📄 other

Longitudinal Hemoglobin A1c–Systolic Blood Pressure Index and Risk of Incident Stroke in Middle-aged and Older Chinese Adults: A Prospective Cohort Study

This prospective cohort study of Chinese adults demonstrates that higher longitudinal hemoglobin A1c–systolic blood pressure index (HSI), particularly persistent elevation over time, is independently associated with an increased, dose-dependent risk of incident stroke.

Original authors: Zepeng Ren, Jianjing Qiao, Xijin Liu, Yike Liu¹, Jiajia Li, xi liu

Published 2026-07-25

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Zepeng Ren, Jianjing Qiao, Xijin Liu, Yike Liu¹, Jiajia Li, xi liu

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine your body as a bustling city where two main utility companies keep everything running: the "Sugar Grid" and the "Pressure Pipes." The Sugar Grid delivers fuel (glucose) to every neighborhood, while the Pressure Pipes push that fuel through the streets. In a healthy city, both systems work smoothly. But sometimes, the Sugar Grid gets sticky and clogged (high blood sugar), and the Pressure Pipes get squeezed too tight (high blood pressure). When these two problems happen at the same time, they don't just add up; they team up to cause chaos, damaging the delicate roads and bridges (your blood vessels) that lead to the city's most important building: the brain.

Scientists have long known that high sugar and high pressure are bad for your brain, but they usually checked them separately, like checking the fuel gauge and the tire pressure on a car one at a time. This study asks a new question: What if we combined them into a single "Stress Score" to see how much total wear and tear they cause together? The researchers wanted to know if this combined score could predict a "traffic jam" in the brain (a stroke) better than looking at the numbers alone, and whether the story changes if we look at just one snapshot in time versus watching the city over several years.

The Big Picture: A Scorecard for Stress

This study, conducted by researchers from Ordos Central Hospital and the People's Hospital of Dongsheng District, looked at nearly 10,000 middle-aged and older adults in China. They created a special "Stress Score" called the Hemoglobin A1c–Systolic Blood Pressure Index (HSI). Think of HSI as a single number that multiplies your blood sugar level by your blood pressure level. It's like a "danger multiplier" that tells you how much combined stress your blood vessels are under.

The team didn't just take one photo of the participants' health; they watched them over time. They looked at three different ways to use this score:

The Snapshot: What was the score at the very beginning?
The Movie: What was the average score over several years (a "cumulative" score)?
The Plot Twist: Did the score go up, go down, or stay the same over time?

What They Found: The Danger of the Double Trouble

The results were clear and loud: The higher the Stress Score, the higher the risk of having a stroke. It wasn't a gentle curve; it was a steep climb.

The Snapshot: People with a higher starting score had a 26% higher risk of stroke for every standard step up in the score. If you were in the top 25% of scores, your risk was twice as high as someone in the bottom 25%.
The Movie (Cumulative Exposure): This was even more revealing. When the researchers looked at the average score over time, the risk gradient got steeper. People who had high scores consistently over the years had a 2.44 times higher risk of stroke compared to those with low scores. This suggests that long-term, chronic stress on your vessels is even more dangerous than a single bad day.
The Plot Twist (Trajectories): The researchers grouped people into four "storylines" based on how their scores changed between 2011 and 2015:
- Low-Low (LL): Staying low and healthy. (The safest group).
- High-Low (HL): Starting high but dropping down. (Risk dropped, but didn't disappear).
- Low-High (LH): Starting low but spiking up. (Risk went up).
- High-High (HH): Staying high the whole time. (The most dangerous group).

The "High-High" group had the highest risk, nearly double that of the "Low-Low" group. Interestingly, the "High-Low" group (those who improved) still had a higher risk than the "Low-Low" group, even after they lowered their scores. This hints at a "vascular memory"—like a scar on a road that remains even after the pothole is fixed. The damage from the earlier high stress didn't vanish immediately.

The Shape of the Danger

The researchers also checked if there was a "safe zone" where the score didn't matter until it got super high. They found no such safe zone. The relationship was a steady, upward climb: as the score went up, the risk went up. It didn't flatten out or dip down; it just kept getting steeper, especially at the higher levels.

Why This Matters

This study suggests that doctors and patients shouldn't just look at blood sugar and blood pressure as two separate numbers to manage. Instead, they should think of them as a team. If both are high, the combined "Stress Score" (HSI) is a powerful warning sign.

The study also highlights that time matters. A single check-up might miss the bigger picture. Someone who has had high scores for years is at greater risk than someone who just had a bad day. Furthermore, while improving your numbers is great (the "High-Low" group did better than the "High-High" group), the study suggests that the damage from years of high stress might linger, meaning prevention is better than cure.

The Bottom Line

The authors found that a higher combined score of blood sugar and blood pressure is strongly linked to a higher chance of stroke in a dose-dependent way (more score = more risk). They didn't prove why this happens biologically, but the data strongly suggests that the combined burden of these two factors is a major driver of brain health issues. The study confirms that looking at long-term trends and the combination of these factors gives a clearer, more urgent warning than looking at them in isolation.

