Int. J. Behav. Med. (2010) 17:58–66 DOI 10.1007/s12529-009-9051-7
Socioeconomic Indicators and Cardiovascular Disease Incidence Among Japanese Community Residents: The Jichi Medical School Cohort Study Kaori Honjo & Akizumi Tsutsumi & Kazunori Kayaba & The Jichi Medical School Cohort Study Group
Published online: 25 June 2009 # International Society of Behavioral Medicine 2009
Abstract Background There has been little research in inequalities in risk of cardiovascular disease incidence by social class in Asia. Purpose The purpose of this study was to examine the association between socioeconomic indicators and risk of stroke and coronary heart disease in Japan. Method Data from the Jichi Medical School Study, a population-based prospective cohort study of approximately 11,000 Japanese men and women, were used. The average follow-up period was 11.7 years. Age- and areaadjusted hazard ratios with 95% confidence intervals (CIs) for education level/occupation were calculated using Cox proportional hazard regression analysis. Results Compared to those who completed education at age 14 or younger, the age and area-adjusted hazard ratios of intraparenchymal hemorrhage incidence for men who
completed education at age 15–17 and at age 18 or older were 0.42 (95% CI, 0.21–0.84) and 0.34 (95% CI, 0.14– 0.84), respectively. The age- and area-adjusted hazard ratios of intraparenchymal hemorrhage and subarachnoid hemorrhage incidence for female white-collar workers compared to female blue-collar workers were 0.28 (95% CI, 0.08– 0.98) and 3.23 (95% CI, 1.29, 8.01), respectively. No associations were found between education level and risk of coronary heart disease among both men and women. Conclusion These results suggest the pattern of social inequalities in health in Japan might be different from that in Western countries. Keywords Cardiovascular disease . Educational status . Incidence . Japan . Occupations
Introduction Study group members are listed in Appendix at the end of this article. K. Honjo Public Health, Department of Social and Environmental Medicine, Graduate School of Medicine, Osaka University, Suita, Osaka, Japan A. Tsutsumi Occupational Health Training Center, University of Occupational and Environmental Health, Fukuoka, Fukuoka, Japan K. Kayaba School of Health and Social Services, Saitama Prefectural University, Koshigaya, Saitama, Japan K. Honjo (*) Public Health, Department of Social and Environmental Health, Osaka University Graduate School of Medicine, 2-2 Yamadaoka Suita Osaka 565-9871, Japan e-mail:
[email protected]
Socioeconomic gradients in cardiovascular diseases, primarily coronary heart disease, are well documented in Europe and the USA [1–3]. Previous studies have reported consistent inverse associations between education level and incidence of coronary heart disease [4–6] and stroke [7–9]. Several studies have reported increased incidence and mortality of stroke and coronary heart disease among manual laborers compared to nonmanual laborers [10–14]. On the other hand, such evidence in Asian countries has been limited [15–17]. Previously conducted studies reported social inequalities in cardiovascular mortality among Japanese males and females [15, 16]. In addition, a recent study identified a U-shaped association between education level and total stroke incidence among Japanese women [17], while no associations were identified between education level and risk of total cardiovascular disease incidence among men and women [18]. Nevertheless, no
Int. J. Behav. Med. (2010) 17:58–66
research has examined the differences in incidence of stroke and coronary heart disease by education level among Japanese men. Also, no research using occupational category as a social class indicator existed in Japan. Because the etiological mechanisms of stroke and coronary heart disease are different, the association between social class indicator and risk of each cardiovascular disease incidence could also be different. Lacunar infarcts and small intracerebral artery lesions are more prevalent than large-artery atherothrombotic ischemic stroke in Japan [19, 20]. Blood pressure, but not serum total cholesterol level, appears strongly associated with stroke, while dyslipidemia and lipid system problems seem to be more important etiologic mechanisms for coronary heart disease [19–22]. Therefore, we set both stroke and coronary heart disease as our outcomes. In addition, although both education level and occupation were often used as indicators of social stratification, each of them is assumed to encompass different aspects of social stratification [23]. These differences in aspects of social class that each indicator reflects may generate different results between education level and occupation. Thus, we conducted this study to examine the association between socioeconomic indicators (i.e., education level and occupation) and risk of stroke and coronary heart disease incidence. As far as we know, this is the first prospective study to examine the association between socioeconomic indicators and risk of stroke and coronary heart disease, at least among men in Asia. The hypotheses are (a) higher education levels are associated with lower risk of stroke and/or coronary heart disease, compared to lower education levels; and (b) white-collar workers have lower risk of stroke and/or coronary heart disease compared to blue-collar workers.
