Composite RC-HbA1c index linked to higher stroke risk in Chinese adults

A simple mathematical combination of two common blood measurements—remnant cholesterol and glycated hemoglobin—was associated with an increased risk of stroke in a large nationwide cohort of middle-aged and older Chinese adults, according to a prospective study published in BMC Neurology. The so-called RC-HbA1c index, calculated by multiplying remnant cholesterol in milligrams per deciliter by HbA1c in percent, was linked to incident stroke over a prespecified seven-year observation period, and it offered a small but measurable improvement in stroke risk discrimination beyond conventional risk factors.

The analysis drew on 8,872 stroke-free adults enrolled in the China Health and Retirement Longitudinal Study, or CHARLS. During follow-up, 468 participants developed stroke. When researchers divided participants into quartiles according to their baseline RC-HbA1c index, stroke incidence rose steadily across the groups, from 3.7 percent in the lowest quartile to 7.3 percent in the highest, a gradient that was statistically significant. Compared with participants in the lowest quartile, those in the highest quartile had a 65 percent higher risk of stroke, with a hazard ratio of 1.65 and a 95 percent confidence interval of 1.18 to 2.31.

The study’s premise was that multiplying the two values into a single index might capture the combined burden of lipid and glycemic metabolic stress on blood vessels more simply than measuring and interpreting each marker separately.

Statistical modeling suggested the relationship was straightforward. Restricted cubic spline analysis, a technique that tests whether an exposure-risk association curves or bends at higher exposures, found no statistical evidence of departure from linearity, meaning stroke risk appeared to rise in approximate proportion to the index across its observed range rather than accelerating past a threshold.

The predictive performance results were more nuanced. Adding the RC-HbA1c index to a basic model of conventional risk factors raised the C-statistic, a measure of how well a model distinguishes people who later have an event from those who do not, from 0.6584 to 0.6665. The index also produced a significant continuous net reclassification improvement of 0.064, indicating that a modest share of participants were shifted into more accurate risk categories.

Head-to-head comparisons sharpened the picture. Against a model containing remnant cholesterol alone, the composite index delivered a small but statistically significant gain in discrimination, with a difference in the C-statistic of 0.00217. However, when the index was compared with a model that included remnant cholesterol and HbA1c as separate predictors, its performance was not significantly different, with a C-statistic difference of −0.00141 that was statistically indistinguishable from zero. In other words, the index condensed the two markers into one convenient number without adding predictive information beyond what the two components already provided together.

The study also examined whether sustained exposure mattered more than a single snapshot. Among 5,412 participants who had measurements at both Wave 1 and Wave 3 of CHARLS, the researchers calculated cumulative RC-HbA1c exposure and used Wave 3 as the landmark for follow-up. High cumulative exposure was associated with nearly double the stroke risk, with a hazard ratio of 1.99 and a 95 percent confidence interval of 1.35 to 2.93.

The authors conclude that the RC-HbA1c index was independently associated with incident stroke and provided a small incremental improvement in discrimination beyond conventional risk factors and remnant cholesterol alone, while performing no better than the two components used separately. They position the index as a potentially simple composite marker of lipid-glycemic metabolic status rather than a ready clinical tool, noting explicitly that its clinical utility and generalizability require further evaluation. Because the cohort consists of middle-aged and older Chinese adults, and because the index’s advantage over simply measuring both markers separately was minimal, the findings are best read as an observational association in a specific population rather than evidence that routine use of the index would improve stroke prevention in practice.

Subject of Research: The association between a composite remnant cholesterol–HbA1c index and incident stroke in middle-aged and older Chinese adults

Article Title: Association of the RC-HbA1c index with incident stroke in middle-aged and older Chinese adults: a nationwide prospective cohort study

Article References: Wu, T., Cao, Y., Wang, Y., Liu, X., & Zhang, H. (2026). Association of the RC-HbA1c index with incident stroke in middle-aged and older Chinese adults: a nationwide prospective cohort study. BMC Neurology. https://doi.org/10.1186/s12883-026-05492-2

Image Credits: AI Generated

DOI: 10.1186/s12883-026-05492-2

Keywords: stroke, remnant cholesterol, HbA1c, CHARLS, prospective cohort, risk prediction, lipid-glycemic index, China, cardiometabolic risk, epidemiology, Association, RC-HbA1c