Dr. Axe Wire · Study decoded · health

Diet Quality Linked To Lower Diabetes Risk In 21,476 Adults

A large cohort study found people eating the highest-quality diets had lower odds of diabetes, partly tied to inflammation markers.

By Dr. Axe Editorial Team · October 15, 2026 · 1 source · Watch

A study published October 15, 2026 in the British Journal of Nutrition found that adults eating the highest-quality diets had 14% to 28% lower odds of diabetes depending on the scoring system used, with part of that link explained by inflammation markers measured in blood.

21,476adults studied

14-28%lower odds by score

9%mediated by inflammation

3diet quality scores used

Researchers followed 21,476 adults in the Fuqing Cohort Study who completed a food frequency questionnaire, then scored diet quality three separate ways: the Healthy Eating Index (HEI), Diet Quality Index (DQI) and Diet Balance Index (DBI). Comparing people in the highest diet quality group with those in the lowest, the adjusted odds of diabetes were 13.7% lower by HEI, 28.1% lower by DQI and 14.8% lower by DBI. The researchers also measured eleven inflammatory markers in blood, which are immune-system signaling proteins and cell counts that rise with inflammation. Higher HEI scores were linked to lower total white blood cell count, neutrophil count, monocyte count, and two inflammation ratios, and these markers accounted for 3.2% to 11.3% of the diet-diabetes link individually. When all eleven markers were analyzed together, they accounted for 8.83% of the total association.

The authors' mediation analysis indicates that poor diet quality may partly raise diabetes risk by driving up inflammatory activity in the blood, rather than through blood sugar effects alone. Diet composition affects circulating immune cells and inflammatory proteins, and chronic low-grade inflammation is an established contributor to insulin resistance, so this fits a plausible biological pathway connecting food choices to metabolic disease.

Lower Diabetes Odds By Diet Quality Score (Highest vs Lowest Group)

21,476 adults, Fuqing Cohort Study

DQI

28.1 % lower odds

DBI

14.8 % lower odds

HEI

13.7 % lower odds

British Journal of Nutrition, 2026, Liu et al.

This is an observational study with diet assessed once at the start by self-report, so recall error is likely, and the inflammation markers only explained a modest share of the overall link, meaning most of the diet-diabetes connection runs through pathways not measured here. Funding and conflicts not reported in the abstract.

Dr. Axe’s Take

This fits what I have believed for years: inflammation sits at the center of metabolic disease, and diet quality shapes that inflammation long before blood sugar numbers go wrong. The fact that white blood cell counts and inflammatory ratios only explained part of the link tells me there are other pathways at work too, likely involving gut health and insulin signaling. This week, I would build meals around vegetables, wild-caught fish, olive oil and fiber-rich plants, and cut back processed grains and added sugar, since those are the foods every diet quality index rewards. Pair that with daily movement to help calm inflammation further.

— Dr. Axe

Take the Metabolic Health Quiz →

Study summary

Who: 21,476 adults in the Fuqing Cohort Study who completed a food frequency questionnaire
Design: Observational cohort study using logistic regression and mediation analysis
Dose: Not applicable; dietary quality scored via HEI, DQI and DBI from FFQ data
Primary result: Highest vs lowest diet quality: 13.7% lower odds (HEI), 28.1% lower odds (DQI), 14.8% lower odds (DBI) of diabetes
Secondary: Inflammatory markers individually mediated 3.2% to 11.3% of the HEI-diabetes link; all 11 markers combined mediated 8.83% of the total association
Funding / conflicts: Funding and conflicts not reported in the abstract.
Date: Oct 15, 2026