Background The association between famine exposure in early life and risk of arthritis (combination of osteoarthritis and inflammatory arthritis) in adulthood is unclear. The aim of this study is to explore the association.
Methods A total of 4124 subjects were selected from the national data of the China Health and Retirement Longitudinal Study in 2011–2012. Doctor-diagnosed arthritis was self-reported in participants’ questionnaire. Birthdates were used to categorise participants into famine-exposed and non-exposed groups. Logistic regression model was used to explore the association of famine exposure in early life with the risk of arthritis in adulthood.
Results The prevalence of arthritis in both infant-exposed and preschool-exposed groups was significantly higher than those in the non-exposed group (35.0% and 30.6% vs 27.3%; p<0.05). Compared with the non-exposed group, the infant-exposed group showed a significantly elevated risk of arthritis in adulthood after adjusting for confounding factors (OR=1.65; 95% CI 1.29 to 2.11; p<0.001). In the stratified analysis, we found that participants who lived in severely affected areas (OR=1.91; 95% CI 1.41 to 2.59; p<0.001), who are female (OR=2.21; 95% CI 1.57 to 3.11; p<0.001) and those with a body mass index ≥24.0 kg/m2 (OR=2.46; 95% CI 1.70 to 3.55; p<0.001) in the infant-exposed group had increased risk of arthritis in adulthood. Similar results were additionally observed when age-balanced control group was used.
Conclusion Great China Famine exposure in infancy may be associated with an elevated risk of arthritis in adulthood, particularly in women and participants with adiposity. These findings suggest nutrition intervention in infancy and weight control in later life may reduce the risk of arthritis in adulthood.
- child health
- epidemiology of chronic non communicable diseases
Statistics from Altmetric.com
Arthritis, including osteoarthritis and inflammatory arthritis, is a common cause of disability, contributing greatly to impaired quality of life and medical care costs.1–5 In the USA, 23% of adults (more than 54 million) have arthritis, and the prevalence is predicted to reach 26% (over 78 million) in 2040.6 The economic burden of arthritis is high. For example, the medical care costs and lost earnings reached as high as $128 billion in the USA in 2003.7
With the rapidly ageing population in low-income/middle-income countries, such as China, the prevalence of arthritis is expected to grow constantly.8 9 According to previous studies, the prevalence of arthritis ranged from 22.0% to 31.4% among middle-aged and older Chinese adults.10 11 Moreover, Qing and colleagues reported the prevalence of symptomatic osteoarthritis and rheumatoid arthritis ranged from 5.1% to 20.8% and from 0.2% to 0.93%, respectively, varying by area, design and age of participants surveyed.12 Several factors, including age, sex, occupation, smoking, obesity, physical activity (PA), hypertension and depression, were associated with arthritis.13 14 Apart from these traditional factors, animal studies additionally suggested that severe malnutrition in early life is associated with arthritis in later life.15 This finding is also supported by the theory of fetal origins of adult disease, which suggests that suffering from undernutrition in early life may change the structure and function of organs, modify the levels of DNA methylation of genes related to metabolic processes, regulating their transcription, and increase the risk of diseases in later life.15 16 Several studies had explored the effect of low birth weight on later risk of osteoarthritis.17 18 However, due to medical ethics, it is inappropriate to examine the association between famine exposure in early life and arthritis in humans. Nevertheless, historical famines such as the Dutch Famine and the Great China Famine (1959–1961) provide us a unique opportunity to investigate this association.
Different from the Dutch Famine, the Great China Famine lasted longer (approximately 3 years), affected more population (more than 662 million) and caused more severe consequences (30 million premature deaths).19 20 Except for these short-term consequences, the famine also resulted in long-term health consequences. For example, studies have found that individuals whose early life was exposed to the Great China Famine have significantly elevated risk of hypertension,21 diabetes mellitus,22 dyslipidaemia,23 metabolic syndrome24 and schizophrenia.25
To our knowledge, only one study assessed the association and found that famine exposure in childhood increased the risk of arthritis in adulthood.26 However, the study had several limitations.27 For example, the difference in age between exposed and non-exposed groups was not considered, and the childhood-exposed group was 6 years older than the non-exposed group. Because age is positively associated with arthritis prevalence,28 it may be the age, rather than the famine exposure, that led to their findings. Therefore, the association between famine exposure in early life and the risk of arthritis later in life requires further clarification.