59
community included those aged 20–69 years, and another community included those aged 35 years and older. The remaining two communities invited all residents. In each community, the local government invited all potential participants by sending invitation letters and making public announcements. The invitation mentioned that persons visiting hospitals or clinics because of cardiovascular diseases did not have to take the examination. The overall response rate was 65.4%; the response rates varied by community (26–96%) [24]. The cohort was composed of 12,490 Japanese residents (4,911 men and 7,579 women) from 12 rural communities across Japan. Of the 12,490 participants, we excluded 97 ineligible participants for two reasons: two had moved away before starting, while 95 rejected the follow-up. We included the remaining 12,393 participants in the study. The present study was approved by each municipal government and by the Ethics Committee for Epidemiological Research at Jichi Medical School. Written informed consent was obtained from all prospective participants. Study Population and Baseline Survey Of the 12,393 participants, we excluded those who reported medical history of cancer, stroke, or heart disease at baseline (n=840); we further excluded those who did not provide information on years of education completed (n= 891) or occupation (n = 130). The remaining 10,640 participants (4,129 men and 6,511 women) were considered to be the study population. The baseline data were obtained between 1992 and 1995 using a standardized questionnaire and physical examination; the study population was followed until the end of 2005. Education Level
Methods Study Cohort The Jichi Medical School Cohort Study was a populationbased prospective study that explored the risk factors of cardiovascular diseases among 12 residential communities [24]. The study makes use of data collected through a mass screening examination program for cardiovascular diseases administrated by the Japanese government since 1983, in accordance with the Health and Medical Service Law for the Aged in Japan. The law requires municipal governments to offer the program to all adult residents who are not offered physical examinations at their workplaces. The target population varied according to each community. Eight communities invited residents aged 40–69 to be subjects of the mass screening examination program. One
Education level was measured by self-reported age at completion of education as part of the baseline questionnaire. Age at completion of education was then categorized into three groups: (a) 14 years old and younger (uncompleted junior high school education); (b) between 15 and17 years old (completed junior high school education); and (c) 18 years old and older (completed high school education). Occupation Category Information on current occupation was obtained in the baseline questionnaire using National Statistics guidelines [25]; subjects were asked to select one of the following: farming/forestry, fishery, security, transportation, construction, production, office worker, professional, service indus-
60
try, homemaker, and retired. The first six job categories, from farming/forestry to production, were categorized as blue-collar jobs, while office worker, professional, and service industry were categorized as white-collar jobs. If subjects reported themselves to be managers or employers, we categorized them as having a white-collar job regardless of their chosen job category. Finally, the categories of retired and homemaker were treated as having no job. Although our occupational categorization was based on simple questions, previous studies in this cohort showed that the occupation category demonstrated reasonable association with psychosocial job characteristics and lifestyle factors [16, 26]. Other Covariates Information on age, gender, marital status, medical history, smoking habits (currently smoking, ex-smoker, never smoked), alcohol intake (drinking status, kinds, frequency, and quantity of drinking alcohol consumed), and physical activity were obtained from the responses to the baseline questionnaire. The amount of ethanol intake per day was calculated by using information obtained by questionnaire. Physical activity index was estimated by calculating the weighted sum of hours spent at five levels of activity during a normal working day, estimated by using the Framingham Study questionnaire [27]. Height, weight, cholesterol level, and blood pressure were measured at the baseline physical examination. Height was measured without shoes. Body weight was recorded with the subject clothed; 0.5 kg in summer or 1 kg in the other seasons was subtracted from the recorded weight. Body mass index (weight (kg)/height (m2)) was calculated by the measured weight and height. We categorized subjects whose body mass index (BMI) was 27 or higher as obese. The systolic and diastolic blood pressures were measured with a fully automated sphygmomanometer, BP203RV-II (Nippon Colin, Komaki, Japan), placed on the right arm of a seated subject who had rested in the sitting position for 5 min before the measurement. Hypertension status at baseline was determined from the baseline questions on medical history of hypertension and measured blood pressure. Participants were classified as hypertensive if (1) their systolic/diastolic blood pressure ≥140/90 mmHg, or (2) they had been clinically diagnosed as hypertensive. Diabetes status was determined by the baseline questions on medical history. Confirmation of Cardiovascular Disease Incidence The endpoints of this study were incidence of total stroke, intraparenchymal hemorrhage, subarachnoid hemorrhage,
Int. J. Behav. Med. (2010) 17:58–66
