What Changed Since WHI 2025

WHI 2025 was one photograph of the urban workforce's metabolic and behavioural health. WHI 2026 is the second edition, on samples several times larger. Reported stress is higher and low Vitamin D is less common; elevated blood sugar and low B12 sit at close to the levels WHI 2025 found; and a long tail of remembered 2025 figures turns out not to be comparable at all.

What moved: high stress up, low Vitamin D down#

Differences between shares are given in percentage points (pp). Stress, sleep and HDL use the earlier edition’s published headline or duration figure; the notes below explain conflicting breakdowns and definition limits.

Stress is higher on both headline measures, although WHI 2025’s published breakdowns do not all agree on the starting point. Both the share above the cut-point and the mean rating are higher: 40.4% against 33.9%, and 5.56 against 5.01. Among female responses the figure is 47.4% of 2,299 against 41.2% of 1,125, and among male responses 37.3% of 5,021 against 30.5% of 2,304. The published wording does not establish the same recall period: WHI 2025 described the item as daily pressure without stating a window, and WHI 2026 asks about the previous week.

WHI 2025's stress headline and sex table conflict with its other breakdowns. All imply higher stress in 2026, but the size of the difference is uncertain; the earlier recall period is also unstated.

Vitamin D moved the other way at all three published thresholds. The shares below 20 ng/mL and below 10 ng/mL are each 5.8 points lower. The share at or above 30 ng/mL rose from 15.9% to 18.6%. These are differences between tested samples; six in ten reports are still below 20 ng/mL.

Low HDL cholesterol is less common at the fixed cutoff of 40 mg/dL: 44.0% of 7,260 reports, against 47.7% in WHI 2025.

Alcohol lines up band for band across the two editions. Any use went from 38.7% of 3,436 responses to 37.3% of 7,249, and beneath that total both editions publish the same four frequency bands: less than once a week, 30.4% to 29.5%; one to two times weekly, 6.7% to 5.4%; three to four times, 1.0% to 1.4%; five or more, 0.7% to 0.9%.

Sleep needs a more qualified comparison. WHI 2025’s duration table put 27.3% of 3,437 responses below six hours. WHI 2026’s two shortest answers, “under 5 hours” and “5 to 6 hours”, hold 30.7% of 7,569 responses — 3.4 points more. The earlier sex table implies about 28.9% on 3,431 responses instead, so the baseline is not fully reconciled. The 5-to-6-hour option also prevents treating the current measure as strictly below six hours. The table compares the published shortest bands, not a confirmed increase on an identical duration definition.

The exercise rows describe the shape of the distribution. Both editions ask the same question — days in a typical week with at least 30 minutes of exercise — and offer the same four answers. The one-to-two-day share is 2.3 points smaller, the zero-day share 2.9 points larger, and the three-or-more share is similar, 37.9% to 37.2%. The samples are separate, so this says the WHI 2026 distribution sits lower, not that particular people stopped exercising.

SGPT at or above 40 U/L is 2.3 points higher. This is the cutoff the two editions share: WHI 2025 counted 335 of 2,039 reports at or above it, on a rule that treats anything under 40 U/L as normal. The narrower SGPT-only measure on the Liver page for WHI 2026 is the above-40 figure, 17.9%. B12 below 200 pg/mL is also 2.3 points higher, while B12 below 300 pg/mL is 1.4 points higher — so the overall B12 share is close to where it was while more of the distribution sits under the lower cutoff.

Exercise comparisons use WHI 2025's count-backed distribution. Other published versions conflict with it, so this is a comparison with that series, not a reconciled baseline.

What held: elevated blood sugar, low B12 and anemia at similar levels#

Blood sugar is the anchor. HbA1c reflects average blood glucose over about the previous three months, with 5.7% the boundary of the prediabetes range and 6.5% the boundary of the diabetes range. A result in the diabetes range usually needs a second test for diagnosis (NIDDK, The A1C Test & Diabetes). HbA1c was at or above 5.7% in 38.7% of 7,088 reports against 37.2% of 1,649 — 1.5 points apart — and the mean landed within a hundredth of a point. Inside the total, the pre-diabetic band is 1.7 points wider and the diabetic band is similar, 13.5% to 13.3%.

The two inflammation means are close. ESR measures how fast red blood cells settle in a tube. Inflammation can make them clump and settle faster, but ESR alone cannot identify its cause (MedlinePlus, Erythrocyte Sedimentation Rate). Among female reports the mean is 19.23 mm/hr across 3,483 reports against 19.3 across 911, and among male reports 8.54 across 4,431 against 8.8 across 1,175 — on a combined tested sample 3.8 times larger. Each sex is compared with its own WHI 2025 value. The two means can be divided, but the result is a ratio of averages, not a ratio of disease burden.

Anemia is a fall inside a level that is still high. Reported anemia among female reports went from 36.5% of 1,214 to 34.6% of 4,375, and more than one in three female reports is still below the cutoff. The male rate is similar on its own cutoff. WHI 2026 uses haemoglobin below 12 g/dL for women and below 13 g/dL for men; WHI 2025 published rates and counts but no numeric threshold, so the two columns are reported rates rather than a matched measurement. A second WHI 2025 publication gives 37.1% and 8.2% for the same measure, without counts or a sample; those are published WHI 2025 values too. The figures above come from the series with published counts; its female and male tested samples sum to 2,719.

The glucose ranking differs between editions#

Both editions publish a per-metro share of HbA1c reports at or above 5.7%, on the same threshold, with the tested sample beside each. The samples matter as much as the shares.

City samples and definitions differ between editions. Their movements are sampling and definition differences, not city health trends.

In WHI 2025, Pune had the lowest share of the five metros and Delhi NCR the highest. In WHI 2026 their positions relative to each other are reversed: Delhi NCR is second-lowest and Pune second-highest. Mumbai has the highest share of these five in WHI 2026, at 46.8% of 872 reports. The five metros are not the whole biomarker sample in either edition.

What the second edition shows#

Two editions of the same report, on samples several times larger the second time, show similar overall levels on several measures: overall HbA1c and its two bands, as well as its mean, B12 below 300 pg/mL, reported anemia in each sex, and mean ESR in each sex. Those are similar measured levels in two tested samples — which is worth knowing, and is not by itself proof of an unchanged rate in the workforce.

What differs is narrower than a year-on-year page usually reports, and more specific. Reported stress is higher among the responses to both editions, and higher among women. The shares below the Vitamin D cutoffs are lower, and the share at or above 30 ng/mL is higher. Six in ten reports remain below 20 ng/mL. Low HDL is lower on the shared cutoff. The shortest sleep bands hold more of the sample. And the exercise distribution sits lower, with a smaller one-to-two-day share.

The second edition also shows how much of a year-on-year table is not a year-on-year finding. The table above separates changes in definitions, conflicting published figures and missing information needed to match the measures. None is evidence that the workforce itself stayed unchanged. The metro ordering rests on WHI 2025 samples of a few hundred reports per city. And one entire dataset has no first frame, so the cost side of WHI 2026 is a baseline rather than a change.