« Back

The intra-uterine growth of babies in rural India in relation to their mothers' nutrition before and during pregnancy (PMNS: Pune Maternal Nutrition Study)

Funded By:  The Wellcome Trust, UK
Month and Year of Commencement:  October 1993
Current Status:  Ongoing

Principal Investigator:  Dr.C.S.Yajnik, Director, Diabetes Unit, KEM Hospital & Research Center, Pune, India
Co-Investigator:  Prof. Caroline Fall, Professor of International Paediatric Epidemiology, Southampton General Hospital,UK

PMNS1        PMNS2
» Background
» Objectives
» Methodology and Design
» Results


The escalating epidemic of type 2 diabetes is usually ascribed to an interaction of a ‘thrifty genotype' and the modern day lifestyle of dietary abundance, physical inactivity and psychosocial stress. In this scheme prevention of diabetes involves changing adult lifestyle. Hales and Barker introduced a paradigm shift in the idea of prevention of diabetes when they suggested that the intrauterine environment is a major determinant of type 2 diabetes. Maternal health and particularly her nutrition thus became the focus of attention.

The PMNS is one of the first prospective community-based study to investigate the relationship between maternal nutrition and fetal growth and its future risk of type 2 diabetes.


To study impact of maternal nutrition before and during pregnancy on fetal growth

Methodology and Design

The study started in 1993 in 6 villages near Pune in collaboration with Environmental Epidemiology Unit of the Medical Research Council, Southampton, UK (Prof David Barker, Prof Caroline Fall and colleagues), Agharkar Research Institute, Pune (Dr Shobha Rao and colleagues), National Informatics Centre, Pune (Mrs. Punam Gupta), and other experts from India and abroad who advised us in various measurements.

In a house-to-house survey we identified over 2500 non-pregnant women who agreed to take part in the study. They were visited every month to record their menstrual dates and every 3 months we measured detailed anthropometry and a blood hemoglobin concentration. Over 800 of these women became pregnant during the study and were intensively studied for anthropometric changes, nutritional intake, physical activity, and circulating nutrient levels. Fetal growth was assessed by antenatal ultrasonography and anthropometric measurements at birth. Growth measurements are done every 6 months after birth. Every 6 years we investigate them to measure various risk factors for type 2 diabetes and cardiovascular disease: DXA body composition, HOMA insulin resistance, glucose tolerance and CVD risk markers.


PMNS has a reputation for one of the most important studies in the world, which has collected prospective high quality information to test the ideas in Developmental Origins of Health and Disease (DOHaD) and the life-course approach to prevention of chronic non-communicable diseases. The community participation is over 90% and at 6 year follow up we investigated over 96% of live children.

  • The Pune newborns were smaller and thinner than the white UK babies (weight 2.7 vs. 3.5kg). However, they were disproportionately small: abdominal circumference and lean tissue measurements were much smaller compared to the subscapular skin fold of the UK babies, thus anticipating the ‘thin-fat' phenotype of adult Indians. We have shown that such a body composition is associated with higher insulin resistance and higher risk of type-2 diabetes. Pune babies also had high cord- blood leptin and insulin concentrations and lower adiponectin concentrations compared to those in the white babies. This indicates that the metabolic-endocrine predictors of type-2 diabetes are also present at higher levels in Indian newborn babies.

  • Maternal measurements predicted ‘like' measurements in the babies: taller mothers had longer babies and mothers with larger skin folds had fatter babies. Maternal energy and protein intakes were unrelated to newborn size but mothers who ate more micronutrient-rich foods (green leafy vegetables, fruit and milk) had larger babies. Mothers with a higher physical workload had smaller babies. In general, these maternal factors predicted overall newborn size, but not the disproportionate ‘thin-fat' phenotype.

  • We observed that 70% had low vitamin B 12 concentrations (< 150 pmol/l), 33% had high tHcy (> 10µmol/l) and 90% had high MMA concentrations (> 0.26µmol/l), a marker of vitamin B 12 deficiency. Only one mother had low folate (< 283 nmol/l) concentration. Thirty five percent of the women were lacto-vegetarian, but milk intakes were generally low. The non-vegetarian women ate relatively small amount of eggs, meat and fish. Intakes of animal foods were directly related to vitamin B 12 and inversely to homocysteine concentrations. Hyperhomocysteinemia (predominantly due to vitamin B12 deficiency) was associated with intrauterine growth retardation.

  • Gestational diabetes was uncommon (1%) in these young rural mothers. However, higher maternal plasma glucose concentrations (in the ‘normal' range) predicted larger newborn skin folds, as did maternal cholesterol and triglyceride concentrations.

  • At 6 years of age these children were small and thin compared to white UK children but had a high percentage body fat. Based on the DOHaD concept, we hypothesized that ‘better' maternal nutrition (higher intakes of micronutrient-rich foods and higher folate, vitamin C and B 12 concentrations) would predict ‘better' outcomes in the children (greater lean mass, less adiposity and lower insulin resistance). The only nutritional measurement that fulfilled this expectation was vitamin B 12; higher maternal concentrations of vitamin B 12 during pregnancy were associated with lower insulin resistance in the child. Unexpectedly, higher maternal folate concentrations were associated with higher insulin resistance and larger fat mass in the child. Children of mothers with low vitamin B 12 and high folate status were the most insulin resistant. Size and body proportions at birth were not related to insulin resistance at 6 years of age.


Vitamin B 12 deficiency appears widespread in Indians, but has received little attention from researchers, clinicians or policy-makers. Deficiency is mainly attributable to low dietary intakes in vegetarians. Vegetarianism originated in the principle of ‘Ahimsa' (not killing animals for food) introduced by Samrat Ashok about 2200 years ago, and is institutionalized in 3 main religions in India: Hindu, Jain and Buddhism. Unlike vitamin B12 deficiency among Western vegans, deficiency in Indians is multigenerational and is present from the earliest stages of intrauterine development.

We speculate that repleting maternal vitamin B 12 stores could improve body composition and insulin sensitivity in the offspring, thus reducing the risk of type-2 diabetes. Another opportunity for intervention may be during the period of adolescent growth.

We followed up the children and their parents during 2006 to 2008, for detail nutritional and metabolic measurements, and cognitive function assessment in children (funded by Dept of Biotechnology, New Delhi). Follow- up of the children has provided growth curves for rural Indian children. It also has provided body composition data which confirms the tracking of the thin-fat Indian phenotype from birth into early puberty. It also confirmed our previous findings in urban cohort that children born small but grown big are at higher risk of adiposity and cardiometabolic disease. It has also provided confirmation that adiposity rebound is linked to risk of future adiposity and type-2 diabetes. We showed for the first time that maternal one-carbon metabolism during pregnancy is an independent predictor of adiposity rebound in the offspring. Neurocognitive analysis demonstrated that maternal one-carbon metabolism during pregnancy influences neurocognitive development and provides potential targets for intervention.

In the extended cohort of the PMNS, we tested (in an RCT) effect of vitamin B12 supplementation on the metabolic-endocrine features, body composition, neurocognitive functions and blood parameters in children during adolescence (2007-2008). We showed that vitamin B12 supplementation in physiological dose (2 mcg/day) was effective in reducing plasma homocysteine concentrations and elevating plasma B12 concentrations within 4 months. Addition of folic acid did not have any additional advantage. The supplementation over 12 months, however, did not show any effect on body composition, metabolic and cognitive functions.

The study is ongoing. We now plan to intervene the children at the age of ~15y with vitamin B12 and other micronutrients in an RCT, for a longer duration and measure the effects in the next generation as well.