Cardiovascular disease is the leading cause of mortality in type 2 diabetes (T2D), yet routine clinical monitoring often focuses on glucose-derived measures rather than insulin. The study tested whether fasting insulin provides information on cardiovascular mortality risk across the dysglycaemic spectrum and whether fasting insulin approaches non-diabetic levels with longer duration of oral-therapy T2D.
The investigators analysed six cycles of the National Health and Nutrition Examination Survey (NHANES) from 2007 through 2018, with mortality follow-up via the National Death Index through 2019. Baseline descriptive analyses used participants with complete kidney-function data classified as normoglycaemic, pre-diabetes, or oral-therapy T2D. Reported baseline group sizes were: normoglycaemic n = 2,913; pre-diabetes n = 3,993; oral-therapy T2D n = 1,381.
For mortality analyses the authors used the covariate-complete samples and applied survey-weighted Cox proportional hazards models. The covariate-complete mortality sample sizes reported were pre-diabetes n = 4,019 and oral-therapy T2D n = 1,387. Models adjusted for age, sex, race/ethnicity, body mass index (BMI), smoking status, physical activity, education, poverty:income ratio, and insulin assay generation. Outcomes included all-cause, cardiovascular, and cancer mortality. The exposure of interest was fasting insulin, modelled per +1 natural-log unit.
Key hazard ratios per +1 natural-log-unit higher fasting insulin were reported as follows:
In the group with oral-therapy T2D, geometric-mean fasting insulin remained above the normoglycaemic referent across T2D duration bands, reported as between 1.6- and 2.3-fold higher. The time-course analysis adjusting for BMI showed an early decline in fasting insulin during the first five years of treated T2D (−4.0% per year, P = 0.018) and little to no change thereafter (−0.1% per year, P = 0.66).
The authors interpret the inverse associations observed in treated T2D (higher fasting insulin associated with lower mortality) as most consistent with survivor effects and the presence of accumulated renal and vascular damage among longer-duration patients, rather than indicating a protective effect of higher insulin.
In this NHANES-based cohort analysis, higher fasting insulin predicted cardiovascular mortality among people with pre-diabetes. In individuals with oral-therapy T2D the relationship between fasting insulin and mortality was inverse; authors attribute this pattern to likely survivor bias and comorbid organ damage in treated T2D rather than a causal protective effect of higher insulin levels.
Across treated T2D duration bands, mean fasting insulin remained elevated compared with the normoglycaemic group (1.6–2.3×), and after adjustment for BMI a decline in fasting insulin occurred in the first five years of oral-therapy T2D but not thereafter.
These findings highlight that fasting insulin carries distinct prognostic information in pre-diabetes and that insulin levels do not simply normalize with treated T2D duration. The authors note that fasting insulin is seldom measured in routine practice relative to glucose metrics.
The analysis used de-identified, publicly available NHANES data collected by the National Center for Health Statistics; NHANES protocols are reviewed and approved by the NCHS Research Ethics Review Board and participants provided written informed consent at enrollment. The preprint authors declared no competing interests.
The report is a preprint and has not undergone peer review; the authors explicitly state that the findings should not be used to guide clinical practice until further validation. Data availability is public via the NHANES website as reported by the authors.
The authors caution that inverse associations in treated T2D are plausibly explained by survivor effects and accumulated renal and vascular damage rather than protective biology. Detailed limitations (for example, residual confounding, assay variation over cycles, or other sensitivity analyses) were not reported in full in the abstract; readers should consult the full preprint for complete methodological detail and limitations. The study used survey weights and adjusted for an insulin assay generation variable, but the abstract does not present additional sensitivity or subgroup results beyond those summarized above.
Fasting insulin may offer prognostic cardiovascular information in pre-diabetes that is not captured by glucose-focused measures. In treated T2D, persistently elevated fasting insulin despite therapy underscores the complexity of metabolic adaptation and the need for further research to determine clinical utility and causal pathways.