Tea and prediabetes: Dark linked to less progression, green linked to less regression. Three-year follow-up in China

An analysis of a large Chinese cohort with prediabetes (n = 2662, ~3-year follow-up) was published in Nutrients. It turned out that daily consumption of dark tea (in the Chinese classification - fermented "dark tea") is associated with a lower chance of progression from prediabetes to type 2 diabetes (OR 0.28; 95% CI 0.11-0.72), while daily consumption of green tea is associated with a lower chance of regression to normoglycemia (OR 0.72; 0.56-0.92). In parallel, dark tea drinkers had lower insulin resistance (TyG index -0.23), while green tea drinkers had higher insulin resistance (TyG +0.05). The associations remained after adjustments for age, gender, BMI, diet, activity, and other factors. The authors carefully emphasize: this is an observational study, and verification in an RCT is needed.

Background

Prediabetes is a "pre-entry" into type 2 diabetes, which hundreds of millions of people already live with today. According to Diabetes Care, in 2021, impaired glucose tolerance (IGT) occurred in 9.1% of adults (≈464 million), and by 2045 the forecast is 10% (≈638 million). Against this background, the World Health Organization records a doubling of the prevalence of diabetes among adults from 1990 to 2022 - from 7% to 14%, with rates and the gap in access to treatment growing especially rapidly in low- and middle-income countries.

The key "brake" on the progression of prediabetes is lifestyle: structured programs with 150-175 minutes of activity per week and a 5-7% weight loss can reduce the risk of developing diabetes by 40-70%; these approaches are also enshrined in current clinical guidelines for diabetes prevention. Against this background, it is logical to look for supporting dietary habits that work "in addition" to the basic scheme.

Tea is one of the most widely consumed beverages in the world, but its “family” is heterogeneous: green tea is almost unfermented, black tea is fully oxidized, and dark tea (Chinese category dark tea - for example, pu-erh) undergoes microbial post-fermentation, due to which the polyphenol composition and biological effects differ significantly. Dark teas produce theabraunins - large pigmented polyphenols, which in preclinical studies are associated with modulation of microbiota, lipid metabolism and improvement of insulin resistance; recent reviews call theabraunin the “bioactive core” of dark tea.

Epidemiological data on tea and diabetes risk are generally encouraging but mixed: meta-analyses of prospective cohorts have shown a reduced risk of T2D with high consumption (≈ ≥3-4 cups/day), although results vary by population and tea type; intervention reviews have more often documented improvements in glycemia with green tea in those already with T2D. These discrepancies encourage more nuanced studies that take into account tea type, dose/frequency, and context (healthy, prediabetes, T2D).

To assess insulin resistance in large population projects, the TyG index is increasingly used - a simple surrogate based on fasting triglycerides and glucose: TyG = ln(TG [mg/dL] × glucose [mg/dL] / 2). It is reproducible, cheap, and correlates well with "classical" markers of insulin resistance, so it is suitable as an outcome and as a mediator in studies on nutrition and metabolism.

Against this background, the Chinese prospective work from Nutrients is interesting for two reasons at once: it separates the tea types (including dark, distinguished by post-fermentation) and looks at clinically significant trajectories of prediabetes - regression to normoglycemia and progression to diabetes - in parallel with the dynamics of TyG as a proxy for insulin resistance. This helps to move away from "average" conclusions about "tea in general" and get closer to what is important to a patient with prediabetes "here and now".

What was done and who was studied

• We recruited 2662 adults with prediabetes from the national SENSIBLE cohort and followed them for approximately 3 years.
• At the start, we found out the type of tea (green, black, dark, other/no tea) and frequency of consumption (daily/sometimes/I don’t drink).
• Two outcomes were assessed: regression to normoglycemia or progression to diabetes, according to ADA criteria; analysis was multinomial logistic regression with a wide range of covariates.

Main results (with figures)

• Dark Tea and the Progression of Prediabetes → Diabetes
  • Compared with non-tea drinkers,daily dark tea drinkers had a 78% lower odds of developing diabetes (OR 0.22; 0.07-0.71); in the overall "any dark tea" model - OR 0.28; 0.11-0.72.
  • The signal was particularly strong in women (OR 0.27; 0.08–0.90) and in non-Han ethnic subgroups (OR 0.18; 0.04–0.80).
• Green tea and regression to normal
  • Daily green tea consumption was associated with lower odds of returning to normoglycemia (OR 0.72; 0.56-0.92); a similar trend was observed for “sometimes” (OR 0.74-0.76).
  • The effect is more pronounced in women and in the Han subgroup.
• Insulin resistance (TyG)
  • Dark tea ↘ TyG −0.23 (p < 0.001);
  • Green tea ↗ TyG +0.05 (p ≈ 0.05);
  • Black tea is neutral. The relationships are stable after corrections.

How to interpret this

• Different teas – different “biology”. The authors remind us that the type of tea matters because processing technologies and microbial fermentation change the polyphenol composition and biological effects. The results indicate that dark tea in people with prediabetes is correlated with lower insulin resistance and a slower “slide” into diabetes, while green tea in this cohort is less likely to “roll back” to normal. This does not cancel out numerous reports of the benefits of green tea in other populations: population differences and the context of prediabetes are likely at work, so interventional and mechanistic studies are needed in this group.

What does this mean for a reader with prediabetes?

• If you enjoy tea daily, dark fermented varieties (known as "dark tea" in Chinese) as part of an overall healthy diet may be a neutral or potentially beneficial option in terms of progression risk.
• Green tea in this study did not speed up progress, but was associated with less regression and slightly higher TyG; this is a reason not to expect an “anti-diabetic” effect by default in people with prediabetes and to focus primarily on the basic measures: weight loss, exercise, sleep, sugar and blood pressure control.
• No tea replaces treatment. The study is observational: it shows associations, not causation; tea selection is an adjunct to a basic plan with a doctor.

Limitations to keep in mind

• This is an observational cohort study, not a randomized trial; residual confounding (eg, dietary factors in tea drinkers) is always possible.
• Tea was assessed at the start and habit dynamics were not monitored; doses/cup strengths were not standardized.
• Subgroups (gender, ethnicity) are sometimes small in number, and estimates are not always precise. However, key relationships are statistically stable across models.

What's next in science

• RCT with dark tea in people with prediabetes: standardize the variety and dose, look at the dynamics of glucose, insulin resistance (HOMA-IR, TyG), glycated HbA1c and the frequency of transition to diabetes.
• Cup-by-cup comparison of green and dark tea in prediabetes, considering microbiota and polyphenol metabolites to explain differential effects.
• Personalization: Check why women and certain ethnic groups respond more strongly - hormonal background, dietary pattern, genetics?

Source: Li T. et al. The Impact of Tea Consumption on Prediabetes Regression and Progression: A Prospective Cohort Study. Nutrients 17(14):2366, 2025. Online: Key numbers, subgroups, and TyG analysis are provided in the text and tables of the article. https://doi.org/10.3390/nu17142366