Author links open overlay panelShigeSongabcPersonEnvelope
Show moreNavigate DownShareShare
Cited ByCite
//doi.org/10.1016/j.socscimed.2010.04.034Get rights and content
Abstract
Using retrospective mortality records for three cohorts of newborns [1956–1958, 1959–1961, and 1962–1964] drawn from a large Chinese national fertility survey conducted in 1988, this article examines cohort mortality differences up to age 22, with the aim of identifying debilitating and selection effects of the 1959–1961 Great Leap Forward Famine. The results showed that the mortality level of the non-famine cohort caught up to and exceeded the level of the famine cohort between ages 11 and 12, suggesting both debilitating and selection effects. Multilevel multiprocess models further established a more direct connection between frailties in infancy and frailties at subsequent ages, revealing the underlying dynamics of mortality convergence between the famine and the non-famine cohorts caused by differential excess infant mortality. These results provide important new insights into the human mortality process.
Introduction
Famine is a catastrophic event, and it has long been known to be associated with increased mortality. Some investigators claim that, in addition to causing an immediate rise in mortality, famine may also have a “long-term” or “sustained” mortality consequence. In a study of the 1974–1975 Bangladesh famine, Razzaque, Alam, Wai, and Foster [1990] showed that the famine-born cohorts continued to have a higher than usual mortality in the second year of life, one year after the famine was over, then followed by a lower than usual mortality between the ages 2 and 5. Kannisto, Christensen, and Vaupel [1997], in their study of the 1866–1869 Finnish famine, reported that the famine cohort had an higher than usual mortality up to age 17 but no cohort difference afterwards.
In both cases, the famine-born cohorts continued to show a higher mortality than the non-famine cohorts after the famine was over; then the cohort difference in mortality either disappeared [in the Finnish case] or even got reversed [in the Bangladeshi case]. Such a pattern, known to demographers as “mortality crossover” [Nam, 1995], was considered as evidence for both “debilitation effect” – a lingering negative impact of poor nutrition and harsh environment on individual health, and “selection effect” – a result of selective frailty processes on cohort-level mortality difference. Debilitation and selection effects may occur in non-famine situations as well. For example, Caselli and Capocaccia [1989] reported a mortality pattern in Italy that was similar to the Bangladeshi case: cohorts that experienced unfavorable conditions early in life tended to show increased cohort mortality up to a certain age, followed by reduced cohort mortality afterwards. The main difference between the Bangladeshi and Italian studies was the timing of the turning point – the crossover point in Bangladesh occurred at age two, while in Italy it occurred at age 45.
Treating mortality crossover as suggestive evidence for debilitation and selection effects represents an important step toward a better understanding of long-term health consequences of famine. Doing so, however, raises several issues that need further discussion. First of all, the above reasoning implicitly assumes that debilitation and selection effects always work against each other – that debilitation increases mortality of the famine cohort while selection decreases it – in determining the observed cohort mortality trend, but not all researchers accept that assumption without reservation [Elo & Preston, 1992]. Second, the nature and dynamics of selection effects, compared to that of debilitation effects, is much less well understood. For example, while debilitation effect has been suggested to be strong at the beginning of the post-famine era and then gets weaker over time, there has been little discussion regarding whether selection effect follows similar or different trends. Without such knowledge, the claim that mortality crossover indicates changes in relative strength of debilitation and selection effects in the post-famine period becomes much less convincing. This leads to the next point. If, for some reasons, no mortality crossover is observed after a famine, what can we say about debilitation and selection?
In this research, I examined long-term mortality consequences of the 1959–1961 Great Leap Forward Famine in China. I first reviewed conceptual and methodological issues in identifying debilitation and selection effects. Then I proposed a statistical approach to isolate selection effect, borrowing strength from the recent developments in multilevel multiprocess latent variable modeling methodology [Lillard, 1993, Panis and Lillard, 1995, Steele et al., 2005]. Using data from a large representative national sample survey conducted in 1988 in China, I [1] identified mortality crossover between the famine and the non-famine cohort; [2] demonstrated that famine may “reduce” mortality level of the famine cohort in the post-famine period by eliminating its frail members first; and [3] showed that both debilitation and selection effects declined with time, and debilitation effect declined faster than selection effect.
