Flat 1.00 to 5.6%. Not a DIY intensive target
This one has trial evidence — in high-risk type 2 diabetes. ACCORD randomized to an intensive sub-6% target and all-cause death went up, HR 1.22.
A three-month glucose average. The overlay is Zhong’s 1.03 per 1% in people without known diabetes — and that association collapses after undiagnosed diabetes is excluded. ACCORD raised death aiming under 6%.
Dose
Flat HR 1.00 from 4.0–5.6%. Then 6.0 = 1.15, 6.5 = 1.35, 7.0 = 1.55, 8.0 = 1.95, 10.0 = 2.7. Council 2026-08-29: 3–0 removed the old 0.88 low-end reward; 2–1 against adding a low-end uptick. 1.15 at 6.0% is not Zhong’s 1.03 per 1%.
1.15 at 6.0% is constructed — Zhong’s continuous slope is 1.03 per 1%, and that association is NS once undiagnosed diabetes is excluded.
Dose-response shape: monotone
How to get there
The behavioural half is the practice: the Zone 2 protocol already on this site is what you actually do. Medication is the other half — intensive lowering toward a sub-6% target is a clinician's call, not a DIY target.
HbA1c blood test, a lab average over about 90 days.
ACCORD randomized high-risk type 2 diabetes to an intensive target under 6% and all-cause death went up, HR 1.22. This is explicitly not a DIY intensive target; the headline already says so, and the on-ramp repeats it.
Evidence, by endpoint
| Endpoint | Grade | Citation |
|---|---|---|
| All-cause mortality per 1% HbA1c, without known diabetes | B | Zhong et al. 2016, Scientific Reports Engine overlay (page meta, not the knot table). 12 studies; ACM dose-response 113,526 people / 11,301 deaths. HR per 1% 1.03 (1.01–1.04), I² 28.9%. Relatively flat below ~5.7%, then steeper. After excluding undiagnosed diabetes (HbA1c ≥6.5%): ACM HR 1.01 (0.99–1.03). After also excluding prediabetes: ACM HR 1.01 (0.98–1.03), NS. Authors: driven by undiagnosed diabetes or prediabetes. |
| All-cause mortality, 6.0–<6.5% vs 5.0–<5.5%, non-diabetic 50+ | C | Schöttker et al. 2016, BMC Medicine IPD of 6 European/US cohorts, 28,681 non-diabetic adults 50+, 6,769 deaths, mean 10.7 y. Increased (6.0–<6.5%) vs low (5.0–<5.5%) ACM HR 1.14 (1.03–1.27) after CV-risk adjustment. Very-low (<5.0%) pooled NS. Authors: data do not support a J-shaped association. Engine 6.0 = 1.15 is nearby; 6.5 = 1.35 is outside this paper. |
| All-cause mortality, categorical U in non-diabetics | C | Cavero-Redondo et al. 2017, BMJ Open 74 studies / 46 in the meta-analysis. Non-diabetic ACM >6.0% HR 1.74 (1.38–2.20); <5.0% HR 1.19 (1.04–1.36). Diabetic >9.0% 1.69; <6.0% 1.57. Authors: optimal 5.0–6.0% without diabetes, 6.0–8.0% with diabetes. Do not paste 1.74 onto the engine’s 1.15 at 6.0%. |
| All-cause mortality from intensive HbA1c lowering in T2D | A | ACCORD Study Group (Gerstein et al.) 2008, New England Journal of Medicine High-risk T2D, intensive target <6.0% (median achieved 6.4%) vs standard 7.0–7.9% (median 7.5%). Primary MACE HR 0.90 (0.78–1.04), NS. ACM 257 vs 203, HR 1.22 (1.01–1.46), P=0.04. Targeting a number this calculator scores as 1.00–1.15 raised death in that trial. |
| All-cause mortality, HbA1c <5.0% vs 5.0–<5.7% | C | Aggarwal et al. 2012, Diabetes Care ARIC, n=13,288. HbA1c <5.0% vs 5.0–<5.7%: ACM HR 1.32 (1.13–1.55); cancer death 1.47 (1.16–1.84). Authors: a generalized marker of mortality risk, mixing healthy people with people whose low HbA1c signals illness. Council minority for a 1.05–1.10 uptick below 5.0%; not implemented. |
| All-cause mortality per 1% in men, including the “normal” range | B | Khaw et al. 2001, BMJ EPIC-Norfolk, 4,662 men 45–79. Each 1% higher HbA1c: +28% ACM (P<0.002). After excluding known diabetes, HbA1c ≥7%, or prior MI/stroke: RR 1.46, P=0.05. 82% of excess population mortality at HbA1c 5.0–6.9%. Lowest rates below 5% — not a low-end-uptick paper. |
| All-cause mortality, prediabetes by definition | C | Cai et al. 2020, BMJ 129 studies, 10.07 million. General-population prediabetes ACM RR 1.13 (1.10–1.17). ACM raised for HbA1c-IEC 6.0–6.4% (1.21, 1.06–1.38) and for IFG/IGT; not for the broader HbA1c-ADA 5.7–6.4% band. Binary prediabetes is not this continuous curve. |
What the dose is
Flat hazard 1.00 from 4.0% through 5.6%. Then the constructed upper limb: 6.0 = 1.15, 6.5 = 1.35, 7.0 = 1.55, 8.0 = 1.95, 10.0 = 2.7. A 2026-08-29 council vote (3–0) removed an old 0.88 reward below 5.6%; a 2–1 vote declined a low-end penalty. Schöttker et al. 2016 was purpose-built to test that J and did not support it after confounders.