In view of these facts, the current study used the national baseline data of the China Health and Retirement Longitudinal Study (CHARLS) to explore the association of famine exposure in early life with the risk of arthritis in later life. These results have a potential to aid the identification of a population group with a high risk of arthritis and the improvement of relative interventions.
Materials and methods
All participants in this study were selected from the national baseline survey of CHARLS. CHARLS is a large national epidemiological survey that aimed to describe the dynamics of retirement and health and its interactions with health and well-being in the Chinese population aged more than 45 years.29 The baseline data of CHARLS were collected through face-to-face household interviews by trained interviewers in 28 provinces of mainland China from June 2011 to March 2012. Information on socioeconomic status, demographics and common adulthood diseases (such as arthritis, hypertension and diabetes mellitus) was collected in the interview. The survey used a stratified four-stage cluster probability sampling method, and participants aged ≥45 years old were enrolled from 10 257 households, 450 villages/neighbourhoods, 150 counties, in 28 provinces.29 Briefly, first, 150 county units were selected randomly by probability sampling method from a sampling frame including all county units and excluding Tibet, stratified by region and county-level gross domestic product. Second, three administrative village level units by probability sampling method were selected randomly from each county units as primary sampling units (PSUs). Third, 80 dwellings were randomly selected from a dwelling unit frame, and 24 age-eligible households were selected from each PSU. Fourth, one member aged ≥45 years old and their spouse were included in the CHARLS baseline survey. In the current study, 4138 participants were selected according to their birthdates. After excluding 14 participants without information on arthritis, a total of 4124 participants were included in the final analysis. All participants in the present study provided written informed consent.
Classification of famine exposure groups
The participants were categorised into four groups based on their birthdates. Because the Great China Famine began in 1959 and ended in 1961, participants (n=439) who were born from 1 January 1959 to 30 September 1959 were exposed to famine both in fetal and infant stages. Similarly, participants (n=635) born from 1 October 1961 to 30 September 1962 were exposed to famine both in fetal stage and non-exposed stages. To minimise misclassification, these participants were excluded. The ranges of birthdate among four groups were as follows: non-exposed group (1 October 1962–30 September 1964), fetal-exposed group (1 October 1959–30 September 1961), infant-exposed group (1 January 1958–31 December 1958) and preschool-exposed group (1 January 1956–31 December 1957).
Classification of famine exposure area groups
Although the Great China Famine affected all regions of mainland China, the severity of famine exposure varied due to weather, agriculture policy and the grain distribution system, which varied among the different regions.30 As shown in our previous study, we used excess mortality during famine exposure periods compared with mortality before the famine as an objective indicator to reflect the severity of famine exposure.30
Excess mortality across mainland China ranged from 14.9% in Tianjin to 474.9% in Anhui province (online supplementary table S1).31 A 50% excess mortality was used as a cut-off to categorise the provinces into severely affected areas (excess mortality ≥50%) and less severely affected areas (excess mortality <50%).
Definition of arthritis
Information on arthritis was self-reported in the face-to-face household interviews. Participants were defined as having doctor-diagnosed arthritis if they answered ‘YES’ to the question ‘Have you ever been diagnosed with arthritis or rheumatism by a doctor?’
Anthropometric measurements were performed from June 2011 to March 2012. Bare height was measured using a portable stadiometer (seca 213, China) and was corrected to 0.1 cm. Fasting body weight was measured using a scale (Omron HN-286, China) and was corrected to 0.1 kg. Both height and weight were measured twice and the average values were used to calculate the body mass index (BMI=height (kg)/weight (m)2). The criteria recommended for Chinese adults were used, and participants were classified according to their BMI as normal weight group (BMI<24.0 kg/m2) and overweight/obesity group (BMI≥24.0 kg/m2).
Demographic, socioeconomic and lifestyle variables were collected using a standard questionnaire. The highest educational attainment of participants and their parents was categorised into the following four levels: primary school or below, junior school, high school, and college or above. Smoking status was categorised into current smoker, former smoker and never smoker. Current smoker was defined as someone who smoked at least one cigarette per day in the last year. Former smoker was defined as smoking >100 cigarettes in their lifetime and has quit more than 1 year ago. Never smoker was defined as smoking ≤100 cigarettes in their lifetime. Drinking status was categorised into never drinker, former drinker and current drinker. Current drinking was defined as someone who drank at least one standard unit per month in the last year (contained 15 g of alcohol). Former drinker was defined as drinking at least one unit per month and has quit more than 1 year ago. Never drinker was defined as someone who never consumed alcohol in their lifetime. The PA levels were collected through the International Physical Activity Questionnaire-Short Form and were categorised into three levels, namely light PA, middle PA and vigorous PA.32
Statistical analysis was performed using SPSS V.20.0 for Windows. Continuous variables (such as age, height, weight and BMI) were presented as mean and SD, and categorical variables (such as sex, arthritis, PA levels, smoking, drinking and educational attainment) were presented as percentages. A p<0.05 was considered statistically significant.