Section snippets
Conceptualizing debilitation and selection
In the context of famine and other natural disasters, the “debilitation” or “scaring” effects refer to the possibility that certain negative conditions [e.g. diseases, malnutrition at birth, and growth retardation] experienced early in life may permanently impair the health of survivors and thus leave an imprint on their mortality risks at all subsequent ages [Almond, 2006, Elo and Preston, 1992]. Selection effects, on the other hand, refer to the possibility that famine survivors [the
Empirical context: the “great leap forward” famine in China
The Great Leap Forward Famine happened on the 10th year of the People’s Republic of China [PRC], which was unexpected at the time because the country was rapidly recovering from century-long warfare, political and social turmoil, and natural disaster and epidemics. Fig. 1 shows the cohort trend for mortality in China [up to age 22], calculated from the 1988 National Survey of Fertility and Contraception. An overall trend of declining mortality from 1948 to 1970 was clear: cohort mortality
Research design
I compared mortality pattern between the pre-famine cohort [1956–1958], the famine cohort [1959–1961], and the post-famine cohort [1962–1964] using data from a large nationally representative sample survey conducted in China in 1988. The comparison between the 1959–1961 and the 1962–1964 cohorts is theoretically more important than the comparison between the 1959–1961 and the 1956–1958 cohorts and thus will be the focus of this research. Since the famine occurred in the middle of a secular
Describing the overall trend of cohort mortality: is there a mortality crossover between ages 0 and 22?
Infant mortality plays an important role in shaping the overall cohort mortality pattern. This is most clearly demonstrated in Fig. 2, which shows the non-parametric Kaplan–Meier survivor function for the three cohorts. For the pre-famine cohort, survival probability declined from 1 at age 0 to 0.80 at age 22; 57% of the change occurred between ages 0 and 1. For the famine cohort, survival probability declined from 1 at age 0 to 83% at age 22; 64% of change occurred between ages 0 and 1; for
Discussion
This study contributes an important case to a small but rapidly growing literature on the potential long-term health consequences of early exposure to adverse environment. It is one of the first studies that focus on the 1959–1961 Great Leap Forward Famine in China, a catastrophic social and humanitarian event in modern human history that caused a tragic loss of over 30 million human lives in a short period of time. Using retrospective mortality record of 207,046 births from 96,203 women, this
References [52]
- S. Song et al.
Famine, death, and madness: schizophrenia in early adulthood after prenatal exposure to the Chinese great leap forward famine
Social Science & Medicine
[2009]
- T. Schultz
Human capital, schooling and health
Economics and Human Biology
[2003]
- C. Riskin
Seven questions about the Chinese famine of 1959–1961
China Economic Review
[1998]
- S.H. Preston et al.
Childhood conditions that predict survival to advanced ages among African–Americans
Social Science & Medicine
[1998]
- J.Y. Lin et al.
On the causes of China’s agricultural crisis and the great leap famine
China Economic Review
[1998]
- L. Lillard
Simultaneous-equations for hazards – marriage duration and fertility timing
Journal of Econometrics
[1993]
- L. Lee et al.
The effects of improved nutrition, sanitation, and water quality on child health in high-mortality populations
Journal of Econometrics
[1997]
- Y. Chen et al.