What the evidence shows
Zhong et al. 2016 (Scientific Reports) is the overlay: without-known-diabetes dose-response HR 1.03 (1.01–1.04) per 1%, relatively flat below ~5.7%. After excluding undiagnosed diabetes the ACM association is 1.01 and not significant. Schöttker’s IPD puts 6.0–<6.5% at 1.14 versus 5.0–<5.5% in non-diabetic adults 50+. Khaw et al. 2001 found 82% of excess population mortality in men at 5.0–6.9% — inside what many people call normal.
What the number is not
It is not Zhong’s 1.03 pasted onto 6.0% (the engine prints 1.15). It is not Cavero-Redondo’s 1.74 / 1.19 U. It is not a licence to intensive-lower toward 5.6% — ACCORD (Gerstein et al. 2008) raised all-cause death (HR 1.22) aiming under 6% in high-risk type 2 diabetes, which is why that endpoint is grade A and the safety line exists. It is not Aggarwal’s <5.0% penalty; that was considered and not implemented.
What “prediabetes” does and does not mean here
Cai et al. 2020 (129 studies, 10.07 million): general-population prediabetes ACM RR 1.13, but the HbA1c-ADA band of 5.7–6.4% did not raise ACM while the tighter IEC band of 6.0–6.4% did (1.21). This slider is a continuous percentage, not a binary prediabetes label, and Huang et al. 2025’s low-end 1.48 is elderly CKD — not this calculator’s population.
What argues against this
Cavero-Redondo et al. 2017 is the strongest opposing shape: non-diabetic ACM HR 1.74 above 6.0% — well above the engine’s 1.15 — and HR 1.19 below 5.0%, a U the engine’s flat-to-5.6% curve does not implement. Zhong et al. 2016’s own 1.03 per 1% collapses to 1.01 (NS) after undiagnosed diabetes is excluded. ACCORD (Gerstein et al. 2008) randomized high-risk type 2 diabetes to an intensive target under 6% and raised all-cause death (HR 1.22). Aggarwal et al. 2012 (<5.0% HR 1.32) was considered and outvoted as a low-end penalty. Huang et al. 2025 (Q1 3.7–5.3% HR 1.48) is elderly non-diabetic CKD — wrong population.
Gotchas
- 1.15 at 6.0% is constructed. Zhong’s continuous slope is 1.03 per 1%, and it is NS after undiagnosed diabetes is out.
- 6.5 = 1.35 is outside Schöttker’s 6.0–<6.5% vs 5.0–<5.5% contrast of 1.14.
- ACCORD’s intensive arm achieved 6.4%, not 5.6%. The trial is a warning against DIY intensive targeting, not a refutation of the observational slope above 6.5%.
- Cai’s prediabetes ACM depends on the definition: IEC 6.0–6.4% raised death; ADA 5.7–6.4% did not.
- Khaw’s lowest rates were below 5%. Do not use Khaw backwards as a low-end-uptick paper.
Safety
- Do not chase a number this calculator scores as 1.00–1.15 with intensive glucose lowering. ACCORD raised death doing that in high-risk type 2 diabetes. Lab targets belong with a clinician.
Educational content only — not medical advice, diagnosis, or treatment. Not a medical device; not FDA evaluated. Consult a qualified healthcare professional about your own health, and call emergency services for urgent symptoms. Full medical disclaimer →