Associations of micronutrients and lipids with prediabetes and glycemic parameters in adolescent girls of the rural DERVAN cohort (DERVAN-9).

Background: We investigated the associations of micronutrients and lipids with prediabetes, glycemic parameters, and glycemic indices among the adolescent girls of the DERVAN (aDolescent and prEconception health peRspectiVe of Adult Non-communicable diseases) cohort study from rural India. Methods: We recruited 1,520 adolescent girls aged 16-18 years. We measured glycemic parameters (glucose, insulin and HbA1C), lipids (total cholesterol, high-density lipoprotein [HDL], low-density lipoprotein [LDL], and triglycerides), and micronutrients (vitamin B12, folate, and vitamin D). Prediabetes was defined using American Diabetes Association criteria (fasting glucose ≥100 mg/dL or HbA1C ≥5.7%). Glycemic indices (insulin resistance, insulin sensitivity, and ß cell function) were calculated using the homeostasis model. Associations of prediabetes, glycemic parameters and glycemic indices with micronutrients and lipids were analyzed by multiple logistic regressions. Results: The median age and Body Mass Index (BMI) were 16.6 years and 17.6 kg/m2, respectively. Overall, 58% of girls had a low BMI. Median vitamin B12, folate, and vitamin D concentrations were 249.0 pg/mL, 6.1 ng/mL, and 14.2 ng/mL, respectively. The deficiencies observed were 32.1% for vitamin B12, 11.8% for folate, and 33.0% for vitamin D. Median total cholesterol, LDL, HDL, and triglyceride concentrations were 148.0 mg/dL, 81.5 mg/dL, 50.8 mg/dL, and 61.5 mg/dL, respectively. Elevated total cholesterol, LDL, and triglycerides were observed in 4.8, 4.0, and 3.8%, respectively, while low HDL was observed in 12.8%. Prediabetes was observed in 39.7% of the girls. Among lipids, total cholesterol and LDL were higher in girls with prediabetes (p < 0.01 for both). In a multivariate model containing cholesterol and vitamin B12/folate/vitamin D, prediabetes was associated with high cholesterol. Prediabetes was also associated with high LDL, independent of folate and vitamin D. Poor insulin secretion was high in those with low vitamin B12. Elevated insulin resistance was associated with low HDL. The likelihood of high insulin sensitivity was reduced in those with high triglycerides. The likelihood of poor ß cell function was high in those with high LDL. Statistical interactions between micronutrients and lipids for prediabetes and glycemic outcomes were not significant. Conclusion: There was a substantial deficiency of micronutrients and an absence of dyslipidemia. Our results indicate the need for lipid and micronutrient-based interventions in adolescence to improve glycemic outcomes. Maintaining adequate storage of not only micronutrients but also lipids in adolescent girls is likely to reduce diabetes risk in adulthood.

Prediabetes in rural adolescent girls from DERVAN cohort: data from the KONKAN region of the state of Maharashtra, India (DERVAN-4).

Background: India is witnessing an epidemic of type 2 diabetes. Overweight/obesity, overnutrition, physical inactivity, and family history are well-known risk factors for diabetes. We investigated the role of undernutrition in the development of diabetes among rural adolescent girls. Methods: DERVAN cohort study was set up in the KONKAN region of the western Indian state of Maharashtra. It enrolled 1,520 adolescent girls (16-18 years old at the time of enrollment). We measured glycemic parameters (glucose, insulin, and HbA1C) and body size using anthropometry and body composition using bioimpedance. Prediabetes was diagnosed using the American Diabetic Association (ADA) criteria. We also calculated various HOMA indices for insulin resistance (HOMA-IR), ß-cell function (HOMA-ß), insulin sensitivity (HOMA-S), and compensatory ß-cell response using a homeostasis model. BMI, body fat%, and waist circumferences were treated as exposures and all the glycemic parameters and indices as outcomes. Results: The median age of the subjects was 16.6 years. The median weight, height, and BMI were 40.7 kg, 151.7 cm, and 17.5 kg/m2, respectively. Prevalence of underweight was 28.8%, and stunting was observed in 30.4%. Thinness and obesity using BMI were observed in 58.4% and 4.2%, respectively. The median body fat% was 22.5, and excess body fat (>35%) was observed in 5.7%. The prevalence of prediabetes was 39.4%. Fasting insulin concentrations, HOMA-IR, and HOMA-ß showed a positive trend across body composition quartiles (p < 0.001). HOMA-S and compensatory ß-cell response showed an inverse trend (p < 0.001). Compared with prediabetic girls in the overweight/obese group, girls most undernourished group had lower median insulin concentrations (8.1 µIU/ml vs. 17.1 µIU/ml), lower HOMA-IR (1.1 vs. 2.3), and lower HOMA-ß (75.6 vs. 129.2) but higher sensitivity (87.4 vs. 43.7) (p < 0.001) for all. Conclusion: We have reported a high prevalence of prediabetes among rural adolescent girls with a very low prevalence of obesity. Prediabetes in obesity is driven by hyperinsulinemia and overworking of the pancreas while poor ß-cell function and poor insulin secretion are major drivers in the undernourished group. The high-risk diabetes screening programs are much needed for the undernourished populations. Caution should be exercised for planning the interventions as overfeeding (or overnutrition) is likely to put the populations at risk of development of obesity and insulin resistance.

Diurnal variation in salivary progesterone in fertile Indian women.

Research question: Is there a diurnal variation in salivary progesterone levels during menstrual cycle among Indian women? Design: A longitudinal study was carried out to measure progesterone in saliva among small cross-sectional sample (n = 31) of fertile Indian women of reproductive age comprising young adults (18-25 years, n = 11), adults (26-38 years, n = 9) and middle aged (39-45 years, n = 11). Saliva samples were collected twice daily (morning and evening) across the entire menstrual cycle of 31 women. Results: Mean ages at enrolment and menarche were 30.6 years and 13.6 years respectively. Fifty-five percent of the women were married. The menstrual cycle range was 20-40 days. After controlling for age and menstrual cycle length, statistically significant diurnal variation in progesterone levels was observed across menstrual cycles with high levels in the morning. Conclusions: This is the first report on salivary progesterone in subjects with Indian ethnicity and could have clinical implications for designing point of care kits for menstrual cycle management, fertility and reproduction.