The long-term health and economic consequences of the 1959–1961 famine in China
Journal of Health Economics
[2007]
- A. Agresti
Categorical data analysis
[2002]
- D. Almond
Is the 1918 influenza pandemic over? Long-term effects of in utero influenza exposure in the post-1940 US population
Journal of Political Economy
[2006]
Famine in China, 1958–1961
Population and Development Review
[1984]
Birth-spacing, fertility and neonatal mortality in India: dynamics, frailty, and fecundity
Journal of Econometrics
[2007]
Demographic responses to famine
Famine, social disruption, and involuntary fetal loss: evidence from chinese survey data
Demography
[2005]
Age, period, cohort and early mortality: an analysis of adult mortality in Italy
Population Studies
[1989]
An introduction to survival analysis using stata
[2004]
Effects of early-life conditions on adult mortality: a review
Population Index
[1992]
Proportional hazards tests and diagnostics based on weighted residuals
Biometrika
[1994]
Parametric frailty and shared frailty survival models
Stata Journal
[2002]
Sample selection bias as a specification error
Econometrica: Journal of the Econometric Society
[1979]
No increased mortality in later life for cohorts born during famine
American Journal of Epidemiology
[1997]
Birth order and health status in a British national sample
Journal of Biosocial Science
[1992]
The causes of China’s great leap famine, 1959–1961
Economic Development and Cultural Change
[2003]
One quarter of humanity: Malthusian mythology and Chinese realities, 1700–2000
[1999]
Premarital cohabitation and subsequent marital dissolution: a matter of self-selection?
Demography
[1995]
2022, Social Science and Medicine Show abstractNavigate Down Very limited studies focused on the early-life adversities on infectious diseases. Taking the Great Chinese famine as a natural experiment, this study re-evaluated the long-term effect of prenatal famine exposure on infectious diseases by using nationally representative data. Using difference-in-difference [DID] models, we analyzed 215,216 adults who participated in the Second National Sample Survey on Disability in 2006 across 734 counties of China to detect the effects of prenatal exposure to the Great Chinese Famine on the risk of infectious diseases in adulthood. Infectious diseases were ascertained by using the combination of self-reports or family members’ reports and on-site medical diagnosis by experienced specialists, and the severity of famine was measured by the cohort size shrinkage index [CSSI] at the county level. All DID estimates of the effects of famine on the probability of infectious diseases were insignificant, with a coefficient of 0.0007 [−0.0024, 0.0026] for all participants and coefficients of 0.0001 [−0.0041, 0.0043] and −0.0002 [−0.0036, 0.0033] for males and females, respectively. That is, the famine cohort dwelling in regions with a greater intensity of famine had similar levels of infectious disease risks than the cohorts with post-famine prenatal exposure experience in less affected famine regions. Furthermore, there were no significant famine and post-famine cohort differences in the DID estimates by examining the variations in subgroups with different types of infectious diseases [trachoma, poliomyelitis, tuberculosis, maternal infections and other infectious diseases]. No significant impact of prenatal exposure to the Chinese famine was observed on the risk of infectious diseases in adulthood. Famine survivors may be “cured” by the famine and were resilient to adverse environments in their life course because selective mortality may weaken the association between adverse prenatal exposure and later health. 2021, Genomics Show abstractNavigate Down Extensive epidemiological studies have established the association between exposure to early-life adversity and health status and diseases in adults. Epigenetic regulation is considered as a key mediator for this phenomenon but analysis on humans is sparse. The Great Chinese Famine lasting from 1958 to 1961 is a natural string of disasters offering a precious opportunity for elucidating the underlying epigenetic mechanism of the long-term effect of early adversity. Using a high-throughput array platform for DNA methylome profiling, we conducted a case-control epigenome-wide association study on early-life exposure to Chinese famine in 79 adults born during 1959–1961 and compared to 105 unexposed subjects born 1963–1964. The single CpG site analysis of whole epigenome revealed a predominant pattern of decreased DNA methylation levels associated with fetal exposure to famine. Four CpG sites were detected with p Cited by [37]
Is it just a cure? Re-evaluating the effects of prenatal exposure to the Great Chinese Famine on the risk of infectious diseases in adulthood
Differential regulation of the DNA methylome in adults born during the Great Chinese Famine in 1959–1961