Early anthropometric indices predict short stature and overweight status in a cohort of Peruvians in early adolescence

Abstract

While childhood malnutrition is associated with increased morbidity and mortality, less well understood is how early childhood growth influences height and body composition later in life. We revisited 152 Peruvian children who participated in a birth cohort study between 1995 and 1998, and obtained anthropometric and bioimpedance measurements 11-14 years later. We used multivariable regression models to study the effects of childhood anthropometric indices on height and body composition in early adolescence. Each standard deviation decrease in length-for-age at birth was associated with a decrease in adolescent height-for-age of 0.7 SD in both boys and girls (all P < 0.001) and 9.7 greater odds of stunting (95% CI 3.3-28.6). Each SD decrease in length-for-age in the first 30 months of life was associated with a decrease in adolescent height-for-age of 0.4 in boys and 0.6 standard deviation in girls (all P < 0.001) and with 5.8 greater odds of stunting (95% CI 2.6-13.5). The effect of weight gain during early childhood on weight in early adolescence was more complex to understand. Weight-for-length at birth and rate of change in weight-for-length in early childhood were positively associated with age- and sex-adjusted body mass index and a greater risk of being overweight in early adolescence. Linear growth retardation in early childhood is a strong determinant of adolescent stature, indicating that, in developing countries, growth failure in height during early childhood persists through early adolescence. Interventions addressing linear growth retardation in childhood are likely to improve adolescent stature and related-health outcomes in adulthood.

Figure 1

Follow-up of participants from original cohort.

Figure 2

(a) Scatter plot of (a) length-for-age intercept and (b) rate of change in length-for-age during early childhood, by adolescent height-for-age in 2008; 2008-2009, Lima, Peru.

Figure 3

(a) Mean LAZ of subjects in early childhood by adolescent HAZ tertile. (b) Mean LAZ of subjects in early childhood by stunting adolescent status. Both stratified by sex.

Figure 4

Scatter plot of (a) weight-for-length intercept, (b) weight-for-length slope 0-5 months during early childhood, and (c) weight-for-length slope 6-30 months during early childhood, by adolescent body mass index for age in 2008; 2008-2009.

Figure 5

(a) Mean WLZ in infancy by adolescent BAZ tertiles. (b) Mean WLZ in during early childhood by adolescent overweight status. Both stratified by sex.

Figure 6

Mean WLZ in early childhood by (a) percent body fat tertiles, (b) Fat mass index tertiles, (c) Fat free mass index tertiles. Stratified by sex.