LoopWHI ’26

The Age & Generation Divide

biomarkers
11,427 reports
survey
9,097 respondents

Key takeaways

  • Different organ systems break in different decades — liver strain and low HDL peak in the 30s, glucose and kidney markers climb straight through to the 60s, and micronutrient deficits are worst in the 20s. There is no single "unhealthy age."
  • Glucose dysfunction doubles between the 20s and the 30s — from 16.7% to 33.0% — then keeps climbing to 69.1% by the 50s. It is the one marker that gets monotonically worse with every decade.
  • Liver strain peaks young. Elevated liver enzymes are most common in the 30s (23.1%) and fall by half by the 60s (9.5%) — the opposite of the curve most people assume.
  • B12 and vitamin D deficiency are worst in the 20s, not in old age. 86.7% of tested 20-somethings are vitamin-D deficient against 69.4% of the over-60s.
  • Stress falls with age while sleep gets shorter — high stress drops from 42.8% in the 20s to 30.5% in the 50s, but short sleep rises from 30.1% to 41.0% over the same span. The two behavioural signals move in opposite directions.

The working body does not deteriorate on a single timeline. Of the ten health signals Loop measured across the 2026 biomarker and survey datasets, no two peak in the same decade. Glucose dysfunction climbs without pause from the 20s to the 60s. Liver strain and low HDL top out in the 30s and recede. Micronutrient deficiency is worst in the youngest workers. Kidney markers wait until the 60s. Self-reported stress falls steadily across the working life even as sleep gets shorter. This is the central finding of the age cut, and it contradicts the way most workforce health is framed — as a steady downhill slope where every marker worsens together with age. India's non-communicable disease burden is itself unusually front-loaded onto the working years: the ICMR-INDIAB study places 101 million adults in clinical diabetes, with onset roughly a decade earlier than in Western cohorts. Inside that population, the age cut shows the disease load does not arrive all at once.

No two health signals peak in the same decade#

The clearest way to see the non-alignment is to put every marker on the same age axis and mark where each one is worst.

Health signal 20s 30s 40s 50s 60+ Worst decade
Glucose dysfunction (HbA1c ≥5.7%) 16.7% 33.0% 52.9% 69.1% 78.3% 60+ (still climbing)
Creatinine ≥1.3 0.6% 0.6% 1.5% 3.1% 5.0% 60+
Triglycerides ≥150 23.1% 37.3% 37.7% 40.2% 35.1% 50s
Short sleep (<6h) 30.1% 29.4% 34.4% 41.0% 50s
Uric acid (high) 16.1% 18.3% 15.3% 13.0% 21.7% U-shaped: 30s and 60+
Low HDL (<40) 41.8% 49.5% 45.0% 36.3% 36.9% 30s
Elevated liver enzymes 21.4% 23.1% 18.3% 13.8% 9.5% 30s
High stress (self-report) 42.8% 39.9% 37.2% 30.5% 20s
B12 deficiency (<300) 76.4% 71.0% 69.9% 60.2% 51.7% 20s
Vitamin D deficiency (<30) 86.7% 82.0% 78.9% 76.5% 69.4% 20s

No two health signals peak in the same decade

Ten markers on a shared age axis and a shared 0–100% scale — grouped by the decade each one is worst

biomarkers n = 11,427 · survey n = 9,097

Worst young — peak in the 20s

High stress (self-report)

B12 deficiency (<300)

Vitamin D deficiency (<30)

Worst in the 30s — peak in early-to-mid career

Elevated liver enzymes

Low HDL (<40)

Worst old — peak in the 50s and 60s (uric acid is U-shaped)

Glucose dysfunction (HbA1c ≥5.7%)

Creatinine ≥1.3

Triglycerides ≥150

Short sleep (<6h)

Uric acid (high) — U-shaped

One axis, ten curves kept apart on purpose: the peaks land in different decades. Survey signals (stress, sleep) have no usable 60+ cell. Biomarker and survey datasets are separate samples, not linked individuals.

Three families fall out of the table. The rising markers — glucose, creatinine, triglycerides — get worse with each decade and reach their worst in the oldest bands. The early-peaking markers — liver enzymes, low HDL, and on the survey side stress and the micronutrient deficits — are worst in the 20s or 30s and improve afterward. And uric acid sits on its own, low in the middle decades and elevated at both ends. The bands where the most repair is possible are not the same for any two systems.