ischemic stroke, and total coronary heart disease. Subjects were followed for a mean of 11.7 years from the time of baseline survey until December 31, 2005. Residential status, including survival status, was confirmed annually through the residential registry in each area. A total of 369 respondents (3.0%) who had moved from their original residential area and five respondents (0.1 %) who were lost to follow-up during the follow-up period were treated as censored at that time. We used the mass screening examination system to check on all participants every year. We asked them directly whether they had a history of stroke and/or heart disease after they first enrolled; if they had, we then asked about the hospital they had visited and when the incident(s) occurred to ascertain the incidence of disease. We directly contacted by mail or phone those participants who had not undergone a screening examination. We also checked the medical records if participants had visited hospitals. Public health nurses also visited the participants to obtain information. If an incidence case was identified, we then filled out a form and duplicated the available computer tomography (CT) films (for strokes) or electrocardiograms (for myocardial infarction). Diagnoses were determined independently by a committee composed of radiologists, neurologists, and cardiologists. Strokes were confirmed according to the criteria of the Yanagawa group for stroke of the Ministry of Health and Welfare, which requires existing focal and inconvulsive neurological deficits lasting 24 h or longer with a clear onset of disease [28]. For each subtype of stroke, such as subarachnoid hemorrhage, intraparenchymal hemorrhage, and ischemic stroke (thrombotic or embolic stroke), a definite diagnosis was established based on examination of CT scans or magnetic resonance images. Myocardial infarction was confirmed in the medical records according to the criteria of the Monitoring Trends and Determinants of Cardiovascular Disease (MONICA) project [29]. Coronary heart disease was defined as acute myocardial infarction or sudden cardiac death within 1 h of onset. Statistical Analyses The gender-specific relationship between education level, occupation, and the age-adjusted selected characteristics of participants at baseline were tested using analysis of covariance for continuous variables and logistic regression for dichotomous variables. Statistical analyses were based on incidence rates of total strokes, total coronary heart disease, and subtypes of stroke during follow-up. For each individual, person-years of follow-up were calculated from the date of her/his baseline survey to the first end point: incidence, death, emigration, or December 31, 2005,
Int. J. Behav. Med. (2010) 17:58–66
61
whichever came first. The incidence rates of total strokes, stroke subtypes, and total coronary heart disease were calculated according to education level and occupation. Hazard ratios (HRs) with 95% confidence intervals (CIs) for education level were calculated after adjustment for age and area, which were considered potential confounding variables, with Cox proportional hazard regression analysis (model 1). A similar analysis was conducted for occupation (model 2). To examine the independent effects of education level and occupation, we then included education level and occupation in the same model (model 3). To examine the mediating effect of behavioral and physical risk factors for cardiovascular disease, we added the hypothesized risk factors (total cholesterol level, physical activity, ethanol intake, marital status, smoking habit, obesity, hypertension, and diabetes) in either model 1 or model 2 (model 4).
Results Table 1 shows the age-adjusted mean values or proportions of selected characteristics of the study population at baseline according to education level and occupation. Seventeen percent of the participants reported having completed their education at age 14 or younger; 39% of the participants were white-collar workers. While little difference was seen in the distribution of cardiovascular risk factors by education level and occupation among men, significant differences were observed among women. The
age-adjusted prevalence of current smoker status, obesity, and hypertension at baseline was inversely associated with education level among women. Blue-collar female workers were more likely to be physically active compared to whitecollar counterparts. Table 2 presents the multivariable HRs (95% CI) of cardiovascular disease incidence according to education level, using the lowest education level group (education completed at age 14 or younger) as a reference group, and occupation, using blue-collar job status as a reference group among men. During the follow-up period, 197 strokes and 57 instances of coronary heart disease were documented. Men who completed their education at age 15 or older had a lower incidence of intraparenchymal hemorrhage compared to those who completed their education at age 14 or younger. These associations did not change significantly after additional adjustments for occupation and behavioral or physical factors. Male white-collar workers had a higher risk of coronary heart disease compared to blue-collar workers, although the association was statistically marginal. No significant associations were identified between education level and/or occupation and risk of total stroke, subarachnoid hemorrhage, or ischemic stroke. Among women, 170 strokes and 27 instances of coronary heart disease were documented during the follow-up period (Table 3). No significant associations were identified between education level and risk of stroke and/or coronary heart disease among women. On the other hand, blue-collar job status was associated with increased
Table 1 Age-adjusted cardiovascular risk characteristics according to education level/occupation by gender, Jichi Medical School Cohort Study, 1992/95 Baseline (N=10,770) Age at education completion
Occupation
Male (n=4,129)
Female (n=6,511)
Male (n=4,129)
Female (n=6,511)
≤14 15–17 ≥18
≤14 15–17 ≥18
Blue White No job
Blue White No job
%
12
52
36 Pa
20
48
32 Pa
49
39
12
Pa
32
27
41
Pa
Age (mean)
65
56
49
