calendar year and month of their first menstruation; at follow-up, the same question was repeated. For breast development and pubic hair growth, at both visits, a self-reported questionnaire adapted from the Pubertal Development Scale (PDS) was independently administered and locally adapted to categorize girls into maturational stages corresponding to Tanner’s Sexual Maturation Scale. Anthropometric status (height, weight and mid-upper arm circumference) was measured at baseline and follow-up visits. Three 7-day food frequencies were assessed across the interim year. With the use of the probit regression, median ages at the onset of the pubertal milestones were calculated. The risk of subsequent 1-year incidence in menarche by nutritional status, as reflected by attained body size at baseline and interim growth velocities, and interim dietary intake was computed as hazard ratios (HRs) from multivariate Cox proportional hazards models. The median (95% confidence interval, CI) age at menarche was 13.17 (13.15-13.19) years. The median (95% CI) age for stages 2 and 3-4 for breast development was 11.02 (11.00-11.04) and 12.82 (12.80-12.83) years, respectively; for pubic hair growth, the age for each respective stage category was 12.93 (12.91-12.94) and 14.29 (14.27-14.31) years. At baseline, menarcheal girls were taller and had higher body mass index (BMI) than their premenarcheal peers of same age. Girls who were premenarcheal at baseline were 13% (HR=1.13, 95% CI 1.12-1.14) and 29% (HR=1.29, 95% CI 1.26-1.33) more likely to attain menarche in the following year per attained cm in height and kg/m² in BMI, respectively. Effect sizes decreased with age. Interim velocities in linear growth (per cm/year) and BMI gain (kg/m²/year) were positively associated with reporting first menstruation by follow-up among girls who were 10 and 11-13 years of age at baseline, respectively; not at older ages. Among initially premenarcheal girls, dietary intake during the interim year was associated with having reported menarche by follow-up. Specifically, mean weekly intake frequencies above the 3rd tertile cutoff for eggs, milk and dairy products and other beverages (tea and sweetened drinks) were associated with 12% (HR=1.12, 95% CI 1.02-1.23), 13% (HR=1.13, 95% CI 1.04-1.24) and 17% (HR=1.17, 95% CI 1.08-1.27) higher likelihood of being postmenarcheal by follow-up compared to intake frequencies in the 1st third for each food group. The observed associations were, however, attenuated toward the null after accounting for socioeconomic factors. On the other hand, irrespective of socioeconomic status, mean intake frequencies in the upper third for fish and seafood and sweets were associated with 11% (HR=0.89, 95% CI 0.81-0.98) and 20% (HR=0.80, 95% CI 0.73-0.88) lower likelihood of reporting menarche in the interim year, respectively, compared to intake frequencies in the lowest third. Attained size and faster growth velocities in height and BMI were associated with earlier occurrence of menarche in rural Bangladeshi girls, most evident at younger ages. Consumption of animal source foods was associated with earlier onset of menarche and, paradoxically, fish and seafood intakes with later menarche. In contemporary rural Bangladesh, better nourished girls, as reflected by their larger body size, faster growth velocity and increased consumption of animal source foods, before accounting for socioeconomic status for the latter, attained earlier menarche than their less well-nourished peers, possibly due to their advanced biological maturity. Future studies that can illuminate biological plausibility of these associations with the extended follow-up period will help better understand adolescent pubertal growth and maturation which to base interventions of public health importance in this region of priority