The need for appropriate 'age-balanced' controls and transparent reporting in Chinese famine studies: a re-analysis of the China Patient-centred Evaluative Assessment of Cardiac Events million persons project.

Abstract

We read with interest the study by Chen et al.1 on an increased risk of adulthood cardiovascular diseases (CVD) after prenatal Chinese famine exposure. We re-examined the data with alternative analytical strategies of using different age-balanced controls and found that overlooking the assumption of a linear relationship between age and CVD among non-famine controls can lead to biased results. We encourage more rigorous epidemiological research methods and greater transparency in disclosing data and findings in future studies. First, Chen et al.1 compared CVD among famine births of 1958–1961 to those among ‘age-balanced’ controls of pre- and post-famine births (1949–1958 and 1963–1972) and found an increased CVD risk (odds ratio, OR = 1.28, 95% CI 1.16–1.41). This estimate was made with adjustments for age, sex, BMI, and others. In contrast, we calculated the crude OR based on CVD cases and the population at risk in Figure 1, which is smaller (OR = 1.03, 95% CI 0.94–1.13). This discrepancy between crude and adjusted ORs should be disclosed and explored because it is unclear the adjustment for which covariate(s) led to a bigger effect estimate.2,3 It is also questionable to adjust for BMI when examining the famine-CVD association because BMI can be the mediator and/or effect modifier in this relationship.2