2.1 Study participants
The Kailuan Study (accession number: ChiCTR-TNC-11001489) was a functional community individual-based cohort study in Tangshan, China, and the specific study design and procedures can be found in the team's previous studies 17,18. The Kailuan study began in 2006 and included 101,510 adults (81,110 males and 20,400 females) aged 18 years or older, all of them completed standard questionnaires (medical history and lifestyle) between 2006 and 2007, underwent health assessments every two years, including physical examinations (waist circumference (WC), weight, height, and blood pressure measurements) and laboratory tests (lipid assessments, fasting blood glucose (FBG), and serum creatinine (SCr)). All participants were followed up until their death or December 31, 2019.
According to the IDF global working definition of MS19, 14,241 MS patients who born between January 1, 1949, and December 31, 1974, were included. Individuals with missing data (n = 69) on WC, high density cholesterol (HDL), triglycerides, diastolic blood pressure (DBP), systolic blood pressure (SBP), fasting blood glucose (FBG), or those with CVD (n = 428) at the baseline survey in 2006 were excluded16.
Finally, 13,744 participants were served as the baseline cohort.
2.2 Famine exposure
Since famine in China occurred concentratedly from 1959 to 1962, we judged the period of famine exposed by birth time. Based on the previous Chinese famine research20, birth year was taken as the basis for classification of famine exposure, all participants were divided into three groups: no-exposed group (born between January 1, 1963 and December 31, 1974), uterine famine exposed group (born between January 1, 1959 and December 31, 1962), childhood famine exposed group (born between January 1, 1949 and December 31, 1958).
2.3 MS definition
MS was defined following the IDF Global Working Definition (IDF criteria) with the following criteria19: the presence of central obesity (waist circumference ≥ 90 cm for males or ≥ 80 cm for females), plus any two of following factors: (i) raised triglyceride level:≥ 1.7 mmol/L (150 mg/dl) or taking triglyceride-lowering medications; (ii) reduced HDL cholesterol:<1.03 mmol/L for men or < 1.29 mmol/L for females, or on lipid-lowering medications; (iii) hypertension: SBP > = 130 mmHg, or DBP > = 85 mmHg, or taking antihypertensive medications; (iv) raised FBG: >=100 mg/dL (5.6 mmol/L), or individuals who have been diagnosed with type 2 diabetes.
2.4 Follow-up and CVD
The starting point was defined as the date of completion of the 2006 annual baseline questionnaire and individuals were followed up until December 31, 2019. During follow-up, CVD incidence was assessed annually, and biochemical markers were collected every two years. The outcome event for the study was the first occurrence of a major CVD, which was defined as the composite of stroke and myocardial infarction (MI)21,22. The Hospital Discharge Register and Municipal Social Insurance Institution database were linked to identify the incidence of CVD based on The International Statistical Classification of Diseases and Related Health Problems 10th Revision (ICD-10) (I61 for intracerebral hemorrhagic stroke, I63 for ischemic stroke, and I21 for MI) 21,23. These two databases are updated annually based on follow-up and cover information on all participants in the Kailuan study. An expert panel collected and reviewed annual discharges records from 11 local hospitals to identify patients who were suspected of CVD.
Incident MI was diagnosed based on the World Health Organization's Multinational Monitoring of Trends and Determinants in Cardiovascular Disease (MONICA) criteria on basis of clinical symptoms and dynamic changes in clinical presentation, cardiac enzymes and electrocardiogram. Incident stroke diagnosed was according to neurological signs, clinical symptoms and neuroimaging (from CT or MRI) on the basis of World Health Organization's criteria. Death data were collected from provincial vital statistics offices, as described in previous studies24.
2.5 Covariates evaluation and measurement
The questionnaire design, anthropometry and laboratory data testing were the same as the literature published by our research group18. The data collected by the research include birth, gender, smoking, drinking, physical exercise, education level, history of medications and other basic information. During the survey, professionally trained medical staff completed the physical examination, including the measurement of their height, weight, waist circumference and blood pressure. Height and waist circumference were measured to the nearest 0.1 cm using the disposable tape measure.
Weight was determined by using calibrated portable digital weighing scale with 0.1 kg precision.
To ensure the reliability of the biochemical measurement results, a venous blood sample was obtained from all subjects who fasted for at least 8 hours before the measurement, and 5 ml of fasting elbow venous blood was collected between 7–9 am on the day of the physical examination, and the blood sample was centrifuged to obtain the upper serum for FBG values and lipid levels. All operations were performed strictly in accordance with the manufacturer's instructions, and blood samples were tested on a Hitachi (7600) automated biochemistry analyzer.
Age was calculated by subtracting the birthday from the beginning date of medical examination. Height and weight were measured to calculate body mass index (BMI) as weight (kg) / height2 (m2). Drinking was defined as consumption on average of 100 ml/day of spirits (alcohol content above 50%) in the past year. Smoking was defined as smoking an average of at least one cigarette per day for more than in the past year. Educational attainment was stratified into two levels: junior high school or below, senior high school or above. Physical exercise was defined as exercise time for > = 30 minutes at a time at least three times per week. Hypertension was defined as SBP over 140 mmHg or a DBP over 90 mmHg, or the fact that the patient was taking antihypertensive medications. Diabetes was defined as FBG > = 6.1 mmol/L, or the fact that the patient was taking hypoglycemic medications.
2.6 Statistical analysis of data
All data processing and analyses were performed using SAS version 9.4(SAS Institute, Cary, North Carolina), and R software version 3.6.0 (R Core Team, Vienna, Austria). The database was established through epidata3.1, entered by uniformly trained medical staff, and uploaded to the Oracle database of Kailuan General Hospital. All statistical tests were 2-sided, and p < 0.05 was considered statistically significant.
The normal or approximate normal distribution of the continuous variables was represented by x̄ ± s, the comparison between groups using analysis of variance, the skewed distribution was represented by M (P25, P75), and the intergroup comparison was used the Kruskal-Wallis test. The percentages described were used categorical variables and compared by χ2 tests.
Person-years of follow-up were calculated from the return date of the baseline questionnaire to the date of CVD diagnosis, death, loss to follow-up (n = 800, 5.82%), or end of follow-up (December 31, 2019) whichever occurred first. The incidence density of CVD in different groups in the MS individuals was calculated by dividing the number of events by the total number of follow-up person-years (1000/person-year), using the Log-rank test for comparison among groups. We used the weighted Cox regression model to analyze the effect of early life famine exposed in the MS individuals on the risk of CVD in adulthood, and the HR (Hazard Ratio) and 95%CI (confidence interval) was calculated. The model adjusted for age, gender, education level, smoking, drinking, physical exercise, BMI, history of diabetes, history of hypertension, low-density lipoprotein cholesterol, using antihypertensive medications, using antidiabetic medications, and using lipid-lowering medications.
Taking CVD as the dependent variable, and famine exposure as the independent variable, a stratified analysis was carried out by gender, smoking, and drinking. To verify the robustness of the results, a sensitivity analysis was performed after removing the individuals who had CVD incidents within two years or lost-to-review individuals.