The χ2 test was used to compare the differences in arthritis prevalence, proportion of sex, and distribution of PA, smoking, drinking and educational attainment among the four groups. Analysis of variance was used to test the differences in age, height, weight and BMI among the four groups.
Logistic regression models were used to calculate the risks of arthritis among the fetal-exposed, infant-exposed and the preschool-exposed groups when the non-exposed group was used as the reference. Interactions between famine exposure and area, sex and BMI groups were tested by adding the multiplicative factors. To examine whether the risk of arthritis was affected by sex, region and nutritional environment in later life, we further conducted stratified analysis in different region, sex and BMI groups.
Participants in the exposed groups were older than those in the non-exposed group. To minimise the influence of difference in age on the association between famine exposure and risk of arthritis, we combined the non-exposed, fetal-exposed and preschool-exposed groups as a control group, which was called ‘Age balanced control group’. Its mean age is equal to the age of the infant-exposed group (approximately 54 years old), and this age-balanced group was used as the control group to further evaluate the risk of arthritis.
The basic characteristics of participants are presented in table 1. A total of 4124 participants were included in the present study. The prevalence of arthritis in the non-exposed, fetal-exposed, infant-exposed and preschool-exposed groups was 27.3%, 27.3%, 35.0% and 30.6%, respectively, and the infant-exposed group had a significantly higher prevalence of arthritis (p<0.05). While the non-exposed group was younger (p<0.001), it was high in overweight/obesity prevalence (p<0.05). In addition, the distribution of smoking status and education level varied among the four groups (p<0.05).
Table 2 presents the association between famine exposure in early life and the risk of arthritis in adulthood. Compared with the non-exposed group, the infant-exposed group had elevated risk of arthritis (OR=1.50; 95% CI 1.21 to 1.85; p<0.001). This association remained after adjustment for sex, BMI, smoking, drinking, PA and educational attainment (OR=1.65; 95% CI 1.29 to 2.11; p<0.001). However, similar associations were not observed for fetal-exposed group and preschool-exposed group after adjustment (p>0.05).
The results of the stratified analysis are presented in table 3. When stratified by famine severity, we found that only the infant-exposed group in severely affected areas had a higher risk of arthritis (OR=1.68; 95% CI 1.28 to 2.18; p<0.001). When stratified by sex, we observed that women, but not men, exposed to the famine in infancy were associated with increased risk of arthritis (OR=2.21; 95% CI 1.57 to 3.11; p<0.001). After being stratified by BMI group, only participants with adiposity in infant-exposed group had a higher risk of arthritis in adulthood (OR=2.46; 95% CI 1.70 to 3.55; p<0.001).
Table 4 presents the risks of arthritis among different famine-exposed groups when different control groups were applied. Compared with the age-balanced control group, the infant-exposed group had a significantly higher risk of arthritis in later life (OR=1.38; 95% CI 1.14 to 1.68; p=0.001). Similar results were observed in women (OR=1.57; 95% CI 1.19 to 2.08; p=0.001), participants with adiposity (OR=1.83; 95% CI 1.37 to 2.45; p<0.001) and those living in severely affected areas (OR=1.51; 95% CI 1.18 to 1.92; p=0.001).
Arthritis is a common disease and a leading cause of disability in Chinese adults. In this study, using the nationally representative data of CHARLS, the reported prevalence of arthritis collected by the questionnaire was 29.0% (1194/4124) in China, and individuals exposed to the Great China Famine in their infancy had a significantly higher risk of arthritis in their adulthood, particularly in women and those with adiposity. These findings could be helpful in identifying a population group with a high risk of arthritis and in improving interventions aimed to reduce the risk of arthritis.
This study identified that those who were exposed to the Great China Famine in infancy had a 65% higher risk of arthritis than the non-exposed group, but not in the fetal-exposed and preschool-exposed groups. To our knowledge, only one study explored the association between famine exposure and risk of arthritis in later life,26 which observed that adults exposed to the famine in the childhood stage had a significantly elevated risk of arthritis (OR=1.573). However, the childhood stage in that study did not distinguish the infant period from the preschool period. Extending the findings of that study, we found that only the infant-exposed group, rather than preschool-exposed group, was related with the elevated risk of arthritis in adulthood. Even when compared with the age-balanced control group, the same result was found (OR=1.38). These findings indicated that exposure to famine in infancy could be a vital period for arthritis, and improving nutritional condition in infants may be an important intervention for reducing the risk of arthritis in later life.
