Adult Metabolic Syndrome and Impaired Glucose Tolerance Are Associated With Different Patterns of BMI Gain During Infancy: Data from the New Delhi birth cohort

During 1969–1972, married women living in a 12-km² area of Delhi (n = 20,755) were followed up (12,13). There were 9,169 pregnancies and 8,181 live births. Trained personnel recorded the babies’ weight and length within 72 h of birth and every 6 months until age 14–21 years. Gaps in funding interrupted measurements in 1972–1973 and 1980–1982. At recruitment, 60% of families had incomes of >50 rupees/month (national average 28 rupees/month) and 15% of parents were illiterate (national average 66%). Nevertheless, 43% of families lived in one room. Hindus were the majority religious group (84%), followed by Sikhs (12%), Christians (2%), Muslims (1%), and Jains (1%).

In 1998–2002 we retraced 2,584 (32%) of the cohort and 1,583 agreed to participate. Data on schooling, occupation, household possessions, alcohol consumption, tobacco use, physical activity, and family history were obtained by questionnaire (12,13). Weight and height were measured using standardized techniques. Waist circumference was measured using fiberglass tape, in expiration, midway between the lower lateral costal margin and the iliac crest, with the subject standing. Blood pressure was recorded using an automated device (Omron 711) with the subject seated, after 5 minutes of rest (mean of two readings). Plasma glucose concentrations were measured fasting and 120-min after a 75-g glucose load. Glucose and fasting cholesterol and triglyceride concentrations were analyzed by enzymatic methods using Randox kits on a Beckman AutoAnalyzer, and HDL cholesterol was measured using the same method after phosphotungstate precipitation. IGT and diabetes were defined using World Health Organization criteria (14). Metabolic syndrome was defined using National Cholesterol Education Program Adult Treatment Panel III criteria (15,16). Insulin resistance (homeostasis model assessment [HOMA]) was estimated (17). The study was approved by the All India Institute of Medical Sciences research ethics committee, and informed verbal consent was obtained.

Data from the whole original cohort was used to derive individual SD scores for BMI and weight at 6 months and birthdays from 1 to 21 years (12). Participants had a mean ± SD of 23 ± 5.5 observations. Interpolated values were used if measurements were made within 6 months (up to 1 year), 1 year (aged 2 years), 1.5 years (aged 3 years), and 2 years (all older ages). Back-transformation provided estimates of measurements at these ages. To measure changes in measurements in early life (e.g., from 2 to 11 years), we regressed SD scores at the end of the interval (11 years) on SD scores at the beginning (2 years) and at all preceding time points (birth, 6 months, and 1 year) and expressed the residuals as SD scores. This method produces uncorrelated variables describing change between specific ages (conditional SD scores). Quadratic terms were included when relationships between measurements at different ages were nonlinear. Associations between size in early life and adult outcomes were examined using regression. Outcomes with skewed distributions (HDL cholesterol and insulin resistance) were log-transformed.

Of the 1,526 subjects attending the clinic, glucose tolerance category was definable for 1,442 and metabolic syndrome for 1,492. Compared with the remainder of the original cohort, more participants were male (58 vs. 52%, P < 0.001), mean birth weight was heavier (2,851 vs. 2,818 g, P = 0.046), and maternal literacy was 6% higher. BMI SD scores differed by −0.10 to 0.06 (mean −0.04) between birth and 21 years and were statistically significant at 11 and 12 years.

The children were short, light, and thin according to an international reference (18), but as adults almost half were overweight or obese (Table 1).

BMI (Table 2) and weight at birth and 2 years were positively related to adult waist circumference and inversely related to 120-min glucose concentrations. Eleven-year BMI was positively related to all outcomes except HDL cholesterol and glucose, and adult BMI was positively related to all outcomes except HDL cholesterol, to which it was inversely related. After adjustment for adult BMI, the associations with earlier BMI reversed, becoming significantly inverse for many outcomes. The associations were little changed after further adjustment for adult lifestyle factors (data not shown).

Subjects with high adult waist circumference, triglycerides, blood pressure, and insulin resistance and those with metabolic syndrome had a higher mean BMI than the cohort mean at all ages from birth (Figs. 1 and 2A). The pattern for metabolic syndrome (Fig. 2A) matched that for overweight/obesity (Fig. 2B) but differed from that for IGT/diabetes (Fig. 2C), which was associated with high BMI during childhood and adolescence but a low BMI from 1 to 4 years.