Technical Summary: Longitudinal Hemoglobin A1c–Systolic Blood Pressure Index and Risk of Incident Stroke

Problem Statement
Stroke remains a leading cause of global disability and mortality, with a particularly high burden among aging populations in China. While hypertension and diabetes mellitus are established, frequently coexisting risk factors for cerebrovascular disease, traditional epidemiological models often evaluate them separately. This approach may fail to capture the integrated cardiometabolic burden resulting from the synergistic interaction of glycemic and hemodynamic stress. Furthermore, single-time-point measurements of risk factors (such as Hemoglobin A1c [HbA1c] and Systolic Blood Pressure [SBP]) do not account for cumulative vascular exposure or longitudinal transition patterns over time. The study addresses the gap in understanding whether a composite index, the Hemoglobin A1c–Systolic Blood Pressure Index (HSI), and its longitudinal trajectories provide incremental value for stroke risk prediction in middle-aged and older Chinese adults.

Methodology
The study utilized data from the China Health and Retirement Longitudinal Study (CHARLS), a nationally representative prospective cohort of adults aged ≥45 years.

Cohorts: Two analytical cohorts were established: a baseline cohort (n=9,342) and a cumulative exposure "landmark" cohort (n=6,071) restricted to participants stroke-free until 2015 to minimize immortal time bias.
Exposure Definition: The HSI was calculated as $(HbA1c (\%) \times SBP (mmHg)) / 100$ $(H b A 1 c (%) \times S B P (mm H g)) /100$ .
- Baseline HSI: Derived from 2011–2012 measurements.
- Cumulative HSI: Calculated as the time-weighted average of HSI values from 2011 and 2015.
- Transition Patterns: Participants were categorized into four groups based on wave-specific median cutoffs: Low-Low (LL), High-Low (HL), Low-High (LH), and High-High (HH).
Outcome: Incident stroke was identified via self-reported physician diagnosis during follow-up surveys (2013, 2015, 2018, 2020).
Statistical Analysis: Cox proportional hazards models (adjusted for demographics, lifestyle, and comorbidities) were used to estimate Hazard Ratios (HRs). Restricted cubic splines assessed nonlinearity. Fine-Gray competing risk models accounted for death as a competing event. Model performance was evaluated using Harrell's C-index, AIC, and likelihood ratio statistics. Sensitivity analyses included excluding early events and adjusting for medication use.

Key Results

Baseline Association: Higher baseline HSI was independently associated with increased stroke risk. Each 1-Standard Deviation (SD) increase in baseline HSI corresponded to a 26% higher risk (HR = 1.26; 95% CI: 1.19–1.33). A graded association was observed across quartiles, with the highest quartile showing a two-fold increased risk compared to the lowest (HR = 2.00).
Cumulative Exposure: Cumulative HSI demonstrated a similar but slightly steeper risk gradient. Each 1-SD increase was associated with a 24% higher risk (HR = 1.24; 95% CI: 1.15–1.34). The risk for the highest quartile (Q4) versus the lowest (Q1) was 2.44 (95% CI: 1.79–3.33), suggesting long-term integrated burden offers additional prognostic information.
Longitudinal Patterns: Compared to the stable Low-Low (LL) group, all other patterns showed elevated risk. The High-High (HH) group had the highest risk (HR = 1.99), followed by Low-High (LH, HR = 1.93) and High-Low (HL, HR = 1.54). Notably, the HL group (decreasing HSI) retained elevated risk compared to LL, though lower than HH.
Dose-Response: Restricted cubic spline analyses revealed nonlinear, monotonic dose-response relationships for both baseline and cumulative HSI, with no evidence of a threshold or J-shaped pattern.
Robustness: Findings remained consistent across competing risk models, subgroup analyses (age, sex, hypertension, diabetes), and sensitivity analyses. Adding HSI to clinical covariates improved model discrimination (C-index increased from 0.657 to 0.673).

Key Contributions

Validation of a Composite Index: The study confirms that the HSI, a simple product of HbA1c and SBP, is an independent predictor of incident stroke, capturing the synergistic burden of glycemic and hemodynamic stress better than isolated metrics.
Longitudinal Perspective: It demonstrates that cumulative exposure (time-weighted average) and longitudinal transition patterns provide risk stratification beyond single-time-point measurements.
Evidence of "Metabolic-Hemodynamic Memory": The persistence of elevated risk in the High-Low (HL) transition group (where HSI decreased over time) compared to the stable Low-Low group suggests a potential "vascular memory" effect, where prior exposure to combined hyperglycemia and hypertension induces lasting vascular vulnerability.
Clinical Utility: The study highlights that HSI is derived from routinely available clinical parameters, offering a cost-effective tool for dynamic, trajectory-based risk assessment in stroke prevention.

Significance and Claims
The authors claim that higher HSI is independently associated with incident stroke in a dose-dependent manner. They posit that long-term cumulative exposure and persistent elevation of HSI provide additional risk stratification beyond baseline measurements. The study advocates for a shift from static risk assessment toward longitudinal, exposure-based approaches for cerebrovascular risk stratification. The findings suggest that while reducing HSI confers meaningful benefit, prior exposure may leave residual vascular risk, underscoring the importance of early and sustained management of both glycemic and blood pressure control. The authors maintain modesty regarding causality, acknowledging the observational nature of the design and potential residual confounding, while noting the need for external validation in other populations.

Technical Summary: Longitudinal Hemoglobin A1c–Systolic Blood Pressure Index and Risk of Incident Stroke

More like this