Indian Babies Are Born Thin and Fat at the Same Time

In the 1990s a research team in Pune began measuring newborns the way a tailor measures a customer. Not just weight. Head, upper arm, waist, and the thickness of a pinch of skin below the shoulder blade. Then they set the numbers beside babies born in Southampton, England.

The Pune babies averaged about 2.7 kilograms. The Southampton babies averaged about 3.5.

The size gap was expected. What the tape measure found underneath it was not. The Indian newborns were smaller in every dimension, but not by the same amount everywhere. The measurements that had shrunk most were the abdomen and the upper arm: the internal organs and the muscle. The one that held up best was the skinfold. The fat.

C. S. Yajnik, who ran the Pune Maternal Nutrition Study, called it the thin-fat baby, and the name stuck. Small Indian babies, the 2003 paper concluded, have small abdominal organs and low muscle mass but preserve their body fat, a build that may persist after birth and predispose them to insulin resistance.

Unpack that last phrase, because everything here rests on it. Insulin is the hormone that tells your cells to pull sugar out of the blood. Insulin resistance means the cells stop responding properly, so the pancreas has to make more and more of it to do the same job. Type 2 diabetes is what happens when the pancreas can no longer keep up and blood sugar stays high.

101 million people, and a longer queue behind them

India’s national survey, ICMR-INDIAB, examined 113,043 adults across 31 states and union territories in phases between 2008 and 2020, projected the totals to 2021, and reported in The Lancet Diabetes and Endocrinology in 2023. It found diabetes in 11.4 percent of adults, roughly 101 million people.

It also found prediabetes in 15.3 percent, about 136 million. Prediabetes means blood sugar above the normal range but not yet high enough to be called diabetes. More Indians are in that queue than have already reached the diagnosis.

The shape is as unusual as the size. A 2022 study in Diabetologia comparing Indian and Scottish records put the median age at diagnosis at 47 in one large Indian cohort, meaning a group of patients tracked together, against 62 for white Europeans. Among Asian Indians diagnosed at 40 or younger, between a quarter and two fifths had a normal BMI, the rough weight-against-height number clinics use to sort patients, judged against the lower threshold used for South Asians. Among young white Europeans with diabetes, 9.3 percent did.

Younger, then, and often at a body size that would not worry a doctor in Europe.

A body budgeted for scarcity

In 1992, in the journal Diabetologia, C. N. Hales and David Barker put forward the explanation that has shaped the field since. They called it the thrifty phenotype, phenotype being the body you actually end up with rather than the genes you started with.

In plain terms: a fetus short of nourishment makes trade-offs, in a sensible order. It protects the brain. It holds on to fat, which is stored fuel. It economises on what it can manage with less of, including muscle, the tissue that soaks up most of the sugar in your blood, and the beta cells, the cells in the pancreas whose job is to make insulin.

That is a shrewd bet if the world outside the womb is short of food too. It is a poor bet if the child grows up on polished rice and refined flour, works at a desk and rides to the office.

The winter that turned into an experiment

You cannot starve pregnant women to test any of this. History did it anyway.

From November 1944 to May 1945 a German blockade cut food supplies to the western Netherlands, and city rations fell to a fraction of what an adult needs. The Dutch kept careful birth records throughout, so researchers could later identify exactly who had been in the womb during those months and follow them for life.

Those people proved more likely as adults to develop type 2 diabetes and heart disease. When in the pregnancy the hunger struck shaped what surfaced later: exposure in the first months, when every organ is being laid down, produced the most pronounced effects, while mid-pregnancy exposure tracked more with the kidneys and lungs.

Then a Dutch team looked at the DNA itself. Roughly six decades on, people conceived during the famine carried slightly fewer chemical tags on a growth gene called IGF2 than their own brothers and sisters conceived before or after it. Those tags, called methylation, sit on top of DNA and help set how loudly a gene gets read; the genetic letters underneath are unchanged. The gap was about two percent, and only in people whose exposure covered conception itself.

A chemical mark is not a disease, and nobody has shown these particular marks cause diabetes. What the finding does establish: a few months of hunger before birth can leave a physical trace still readable sixty years later.

What the evidence does not say

Put the pieces together and one claim stands up: undernutrition around the time of birth raises an individual’s later risk of type 2 diabetes. India has plenty of that risk in circulation now, not only in its past. In an analysis of the most recent National Family Health Survey, covering 2019 to 2021, 17.1 percent of Indian babies were born under 2,500 grams.

A version of this story travels well online: the famines of British rule starved a population into a metabolic curse its great-grandchildren are still paying off. Those famines were real, enormous and shaped by policy, as the grain that kept leaving a starving country and Bengal in 1943 both show, but the causal chain from them to India’s diabetes clinics has not been demonstrated, and three things stand in its way.

The first is that famine studies measure the wrong people. The Dutch work followed those who were fetuses during the famine, born in 1944 and 1945. It says nothing about descendants born generations later.

The second is that no Indian study of the kind exists. Nobody assembled a group exposed in the womb to the famines of the 1870s or 1943 and tested them for diabetes decades later. The Indian argument is borrowed from the Netherlands and from China.

The third is that the Chinese loan has gone bad. The Great Leap Forward famine of 1959 to 1961 produced the largest body of famine and diabetes research, most of it reporting more diabetes among the famine-exposed. In 2022 Chihua Li and L. H. Lumey pooled those studies in Nutrients and found that eighteen of the twenty-three had compared famine births only with people born after the famine, who were in every case at least three years younger. Diabetes climbs with age. With comparison groups matched for age, the apparent effect shrank from roughly a 50 percent excess risk to 12 percent. That smaller figure is still real, so something may survive the correction, but it is a fraction of what was claimed. Their verdict on the idea that the famine explains China’s epidemic: premature.

Capacity on one side, load on the other

The framing that survives all this comes from a 2016 paper in Frontiers in Public Health by Jonathan Wells and colleagues, Yajnik among them.

They split the problem in two. Metabolic capacity is the body’s ability to handle fuel safely: how much insulin the pancreas can make, how much sugar the muscles can take up, both heavily dependent on growth before and just after birth. Metabolic load is what that machinery must cope with: body fat, refined carbohydrate, sitting still. Diabetes, in their words, occurs when the load exceeds the capacity.

That model fits the Indian timeline better than any famine does. Generations of poor nutrition may have left the population’s capacity low. What changed recently, and fast, is the load: cities, cheap refined carbohydrate, cooking oil, office work, two-wheelers. Early undernutrition loads the gun. The modern environment pulls the trigger.

Genes are in the room as well. The same 2022 Diabetologia study found Asian Indians had lower genetically determined beta cell function than white Europeans, with about 45 percent carrying eight or more gene variants linked to weaker insulin production, against 25 percent. Wells and his co-authors judge the genetic share real but probably small. Either way, a purely environmental account is incomplete.

The lever nobody can pull quickly

What follows from the science is unglamorous. If capacity is built before birth and in the first years of life, the nutrition of girls, of women before and during pregnancy, and of small children is the part of the system still open to change. None of that is advice for anyone already living with a diagnosis, which is a matter for their doctor; it is a point about where the population lever sits.

The catch is the timetable. Better-fed mothers today do not show up in diabetes statistics for thirty or forty years, longer than any budget cycle and most political careers.

The 101 million is already fixed. The number three decades from now is being set, gram by gram, in maternity wards this week.