Over the past 10 years, 17 studies have reported that antenatal exposure to the Chinese Famine across the first 3 years of life increased the subsequent risk of type 2 diabetes mellitus (T2DM) and/or hyperglycaemia. A 2018 review also concluded that the famine has contributed to China’s current T2DM epidemic1. However, in a Comment published in Nature Reviews Endocrinology (Li, C. et al. The effect of the Chinese Famine on type 2 diabetes mellitus epidemics. Nat. Rev. Endocrinol. 18, 313–314 (2019))2, which included discussion of a 2017 meta-analysis of published Chinese famine studies using age-matched controls3, Chihua Li and colleagues found that famine was not associated with an increased risk of T2DM.

We agree with Li and colleagues that both the age of assessment and severity of exposure contribute to the differences in findings. In the 2002 China National Nutrition and Health Survey (CNNHS), the mean ages for participants in the non-exposed, fetal-exposed, early-childhood-exposed, mid-childhood-exposed and late-childhood-exposed cohorts were 39, 42, 45, 47 and 49 years, respectively4. Similarly, in the Survey on Prevalence in East China of Metabolic Diseases and Risk Factors (SPECT-China), the mean ages for the non-exposed, fetal-exposed, childhood-exposed and adolescence-exposed or adult-exposed cohorts were 40–51, 52–55, 56–65 and 66–93 years, respectively5. Such differences in ages of assessment will undoubtedly have affected the likelihood of receiving a diagnosis.

Equally important, Chinese Famine exposure varied substantially, indicated by very different mortality across provinces6. Such differences in exposure will have effects on the association with T2DM4,7,8. In the severely affected famine areas in the CNNHS study, fetal-exposed participants had a clear increased risk of hyperglycaemia (OR 3.92; 95% CI 1.64–9.39), but no risk was apparent in less severely affected areas (OR 0.57; 95% CI 0.25–1.31)4. Similarly, in the SPECT study in Shandong and Anhui provinces, we again found that severe fetal exposure to famine resulted in an increased risk of T2DM (OR 1.59; 95% CI 1.11–2.30) but found no association in less severely affected areas8.

Moreover, in our 2017 report from the China Health and Retirement Longitudinal Survey (CHARLS), fetal exposure to famine also resulted in an increased risk of T2DM after participants were age-matched (combination of the non-exposed cohort 1962–1964 and pre-famine cohort 1956–1958) (OR 1.37; 95% CI 1.07–1.75)9. Given that CHARLS is a nationally representative sample of those >45 years of age, we were surprised that it was not included in the meta-analysis.

In this context, we re-analysed six studies (five that were also included in the Review by Zimmet et al. and one from the CHARLS) with age-matched controls1,8,9, taking into account severity of famine exposure. The summary estimate showed small but significant overall effects on the risk of T2DM (OR 1.25; 95% CI 1.15–1.37). However, we found a clearer association in severely affected areas (OR 1.38; 95% CI 1.11–1.72) and no association in less severely affected areas (OR 0.94; 95% CI 0.57–1.57) (Supplementary Tables).

Based on these findings, we conclude that the severity of antenatal exposure to the Chinese Famine determined heightened risk for later T2DM. Given that severe exposure was widespread in China (15 of 28 provinces had an over 50.0% higher mortality than the 3 years (1956–1958) immediately preceding the famine (1959–1961)6,9), it seems likely that the Chinese Famine has indeed substantially contributed to the current T2DM epidemic. We agree with Li and his colleagues that, regardless of earlier famine exposure, healthy lifestyles will reduce the burden of T2DM, but this does not diminish the significance of the Chinese Famine effects.

There is a reply to this letter by Li, C. et al. Nat. Rev. Endocrinol. https://doi.org/10.1038/s41574-019-0302-7 (2019).