We thank Zou et al. (Early-life exposure to the Chinese Famine and subsequent T2DM. Nat. Rev. Endocrinol. https://doi.org/10.1038/s41574-019-0299-y (2019)1 for their interest in our Comment (Li, C. et al. The effect of the Chinese Famine on type 2 diabetes mellitus epidemics. Nat. Rev. Endocrinol. 15, 313–314 (2019))2 and welcome the opportunity to address their concerns about the completeness of the data we presented.

Zou et al. re-analysed the five studies included in a Review by Zimmet et al.3 together with their own study that combined pre-famine and post-famine controls from the 2011 China Health and Retirement Longitudinal Study4. They claim a statistically significant overall effect on the risk of type 2 diabetes mellitus (T2DM) from an odds ratio of 1.25 (95% CI 1.15–1.37) for all studies combined using age-balanced controls and a random-effects model. Using the data as provided by Zou et al. in their supplementary table 1 (ref.1), we undertook the same analysis, including their added study, and still find a non-significant overall effect on the risk of T2DM with an odds ratio of 1.05 (95% CI 0.87–1.26). A summary of our calculations is attached (Supplementary Fig. 1).

For their re-analysis of studies in severely affected famine areas, Zou et al., in our view, should have been more careful in their choice of studies. The findings from the included Suihua cohort show an increase in T2DM in famine births compared with post-famine births5, but demographic and population census data suggest that the study area was in one of the provinces least affected by the famine6,7. This calls for a re-examination of their findings on severe famine exposure effects.

We do not dispute that Zou et al. could be correct in believing that the famine had a marked effect on current and future T2DM epidemics in China. This conclusion, however, cannot be drawn from current studies. For example, the Ukraine and Dutch famines had a statistically significant effect on later T2DM8,9, and the same could be true in China. Ideally this would be demonstrated in studies in which first, the timing and severity of the famine at the time of birth can be documented in well-defined study populations; and second, the populations can be followed over the life course for disease onset and mortality. The better that future studies are able to meet these goals, the better we will be able to determine the long-term impact of the Chinese Famine.