Glucose is the one curve that never bends back#

Glucose dysfunction is the exception that makes the rule legible: it is the only signal that rises in every single decade, with no peak and no plateau inside the working life.

Age band Dysfunction rate n
20–29 16.7% 2,050
30–39 33.0% 2,627
40–49 52.9% 1,001
50–59 69.1% 751
60+ 78.3% 590

Glucose is the one curve that never bends back

Dysfunction rate (HbA1c ≥5.7%) by age band — the reference curve the rest of the page is measured against

16.7% → 33.0% is a doubling between the 20s and the 30s. The 50% line falls between the 40s and 50s. This is the only marker on the page that rises in every decade.

One in six in the 20s. One in three in the 30s. One in two in the 40s. The steepest single jump is the 20s-to-30s doubling, and it lands in the decade when work, parenting, sleep debt, and lengthening evening eating windows all intensify together. The full mechanism and the city and gender splits are on the Blood Sugar page; what matters here is the shape. Glucose is the marker that does not recover. Every other system on this page has a decade where its at-risk rate stops climbing. Glucose does not.

Liver strain and low HDL peak in the 30s, not in old age#

The two markers that contradict the "worse with age" prior most sharply are liver enzymes and HDL. Both are worst in early-to-mid career and improve from the 40s onward.

Age band Elevated liver enzymes Low HDL (<40)
20–29 21.4% 41.8%
30–39 23.1% 49.5%
40–49 18.3% 45.0%
50–59 13.8% 36.3%
60+ 9.5% 36.9%

Liver strain and low HDL peak in the 30s, not in old age

Elevated liver enzymes and low HDL by age — both top out in early career while glucose (faded) keeps climbing

n = 11,427
Glucose dysfunction (reference)Elevated liver enzymesLow HDL (<40)

Liver strain halves from 23.1% in the 30s to 9.5% by the 60s; HDL eases from 49.5% to 36.9%. The faded line is the glucose curve from above — the contrast in shape is the point.

Liver strain at the 30s peak runs at 23.1% and falls to under 10% by the 60s — a halving across the back half of the working life. The Liver page reads this early peak in detail, including the NAFLD link to insulin resistance that ties it back to the glucose curve above. Low HDL follows the same arc, worst in the 30s at nearly one in two, easing to roughly one in three by the 50s. Neither pattern is what an actuary modelling age alone would predict.

The HDL story rhymes. HDL — the lipid fraction that clears cholesterol from artery walls — is suppressed by the same metabolic load that elevates liver enzymes, and the suppression is heaviest in the 30s. The survivorship caveat applies at the oldest end: some of the apparent improvement after 50 reflects the people with the worst HDL having already moved into cardiac care or out of the working sample. But the 30s peak itself is robust and sits on the largest age bands in the data.

Micronutrient deficiency is a young-worker problem#

If liver and HDL break the age prior, the vitamins shatter it. Both vitamin D and B12 deficiency are most common in the youngest workers and least common in the oldest.

Age band Vitamin D deficiency (<30) B12 deficiency (<300)
20–29 86.7% 76.4%
30–39 82.0% 71.0%
40–49 78.9% 69.9%
50–59 76.5% 60.2%
60+ 69.4% 51.7%

Micronutrient deficiency is a young-worker problem

Vitamin D and B12 deficiency by age — the only two markers that fall continuously from the 20s

n = 11,427
Vitamin D deficiency (<30)B12 deficiency (<300)

The 20s endpoints — 86.7% vitamin D and 76.4% B12 — are the highest at-risk rates anywhere in the dataset. Both slope down where glucose slopes up.

Vitamin D deficiency among tested 20-somethings is 86.7% — the highest at-risk rate of any marker, in any age band, on this page. It eases by 17 points across the working life but never approaches normal. B12 deficiency follows the same downhill slope, from 76.4% in the 20s to 51.7% past 60.

Kidney and uric-acid markers wait for the oldest decades#

At the far end of the timeline sit the signals that stay quiet through most of working life and surface late. Creatinine — the kidney-filtration marker — is near zero until the 40s and only climbs meaningfully in the 50s and 60s.