We speculated that the underlying mechanism linking famine exposure in infant stage and arthritis may be related with the abnormal metabolic profiles in later life. Adults who were exposed to the Dutch Famine in fetal stage were twice as likely to consume high-fat food and were less physically active than the unexposed groups.33 Overnutrition in later life triggered inflammation and led to abnormal metabolites, excessive secretion of proinflammatory cytokines and adipokines, and deregulated microRNAs, which may play a vital role in the pathophysiology of osteoarthritis.34 However, mechanisms at stake remain highly speculative, and future study is needed for further investigation and clarification.
A significant sex difference was observed in this study. We found that the significant association between famine exposure in infancy and risk of arthritis in adulthood was only detected in women, rather than in men. We speculated that this result may relate to gender bias in traditional Chinese culture, and girls may suffer more from severe food shortage than boys during the famine period.35 Furthermore, survivor bias might be another reason that led to the sex difference. Because men are more sensitive to food restriction than women and more likely to die,36 male survivors may be relatively healthier than female survivors.37
In addition, we also observed that the association was obvious in participants who are overweight and with obesity. This finding indicates that participants with rich nutrition condition in later life would enhance the adverse effects of famine. In previous studies, similar patterns were observed in the associations between famine exposure and risks of hypertension21 and type 2 diabetes mellitus.38 The mismatch between poor nutrition environment during early life and rich nutrition environment in later life may lead to the increased risk of arthritis in later life.39 40
The current study has several limitations. First, information on doctor-diagnosed arthritis was self-reported in the participants’ questionnaire, which may underestimate the prevalence of arthritis. However, this potential random misclassification may underestimate the magnitude of the association and would unlikely alter our observed relative risk. Second, the selection bias caused by severe famine is irremovable. The famine left healthier and stronger survivors, and people with severe metabolic abnormalities would be less likely to survive. Third, excess mortality was used to reflect the severity of famine exposure. However, confounding factors, such as weather and infections, may also have contributed to the premature deaths, but were not taken into consideration in this study. Additionally, information on personal severity of the famine, such as individual data on total daily calorie intake, was unavailable, and future studies with individual data are warranted.
Using large, nationally representative data, this study found that individuals exposed to the Great China Famine in their infancy had an elevated risk of arthritis in their adulthood, particularly in women and those with adiposity. These findings suggest that interventions aimed to improve nutrition in early life, as well as to control obesity in later life, could be possible approaches to reduce the risk of arthritis, and further epigenetic investigations could be helpful to elucidate the underlying mechanism.
What is already known on this subject
Existing research suggests that famine exposure in early life linked with non-communicable diseases, including hypertension, diabetes mellitus, dyslipidaemia, metabolic syndrome and chronic lung diseases.
However, to our knowledge, no studies using data of a large national epidemiology survey have estimated the association between famine exposure in early life and risk of arthritis in adults who are in their 50s.
What this study adds
The prevalence of arthritis was 29.0% (1194/4124) in Chinese adults aged over 45 years, and individuals exposed to the Great China Famine in their infant stage of life had a significantly higher risk of arthritis in their adulthood, especially in women and those with body mass index ≥24.0 kg/m2.
These findings could be helpful in identifying a population group with a high risk of arthritis and in improving interventions aimed to reduce the risk of arthritis.
The authors thank the CHARLS team for sharing the data and providing the sample training on using the data. The authors also thank all the interviewers who participated in the project.
Contributors JM, ZZ and BD designed this study and supervised the data analysis. ZW and ZZ carried out the initial analysis. LA critically reviewed the manuscript. All the authors were involved in writing the paper and had final approval of the submitted and published version.
Funding This work was granted by the National Natural Science Foundation of China (NSFC 81402692) awarded to ZZ and the Excellent Talents Fund Program of Peking University Health Science Center (BMU2017YJ002) awarded to BD.
Competing interests None declared.
Patient consent Obtained.
Ethics approval The present study is a secondary analysis of the CHARLS public data, which was approved by the Medical Ethics Committee of Peking University, and the Medical Ethics Committee of Peking University has granted current study exemption from ethics review.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.