Findings from a recent study published in Journal of Global Health examined the relationship between physical activity (PA), diabetic retinopathy (DR), and retinal vessel diameters in patients with type 2 diabetes (T2D), using both cross-sectional and longitudinal data from a large Chinese cohort.
Give me some background first.
While DR is a common microvascular complication of diabetes, the evidence linking PA to DR has been inconsistent.
- In fact, cross-sectional studies have gone both ways: some find that regular PA is tied to lower DR prevalence, while others report no meaningful association.
And for context?
The American Diabetes Association recommends that adults with diabetes get at least 150 minutes of moderate-intensity exercise or 75 minutes of vigorous-intensity exercise per week, spread across at least three days.
Those guidelines are well-supported for reducing macrovascular risk. Whether they translate to microvascular benefits, particularly for the eyes, is less clear.
Now, talk about the study.
Investigators used data from the Shanghai Cohort Study of Diabetic Eye Disease (SCODE), which ran from 2017 to 2020.
The study had two components:
- A cross-sectional analysis of nearly 43,000 type 2 diabetes patients
- A 3-year longitudinal follow-up of a subset without DR at baseline.
The question: Does physical activity protect against DR incidence, and if so, which exercise components matter most?
Retinal vessel diameters were measured using computer vision and deep learning. PA data, including intensity, frequency, and duration, were collected through physician interviews.
Who was included in the study?
Cross-sectional arm: 42,992 T2D patients, mean age 64.4 years, 57.6% female, and a mean diabetes duration of 8.25 years.
- About 20.2% had DR at baseline, with 6.5% mild, 11.6% moderate, and 1.7% severe nonproliferative DR (NPDR). Just 0.3% had proliferative DR (PDR).
- Baseline characteristics: mean hemoglobin A1c of 7.24%, mean weekly exercise duration of 245 minutes.
To note: Only 16.5% of T2D patients met the study’s definition of “active” PA, and 23.5% engaged in moderate-to-high intensity exercise.
As for the longitudinal arm: 3,669 patients without DR at baseline, mean age 63.1 years.
- Patients were grouped as physically active or inactive and followed for 3 years.
Findings?
We’ll start with the longitudinal cohort, in which 833 patients (22.7%) developed DR by 2020.
PA was a protective factor against developing DR, with a hazard ratio (HR) of 0.812 (95% confidence interval [CI]: 0.679 to 0.971) in the fully adjusted model.
- By 2020: DR incidence within this cohort was 18.95% and 23.45% for active and inactive patients, respectively.
And in general?
Within this cohort, moderate-to-high intensity exercise was the only exercise component that significantly reduced DR risk after multivariate adjustment (HR = 0.823; 95% CI: 0.704 to 0.963).
- Exercise frequency of three or more times per week and total duration of 150 or more minutes per week did not reach statistical significance on their own.
Now to the cross-sectional cohort.
Among these T2DM patients: PA was not significantly associated with overall DR prevalence. However, after multivariate adjustment, it was protective against moderate and severe NPDR, specifically.
Noted. What else was discovered?
Subgroup analyses painted a more nuanced picture.
For patients already exercising three or more sessions per week, adding moderate-to-high intensity made a significant difference (HR = 0.812; 95% CI: 0.675 to 0.977).
For those below three sessions per week, simply increasing frequency helped (HR = 0.698; 95% CI: 0.526 to 0.926).
And what was the dose-response analysis?
Dose-response curves showed a U-shaped pattern:
- DR incidence dropped as exercise frequency increased, hitting its lowest point at roughly 4.67 to 5.09 sessions per week before ticking back up
- A similar curve appeared for weekly duration, with the sweet spot around 178.82 to 193.24 minutes per week.
Worth noting: Among patients doing moderate-to-high intensity exercise, bumping up to high intensity provided no additional benefit (P = 0.48 in the multivariate model).
How about on the vascular side?
PA significantly widened peripheral retinal arteries and increased the peripheral arterio-venous ratio (PAVR) over the 3-year follow-up.
- PAVR was independently associated with both DR prevalence and incidence, suggesting that changes in smaller peripheral vessels may serve as an early indicator of DR risk.
Further: A two-sample Mendelian randomization (MR) analysis supported a causal link: increased duration of moderate-intensity activity was associated with reduced DR risk (OR = 0.45; 95% CI: 0.27 to 0.77, P = 0.003).
- Vigorous activity showed no significant association.
Any limitations?
PA data were self-reported through physician interviews, introducing recall bias.
The study didn’t differentiate between specific exercise types, and only 2.6% of participants engaged in high-intensity exercise, making those findings harder to generalize.
Plus:
- The 3-year follow-up may not have been long enough to fully capture DR progression or vascular changes
- DR severity was only assessed in 2020, not annually, which limited the analysis of disease progression over time.
- The cohort consisted entirely of Chinese patients
See here for other considerations.
Expert opinion?
No outside expert commentary was included in the study.
That said: The authors argued that the metabolic benefits of exercise on blood glucose control are well-established and likely contribute to DR protection.
- But they also noted that PA showed a protective effect even without demonstrating significant metabolic improvements in this particular cohort, pointing to retinal vascular changes as a possible independent mechanism.
Their position: The peripheral retinal vessel changes observed in the active group suggest PA may improve microvascular health directly, not just through better glucose numbers.
Anything else?
One finding that stood out: Poorer blood pressure control was actually observed in the active group.
- The authors attributed this to the active group having a higher proportion of males, smokers, and drinkers, rather than exercise itself worsening blood pressure.
They also flagged that most patients in the cohort already met frequency and duration targets—and that the gap was in intensity, as only a minority exercised at moderate-to-high intensity.
Take home.
For clinicians managing T2D patients, intensity appears to matter more than frequency or duration alone when it comes to DR prevention.
As such: Encouraging patients to move beyond low-intensity exercise toward moderate-intensity activity, ideally around five sessions per week totaling at least 180 minutes, may offer meaningful retinal protection.