Age band Creatinine ≥1.3 Uric acid (high)
20–29 0.6% 16.1%
30–39 0.6% 18.3%
40–49 1.5% 15.3%
50–59 3.1% 13.0%
60+ 5.0% 21.7%

Kidney and uric-acid markers wait for the oldest decades

Creatinine and uric acid kept on separate axes — their scales (0–5% vs 13–22%) are too different to share

n = 11,427

Creatinine ≥1.3 — flat, then a late upturn

Uric acid (high) — U-shaped

Creatinine is flat through the 20s and 30s, then rises eightfold to 5.0% by the 60s — the latest-breaking curve on the page. Uric acid is U-shaped: a 30s local peak (18.3%), a 50s trough (13.0%), and a 60+ high (21.7%).

Creatinine elevation runs below 1% through the entire first half of working life, then rises eightfold to 5.0% by the 60s. It is the most age-loaded marker in the dataset and the one closest to the naive "worsens with age" assumption — which is precisely why it is the exception rather than the pattern. Uric acid is stranger still: elevated in the 30s, lowest in the 50s, highest past 60. The dip-then-rise gives it two separate at-risk populations — younger men with diet-and-alcohol-driven hyperuricemia, and older adults whose kidneys clear urate less efficiently.

Stress falls with age; sleep gets shorter#

The behavioural layer, drawn from 9,097 survey responses, moves in two directions at once. Self-reported high stress declines steadily across the working life. Short sleep does the opposite.

Age band High stress Short sleep (<6h)
20–29 42.8% 30.1%
30–39 39.9% 29.4%
40–49 37.2% 34.4%
50–59 30.5% 41.0%

Stress falls with age; sleep gets shorter

High stress and short sleep by age band, on one 0–50% axis — the two signals trace an X

n = 9,097
High stressShort sleep (<6h)

Stress falls from 42.8% in the 20s to 30.5% in the 50s; short sleep climbs from 30.1% to 41.0% over the same span. The 50s band carries the smallest sample (sleep n=229, stress n=226) — directionally reliable, not decimal-precise.

The youngest band reports the most stress — 42.8% in the high range — and the rate falls in every decade after, to 30.5% in the 50s. Short sleep runs the other way: roughly stable through the 20s and 30s, then climbing to 41.0% past 50 — the same post-40 deterioration the Sleep page traces to age-driven changes in sleep architecture rather than mood. The two signals diverge from the 40s onward, where stress is already easing but sleep is getting worse.

Which decade prevention pays off in depends on which system#

The non-alignment has a direct operational consequence: there is no single best age to intervene, because the highest-leverage intervention is different for each system.

  • The 20s are the window for the micronutrient and HDL deficits — the markers already at their worst before career intensity peaks. Vitamin D and B12 are correctable with a tablet; HDL responds to movement. Cheap fixes, earliest payoff.
  • The 30s are the window for glucose and liver. Glucose has just doubled and is still reversible; liver strain is at its peak and is the most behaviour-driven, most recoverable signal on the page. This is the decade where prevention buys back the most future disease.
  • The 50s and 60s are where the kidney and uric-acid curves finally surface — late enough that the work there is management, not prevention.

The body keeps several clocks, and they do not agree#

The instinct behind most workforce wellness is that health is one curve that bends downward with age, and that the older the worker, the bigger the risk. The 2026 data does not support a single curve. It supports at least three. One climbs without pause — glucose, then kidney function behind it. One peaks in the 30s and recedes — liver, HDL, the metabolic strain of peak career load. One is already worst in the 20s and improves only because the deficit gets diagnosed and treated — the vitamins. Stress and sleep, the behavioural layer, pull apart in opposite directions across the same years.

References

  1. 1Anjana, R.M., Unnikrishnan, R., Deepa, M. et al. Metabolic non-communicable disease health report of India: the ICMR-INDIAB national cross-sectional study (ICMR-INDIAB-17). The Lancet Diabetes & Endocrinology 11, 474–489 (2023). https://doi.org/10.1016/S2213-8587(23)00119-5
  2. 2Eslam, M., Newsome, P.N., Sarin, S.K. et al. A new definition for metabolic dysfunction-associated fatty liver disease: An international expert consensus statement. Journal of Hepatology 73, 202–209 (2020). https://doi.org/10.1016/j.jhep.2020.03.039
  3. 3Holick, M.F. Vitamin D deficiency. New England Journal of Medicine 357, 266–281 (2007). https://doi.org/10.1056/NEJMra070553
  4. 4Ohayon, M.M., Carskadon, M.A., Guilleminault, C. & Vitiello, M.V. Meta-analysis of quantitative sleep parameters from childhood to old age in healthy individuals. Sleep 27, 1255–1273 (2004). https://doi.org/10.1093/sleep/27.7.1255