Cardiovascular

Association between nutritional‑inflammatory status and diabetic retinopathy in adults with diabetes: insights from a US population‑based survey and a Chinese hospital‑based clinical cohort.

TL;DR

This dual-cohort study provides robust and cross-ethnic evidence that PNI and GNRI are inversely linked to diabetic retinopathy, whereas ALBI and NPAR show positive associations, with hospital-based data further linking these indices to DR severity.

Key Findings

Higher prognostic nutritional index (PNI) was associated with significantly reduced diabetic retinopathy risk in the NHANES cohort.

  • The highest versus lowest tertile of PNI was associated with an OR of 0.57 (95% CI: 0.41–0.80) after multivariable adjustment.
  • A significant dose-response trend was observed across PNI tertiles.
  • RCS analysis confirmed a nonlinear association between PNI and DR.
  • Threshold effect analysis identified a cohort-specific inflection point at PNI = 56.00 in NHANES and PNI = 42.85 in the clinical cohort.
  • The protective association was stronger among overweight/obese individuals in subgroup analyses.

Higher geriatric nutritional risk index (GNRI) was associated with significantly reduced diabetic retinopathy risk in the NHANES cohort.

  • The highest versus lowest tertile of GNRI was associated with an OR of 0.56 (95% CI: 0.41–0.76) after multivariable adjustment.
  • A significant dose-response trend was observed across GNRI tertiles.
  • RCS analysis confirmed a nonlinear association between GNRI and DR.
  • Threshold effect analysis identified inflection points at GNRI = 98.27 in NHANES and GNRI = 90.39 in the clinical cohort.
  • Analysis was conducted using weighted multivariable logistic regression in 3,715 adults with diabetes from NHANES (2005–2018).

Higher albumin-bilirubin (ALBI) score was associated with elevated diabetic retinopathy risk in the NHANES cohort.

  • The highest versus lowest tertile of ALBI was associated with an OR of 1.58 (95% CI: 1.18–2.19) after multivariable adjustment.
  • A significant dose-response trend was observed across ALBI tertiles.
  • Subgroup analyses showed stronger risk associations for ALBI in women.
  • These associations were corroborated with even stronger effect sizes in the hospital-based clinical cohort of 252 type 2 diabetes patients.

Higher neutrophil percentage-to-albumin ratio (NPAR) was associated with elevated diabetic retinopathy risk in the NHANES cohort.

  • The highest versus lowest tertile of NPAR was associated with an OR of 1.37 (95% CI: 1.02–1.83) after multivariable adjustment.
  • A significant dose-response trend was observed across NPAR tertiles.
  • Associations were corroborated with stronger effect sizes in the hospital-based clinical cohort.
  • Analyses used weighted multivariable logistic regression, restricted cubic splines, and subgroup analyses.

PNI and GNRI declined progressively with increasing diabetic retinopathy severity in the hospital-based clinical cohort, while ALBI and NPAR rose incrementally.

  • The hospital-based clinical cohort comprised 252 type 2 diabetes patients from a tertiary eye center in China.
  • All four indices demonstrated graded associations with DR severity, not just DR presence versus absence.
  • Effect sizes for the association between nutritional-inflammatory indices and DR were stronger in the hospital-based cohort than in NHANES.
  • These findings extended results beyond binary DR classification to a severity spectrum.

Nonlinear associations between PNI/GNRI and diabetic retinopathy were confirmed by restricted cubic spline analyses with cohort-specific inflection points identified by threshold effect analysis.

  • RCS confirmed nonlinear relationships for both PNI and GNRI with DR.
  • NHANES inflection points: PNI = 56.00 and GNRI = 98.27.
  • Clinical cohort inflection points: PNI = 42.85 and GNRI = 90.39.
  • Three sensitivity analyses were conducted to confirm the robustness of findings.
  • The dual-cohort design spanning US (NHANES) and Chinese hospital populations provided cross-ethnic evidence.

The study analyzed data from two distinct cohorts representing different ethnic populations and clinical settings to examine nutritional-inflammatory indices and diabetic retinopathy.

  • The NHANES cohort included 3,715 adults with diabetes from the US population survey conducted between 2005 and 2018.
  • The hospital-based clinical cohort included 252 type 2 diabetes patients from a tertiary eye center in China.
  • Indices examined included PNI, GNRI, ALBI score, and NPAR.
  • Statistical methods included weighted multivariable logistic regression, restricted cubic splines, subgroup analyses, threshold effect analysis, and three sensitivity analyses.

What This Means

This research suggests that simple, routinely available blood test markers that reflect both nutritional status and inflammation levels are meaningfully linked to the risk of diabetic retinopathy (DR), the leading cause of vision loss in working-age adults with diabetes. The study examined four composite indices — the Prognostic Nutritional Index (PNI), Geriatric Nutritional Risk Index (GNRI), Albumin-Bilirubin (ALBI) score, and Neutrophil Percentage-to-Albumin Ratio (NPAR) — in two separate groups: over 3,700 people with diabetes from a large US national health survey, and 252 type 2 diabetes patients from a specialized eye clinic in China. Higher PNI and GNRI scores, which reflect better nutritional status, were associated with a roughly 40–44% lower likelihood of having DR, while higher ALBI and NPAR scores, which reflect worse liver function and more inflammation, were associated with 37–58% higher likelihood of DR. Importantly, in the Chinese clinical cohort, these indices also tracked with how severe the DR was, not just whether it was present. This research suggests that the development of diabetic retinopathy is not driven solely by blood sugar levels but also by a person's nutritional and inflammatory status. The finding that these associations held across two very different populations — American adults from a diverse national survey and Chinese patients at a specialist eye clinic — strengthens the argument that these markers may be broadly relevant regardless of ethnicity or clinical setting. The researchers also found some variation by subgroup: the protective effect of better nutritional status (as measured by PNI) appeared stronger in people who were overweight or obese, and the risk associated with worse liver-related scores (ALBI) was more pronounced in women. This research suggests that these indices, which are calculated from standard blood tests that are already routinely collected in clinical care (such as albumin, bilirubin, and white blood cell counts), could serve as practical, low-cost tools to help identify people with diabetes who are at higher risk for developing or worsening diabetic retinopathy. This could potentially support earlier screening and more targeted monitoring, though the study's observational design means it cannot establish that improving these indices would directly prevent or slow retinopathy.

Have a question about this study?

Citation

Deng J, Pan L, Zhang Q, Wang M, Luo S, Li Y. (2026). Association between nutritional‑inflammatory status and diabetic retinopathy in adults with diabetes: insights from a US population‑based survey and a Chinese hospital‑based clinical cohort.. Frontiers in endocrinology. https://doi.org/10.3389/fendo.2026.1883581