Conditional regression analyses showed (Table 3) that greater BMI gain from birth to 2 years was associated with higher adult waist circumference and systolic blood pressure and lower 120-min glucose concentration. Weight gain in infancy was more strongly related to adult risk factors than BMI gain (Table 4, Fig. 2) and showed additional positive associations with triglycerides, insulin resistance, and metabolic syndrome. Greater BMI/weight gain from 2 to 11 years was associated with higher waist circumference, triglycerides, systolic blood pressure, and insulin resistance and a higher risk of IGT/diabetes and metabolic syndrome. Greater BMI/weight gain between 11 years and adulthood was associated with an increase in all risk factors (lower HDL cholesterol).

The inverse association between infant BMI gain and adult IGT/diabetes was stronger in subjects with lower birth weight (OR 0.74 [95% CI 0.58–0.95] for subjects weighing <2,850 g [median] compared with 1.05 [0.81–1.36] for subjects weighing ≥2,850 g, P_interaction = 0.01). There were no significant interactions at other ages or for other outcomes.

Mean ± SD age at adiposity rebound (lowest recorded childhood BMI) was 6.6 ± 1.7 years. Earlier rebound was associated with increased adult metabolic syndrome (P = 0.07) and IGT/diabetes (P = 0.04) and higher waist circumference (P < 0.001), systolic blood pressure (P = 0.052), triglyceride concentration (P = 0.054), and 120-min glucose concentration (P = 0.01). These associations became nonsignificant after adjustment for adult BMI.

The Delhi cohort represents an affluent, well-educated section of Indian society that has undergone considerable “transition.” As children they were thin, but as young adults almost half were overweight and 29% had metabolic syndrome. Higher levels of all risk factors except IGT/diabetes were associated with BMI or weight above the average for the cohort as a whole (Figs. 1 and 2) and more rapid BMI or weight gain than the cohort average (Tables 3 and 4) from birth onward.

Strengths of the study were that it was population-based and children were measured by trained personnel, with exceptionally frequent follow-up throughout childhood. As with other birth cohorts, there was considerable loss to follow-up and participants are likely to be unrepresentative of the original sample. However, differences in their childhood sizes were small, and in a within-sample analysis, loss to follow-up would introduce bias only if associations between early BMI/weight and later disease differed between those studied and not studied, which seems unlikely given that inclusion was based only on subjects’ availability.

Studies in high-income countries have shown increased metabolic syndrome in adults of lower birth weight (19). In Delhi, after adjustment for adult BMI, there were inverse associations with BMI at birth for metabolic syndrome and its components (Table 2), but these resulted from positive associations with childhood BMI gain, not from lower BMI at birth (Table 3). The absence of associations between metabolic syndrome and small size at birth in this population may be due to their young age, low mean birth weight, or different newborn body composition (20).

Consistent with studies of adults in high-income countries (8,9), greater infant BMI/weight gain was associated with a lower risk of diabetes, especially in lower-birth-weight infants. However, it was associated with an increased risk of metabolic syndrome and its components, which is consistent with recent studies showing higher BMI, blood pressure, and insulin concentrations in children who had greater infant weight gain (10,11,21). Understanding these apparently paradoxical findings is important. Effects may differ among populations according to body composition at birth and fat and lean mass accrual during infancy and may vary for different outcomes according to critical periods of development for different tissues. In developing countries, greater infant weight gain is beneficial for survival, growth, and neurocognitive development (22). However, it may become disadvantageous as obesity-related adult chronic diseases emerge (23). The balance of benefits and risks will become clearer as the cohort ages enough to assess cardiovascular disease and mortality. In an intervention study with relevant adult outcomes, protein-energy supplementation in infancy produced no increase in adult cardiovascular risk factors (24).

A clear message from our study, consistent with studies in high-income countries, is that rapid BMI gain in childhood and adolescence and earlier adiposity rebound are associated with adult metabolic syndrome and IGT/diabetes. This result probably reflects the known correlation between childhood and adult BMI. Thus, even in underweight children in developing countries, increasing BMI SD scores (“becoming obese relative to oneself”) is a risk factor for later disease. Reinforced by evidence that risk factors in Indian children are already high (25), our study supports efforts to prevent childhood obesity. It also suggests that interventions to control adiposity should be targeted not only to obese children, but also to “normal” weight children with rising BMI SD scores.

The original study was funded by the Indian Council of Medical Research and National Institutes of Health (U.S.); the current study was funded by the British Heart Foundation and the Medical Research Council (U.K.).

We thank the participants and field and laboratory staff. We acknowledge Dr. Shanti Ghosh and I.M. Moriyama who initiated the cohort study with Dr. Bhargava, Vinod Kapani for technical input, and Rajeshwari Verma and Bhaskar Singh for maintaining liaison with the cohort.