Preplanned Studies: A Systematic Analysis of Early Life Exposure to the Chinese Famine (1959–1961) and the Health of Older Adults — China, 2008–2023

The increasing aging population and the prevalence of chronic diseases necessitate a comprehensive life-course approach to address the growing demand for cost-effective prevention, treatment, and long-term care. To achieve healthy aging, it is crucial to understand the factors influencing disease development throughout an individual’s life and identify opportunities for effective interventions. This study examines the latest evidence on how early life circumstances impact health in old age, distinguishing between exposures in utero and during childhood in various environments. Furthermore, we conduct a meta-analysis utilizing studies on the effects of the Chinese Famine (1959–1961) on the health of older adults, discussing potential mechanisms underlying these associations.

Prenatal conditions have a significant impact on healthy aging, as evidenced by numerous studies (1–2). The Fetal Origin Hypothesis suggests that disruptions to the prenatal environment, including maternal and fetal health, social and economic shocks, and environmental pollution, can have long-term consequences on developmental health and well-being. These effects become apparent later in life, with an increased susceptibility to diseases persisting into old age.

Childhood represents a critical period during which adverse circumstances can have long-lasting effects on health (3). Figure 1 displays the variation in self-rated health among respondents from the China Health and Retirement Longitudinal Study (CHARLS) national sample over three waves, based on retrospective reports of health status before age 16. The data reveals a significant association between good health before age 16 and a higher likelihood of reporting excellent or very good health after the age of 60, with a difference of 5–10 percentage points. Additionally, various other childhood factors, including traumas and adversity, neighborhood safety and cohesion, education, friendship, parent-child relationship, parenting skills, natural environments, exposure to famine, parental health behaviors, disease infections, access to healthcare, parental socioeconomic status, and home and social environments, have also shown associations with health outcomes in later life.

Figure 1. 

Self-rated health among middle-aged and older adults by child health.

Note: In this figure, we included data from the longitudinal survey, ensuring that each respondent is counted only once. If a respondent participated in multiple waves of the survey, we used information from their earliest participation for plotting purposes. The respondents featured in this figure are middle-aged and older adults. Data: CHARLS National Sample (2011, 2013, 2015).

Abbreviation: CHARLS=China Health and Retirement Longitudinal Study.

Multiple risk factors in utero and early childhood have been identified as contributors to various adverse health outcomes, including but not limited to diminished height, disability, premature mortality, depressive symptoms, schizophrenia, cognitive impairment, metabolic syndrome, frailty, reduced lung function, arthritis, anemia, diabetes, coronary heart disease, stroke, and fatty liver disease.

Numerous studies in China have explored the association between early life risk factors and aging, with a particular focus on exposures to the Chinese Famine (1959–1961) (4). It is worth noting that a significant proportion of the Chinese population over the age of 60 today was exposed to famine during their early years. Both animal and human studies have indicated that the famine may have a substantial long-term impact on the prevalence of chronic diseases. The international literature has extensively documented the link between early-life undernutrition and detrimental health outcomes in later life (5).

We conducted a comprehensive search of publications, including journal articles, degree theses, and conference manuscripts, in various databases such as PubMed, Embase, Chinese Wanfang Data, and Chinese National Knowledge Infrastructure (CNKI) databases. The search was conducted until August 31, 2023. The keywords used for the search were [(China OR Chinese) AND (famine OR undernutrition OR starvation OR malnutrition)] OR great leap forward OR great famine. The English language was used for the search in PubMed and Embase, while the Chinese language was used for Wangfang and CNKI. For further details, please refer to the Supplementary Material.

The article selection process involved an independent review of titles and abstracts by two researchers. Data extraction included information on authors, publication details, analytical methods, study size, exposure definitions, control selections, conditions studied, and reported results. Excel spreadsheets were used to collect this data. Additionally, the researchers independently extracted data on the number of disease events and at-risk populations.

The quality of the literature was assessed using the Newcastle-Ottawa scale, which evaluates three main perspectives: study group selection, comparability of study groups, and identification of relevant exposures or outcomes in case-control or cohort studies. Each perspective was assigned a score of “good”, “fair”, or “poor” based on predefined criteria. Only literature of “good” quality in any perspective was included in the study, with specific criteria of achieving 3 or 4 points in the selection domain, 1 or 2 points in the comparability domain, and 2 or 3 points in the outcome/exposure domain.

The data analysis in this study utilized the “meta” package (version 6.5-0) in R (version 4.2.3; R Core Team, 2023) as the third-party software. Heterogeneity between studies was assessed using the I² statistic and analyzed using the χ² test. For I²<50%, a common-effects model (also known as the fixed-effect model) was employed for the meta-analysis. For I² values above or equal to 50%, a random-effects model was utilized. Statistical significance was defined as P<0.05.

As shown in Figure 2, data from 30 studies could be used for a meta-analysis of overweight and obesity, diabetes, hyperglycemia, metabolic syndrome, schizophrenia, depression, and arthritis. These studies provided data on famine births and post-famine births. Several reports findings for more than one health condition. The forest plot shows effect estimates for health conditions comparing famine births with post-famine births, which servies as controls.

Figure 2. 

Effect estimates on selected health conditions comparing famine births with controls (post-famine births, meta-analysis).

Note: Summary estimates for the Mantel-Haenszel fixed-effects model and DerSimonian-Laird random-effects model are provided. Each outcome is represented by a box and horizontal lines, indicating the odds ratio (OR) and 95% confidence intervals (CIs). The size of each box reflects the weight of the report for that particular outcome. Diamonds represent the 95% CI for pooled effect estimates, centered on the pooled odds ratio of either the fixed-effects model or the random-effects model for each outcome. The numbering of literature in this Table corresponds to the reference list in the Supplementary Material.

The meta-analysis findings indicate that there is a significant increase in the odds of overweight [odds ratio (OR) 1.14; 95% confidence interval (CI): 1.02, 1.27], hyperglycemia (OR 1.27; 95% CI: 1.13, 1.42), metabolic syndrome (OR 1.31; 95% CI: 1.18, 1.45), and schizophrenia (OR 1.52; 95% CI: 1.29, 1.78). However, no significant effects are observed for obesity (OR 1.18; 95% CI: 0.99, 1.40), diabetes (OR 1.32; 95% CI: 0.99, 1.75), hypertension (OR 1.07; 95% CI: 0.92, 1.25), depression (OR 1.65; 95% CI: 0.98, 2.78), or arthritis (OR 1.04; 95% CI: 0.86, 1.26). The estimates from both common-effects and random effects models were similar, with the random-effects model having wider 95% CIs due to between-study heterogeneity.