Blood Sugar

Of 7,088 HbA1c readings Loop ran on urban working professionals in the latest 12-month window, 38.7% sit at or above the threshold where the body has stopped managing blood sugar well — virtually unchanged from the 37.2% reported in WHI 2025 on a sample four times smaller. In a country where the ICMR-INDIAB national study already places 101 million adults in clinical diabetes and another 136 million in pre-diabetes (Anjana et al., Lancet Diabetes & Endocrinology, 2023), this band — the workforce's early warning, in front of the diagnosed group — is where prevention is still cheap, still effective, and still mostly missed.

The distribution:

The population mean of 5.80% sits inside the normal range only by a hair. A healthier population would centre near 5.2–5.3%. What's moving here is the centre of the distribution, not just the tail.

The headline rate held across a sample that grew 4×#

All thresholds match exactly across years. The 2026 sample is broader — more contributing companies, wider geographic spread — and gave the same answer. The pre-diabetic band widened by 1.7 points; the diabetic share held; the average mean barely moved.

Pre-diabetes is where the action is. It's the band where the pancreas is still keeping up, the patient is still asymptomatic, and twelve weeks of structured intervention can still bend the trajectory. It's also the band the Indian healthcare system has the hardest time seeing.

Detected risk runs ~7× ahead of diagnosed cases#

Biomarker prevalence and self-reported diagnosis don't agree at audience level.

For every employee who knows they have diabetes, biomarker evidence suggests at least one more does. For every diagnosed case, five or six more sit in the early-warning band where the disease has not yet arrived but the trajectory toward it is well underway.

The mechanism is no mystery. Most Indian working professionals don't get a fasting blood panel unless something has already gone wrong. Three quarters of our survey audience hadn't done a full-body diagnostic test in the last twelve months — even though their employer's group policy almost certainly covers one. The Preventive Care page traces this screening gap directly, and shows it widening down the income scale.

Glucose dysfunction roughly doubles between the 20s and the 30s#

One in six in the 20s. One in three in the 30s. One in two in the 40s. Seven in ten in the 50s. By 60, four in five.

The slope between 30 and 50 is steeper than any other twenty-year stretch a working adult lives through. The 30s, in our data, are where the curve breaks — and the contributing pressures aren't a mystery. Sleep compressed by infant care, commute, and screens. Eating windows that lengthen as evenings stretch later. The intensification of work in a decade that intensifies in every other way. The biomarker doesn't tell you which of these matters most. It tells you the body is registering all of them at once.

For an employer staffing senior bands, this stops being a wellness story and becomes a workforce-continuity one. The 40s and 50s are where most operating teams, managerial layers, and institutional memory live. They are also where the majority of the workforce now sits in pre-diabetic or diabetic glucose ranges.

The gender gap is age-dependent, not flat#

The 2025 female sample was 770; the 2026 sample is 3,262 — four times larger and more representative. The cleaner read is that the new estimate is the more accurate one, not that women's health deteriorated in twelve months.

But the population-average framing — "women have caught up to men" — is incomplete. The gender pattern depends on age.

In the 20s, women are marginally worse off than men. In the 30s and 40s the curve flips — men pull ahead, and the gap is sharpest in the diabetic range, where men in their 30s are 2.4× more likely than women in the same band to be in diabetic-range glucose (10.3% vs 4.3%). The gap closes from the 50s onward, and by 60 the rates are identical.

The clinical implication is concrete. Screening protocols and follow-up intensity calibrated against population averages will under-screen the 35-year-old male and over-screen the 35-year-old female — and the gap is largest in the band where chronic-disease management is also cheapest.

Dependents are 20 points worse than employees#

The biomarker file separates "self" from "others" but doesn't break dependents into spouse, child, parent. (The claims data does — see the Family Claims page.) The dependent population skews older, more often female, and frequently includes parents and parents-in-law covered under the policy.

More than one in two dependents already shows glucose dysfunction, against fewer than one in three employees. The 20-point gap is partly biology — dependents are older. Most of it is access. Employees sit inside the workplace screening default; dependents don't. The employee is at the annual checkup, the company webinar, the on-site fitness initiative. The parent is not.

The disease shows up in cardiac, renal, and ophthalmic admissions#

The same broad audience whose biomarkers show 38.7% glucose dysfunction is the audience whose hospital admissions concentrate in metabolic-syndrome-adjacent categories.

The endocrine and metabolic line is the most direct link to glucose dysfunction. But the more revealing read is the downstream conditions. Cardiac and circulatory admissions — the rupee tail the Top Categories by Incurred page tracks in full — account for roughly one rupee in every twelve of the policy's spend, at over a lakh per claim. Eye and renal admissions — diabetic retinopathy, diabetic nephropathy in the patients who reach them — account for another big chunk.

Mumbai now leads the metros#

The WHI 2026 biomarker file records a six-digit pincode for every report. Mapping the first three digits to a metro bucket via the India Post allocation scheme covers 87% of the sample. The picture is more interesting than the 2025 ordering suggested.

Mumbai leads on every glucose metric in the dataset — highest dysfunction rate, highest diabetic-range share, second-highest average HbA1c. The 2026 sample composition is broadly stable across years; the worsening is real. Within Mumbai, 49.8% of women tested are in the dysfunction band (n=426) against 43.9% of men. Combined with the city's female anemia rate (52.3%) and female HDL <50 rate (78.1%), Mumbai is uniformly hostile to women's biomarkers in this dataset.

The Bengaluru advantage is concentrated in working-age cohorts. In the 20s, Bengaluru sits at 13.3% dysfunction against Mumbai's 19.5% and Pune's 24.2% — a six-to-eleven-point lead. By the 60s, all three metros converge between 78% and 86%. Whatever Bengaluru's working life is doing right — lower stress, hybrid prevalence, age skew — doesn't survive into the post-retirement window.

Two movements warrant a methodological caveat. Pune's jump from 30.2% to 43.2% is mostly sample composition — 2025 sampled 384 reports; 2026 sampled 2,562 (seven times larger, broader spread). The 2026 number is the more reliable estimate. Delhi's apparent drop from 43.1% to 32.7% is the same artefact in reverse — Delhi's 2025 n=130, 2026 n=689. Delhi sits closer to the middle of the metro distribution than the 2025 ordering placed it.

What survives across both years: every major metro is in a range of 30–47% on glucose dysfunction. Bengaluru is consistently at the lower end; Mumbai is consistently at the upper end.

Fasting glucose corroborates the HbA1c picture#

One in five of the fasting-glucose subgroup shows impaired fasting glucose. The fasting test is more sensitive to recent behaviour, less to long-term trend. The 20% reading is corroboration of the HbA1c headline, not a substitute for it — two witnesses to the same metabolic state, looking at it through slightly different windows.

The bottleneck isn't the test — it's the callback#

Indian primary care, as the working professional encounters it, does not routinely screen for glucose dysfunction in the absence of symptoms. The general practitioner is paid by visit, not by panel of patients tracked over time. The corporate annual checkup, when it happens, produces a PDF and an empty inbox. No callback. No follow-up appointment. No one whose job it is to read the number and act on it.

The HbA1c test itself costs roughly ₹400 retail, less in bulk. The bottleneck isn't the test. It's what doesn't happen after the result comes back.