Meta-analysis of individual participant data (IPD meta-analysis) has several advantages over meta-analysis using aggregated published data, including the possibility of using statistical methods such as a fine stratification analysis, interaction analysis between 2 risk factors, and absolute risk estimation. The Evidence for Cardiovascular Prevention from Observational Cohorts in Japan Study (EPOCH-JAPAN), which was initiated in 2005, is a collaborative research project for IPD meta-analysis and includes 13 participating cohort studies in Japan. We generated 2 pooled databases with data on all-cause mortality (n = 199 047) and cardiovascular outcomes (n = 90 528) and applied a stratified Cox model to account for the different baseline hazards between cohorts. The results of our analyses show the age- and sex-specific associations between all-cause and cardiovascular disease mortality and established cardiovascular risk factors (blood pressure, smoking, total cholesterol, proteinuria, and kidney function). During the 9 years of its existence, the results generated by EPOCH-JAPAN have had important implications for clinical medicine and public health policy in Japan. The project is expected to draw upon new analytical methods such as interaction analysis and absolute risk evaluation in the near future. We believe that, over the next decade, this project will continue to provide new insights that can be applied to research on other Asian populations.
Background: We identified factors associated with delayed first consultation for breast symptoms (patient delay), delayed diagnosis after first consultation (doctor delay), and advanced pathologic stage at presentation among 180 women with breast cancer in Thailand. Methods: In this cross-sectional study 180 patients with invasive breast cancer were interviewed about potential risk factors and markers of delayed presentation. Patient delay was defined as time from onset of symptoms to first consultation with a health care provider, and doctor delay was defined as time from first consultation with a health care provider to diagnosis of breast cancer. Linear regression and logistic regression were used for the data analyses. Results: Among the 180 patients, 17% delayed seeking consultation for longer than 3 months, and 42% reported a doctor delay of longer than 3 months. In multivariate linear analysis, a significant increase in patient delay was associated with higher family income and smoking; factors associated with increased doctor delay were previous breast symptoms, self-treatment, and travel time to the hospital. In multiple logistic regression, doctor delay was related to age at first birth (P = 0.003), previous breast symptoms (P = 0.01), and number of consultations with a surgeon before diagnosis (P = 0.007). Regarding stage of breast cancer, there were significant associations with age at diagnosis (P for trend = 0.04), education (P for trend = 0.01), family income (P for trend = 0.02), time to referral (P = 0.01), and number of consultations with a surgeon before diagnosis (P < 0.01). Conclusions: Hospital referral from a health care provider was a major contributor to delayed diagnosis. Breast cancer awareness campaigns in Thailand should target individuals in low- and high-income groups, as well as practitioners.
Background: Although obesity is increasing worldwide and becoming a major public health problem, some countries report a trend toward stabilization. We investigated prevalence trends in overweight/obesity and obesity among Korean adults during a 12-year period. Methods: This study was based on the Korean National Health and Nutrition Examination Survey (KNHANES) I (1998), II (2001), III (2005), and IV (2007–2009). The χ2 and ANOVA tests were used to compare the prevalence and mean values for age and BMI, respectively. P-values for trends were determined by linear and logistic regression analyses, with KNHANES phase as the continuous variable. Results: The prevalences of overweight/obesity in KNHANES I through IV were 50.8%, 57.4%, 62.5%, and 62.6%, respectively, among men (P for trend = 0.002, β = 0.021) and 47.3%, 51.9%, 50.0%, and 48.9% among women (P for trend = 0.017, β = −0.015). The respective prevalences of obesity were 26.0%, 32.4%, 35.1%, and 36.3% among men (P for trend = 0.006, β = 0.018) and 26.5%, 29.3%, 28.0%, and 27.6% among women (P for trend = 0.143, β = −0.008). During the same period, the respective prevalences of grade 2 obesity (BMI ≥30 kg/m2) were 1.7%, 2.8%, 3.6%, and 3.8% among men (P for trend = 0.075, β = 0.005) and 3.0%, 3.5%, 3.4%, and 4.0% among women (P for trend = 0.398, β = 0.003). Conclusions: The prevalences of overweight/obesity and obesity showed an upward trend among men during the 12-year period, whereas the prevalence of overweight/obesity slightly decreased among women from 2001.
Background: We attempted to identify the pathways by which demographic changes, socioeconomic inequalities, and availability of health factors influence life expectancy in low- and lower-middle-income countries. Methods: Data for 91 countries were obtained from United Nations agencies. The response variable was life expectancy, and the determinant factors were demographic events (total fertility rate and adolescent fertility rate), socioeconomic status (mean years of schooling and gross national income per capita), and health factors (physician density and human immunodeficiency virus [HIV] prevalence rate). Path analysis was used to determine the direct, indirect, and total effects of these factors on life expectancy. Results: All determinant factors were significantly correlated with life expectancy. Mean years of schooling, total fertility rate, and HIV prevalence rate had significant direct and indirect effects on life expectancy. The total effect of higher physician density was to increase life expectancy. Conclusions: We identified several direct and indirect pathways that predict life expectancy. The findings suggest that policies should concentrate on improving reproductive decisions, increasing education, and reducing HIV transmission. In addition, special attention should be paid to the emerging need to increase life expectancy by increasing physician density.
Background: The 2004 Niigata-Chuetsu earthquake of Japan caused considerable damage. We assessed long-term changes in psychological distress among earthquake victims during the period 5 years after the earthquake. Methods: The participants were people aged 18 years or older living in Yamakoshi, a community in Niigata Prefecture near the epicenter. A self-administered questionnaire survey was conducted annually for 5 consecutive years after the earthquake. Response rates were 1316/1841 (71.5%) in 2005, 667/1381 (48.3%) in 2006, 753/1451 (51.9%) in 2007, 541/1243 (43.5%) in 2008, and 814/1158 (70.3%) in 2009. The questionnaire asked about demographic characteristics, including sex, age, employment status, social network, and psychological status. Psychological distress was assessed using the 12-item General Health Questionnaire and was defined as a total score of 4 or higher. Results: The overall prevalence of psychological distress decreased (P < 0.0001) gradually from 2005 (51.0%) to 2008 (30.1%) but tended to increase from 2008 to 2009 (P = 0.1590). Subgroup analyses showed that prevalence did not decrease over the 5-year study period among participants with poor social contact (P = 0.0659). From 2008 to 2009 prevalence increased in women (+7.5%, P = 0.0403) and participants aged 65 years or older (+7.2%, P = 0.0400). Conclusions: The prevalence of psychological distress in Yamakoshi people decreased steadily during the 4 years immediately after the earthquake but appeared to increase thereafter. The earthquake victims are still reestablishing their lives. Thus, continued attention should be focused on maintaining and further assessing their mental health.
Background: We examined the psychometric properties of the Korean version of the 8-item Morisky Medication Adherence Scale (MMAS-8) among adults with hypertension. Methods: A total of 373 adults with hypertension were given face-to-face interviews in 2 cardiology clinics at 2 large teaching hospitals in Seoul, South Korea. Blood pressure was measured twice, and medical records were reviewed. About one-third of the participants (n = 109) were randomly selected for a 2-week test-retest evaluation of reliability via telephone interview. Results: Internal consistency reliability was moderate (Cronbach α = 0.56), and test-retest reliability was excellent (intraclass correlation = 0.91; P < 0.001), although a ceiling effect was detected. The correlation of MMAS-8 scores with scores for the original 4-item scale indicated that convergent validity was good (r = 0.92; P < 0.01). A low MMAS-8 score was significantly associated with poor blood pressure control (χ2 = 29.86; P < 0.001; adjusted odds ratio = 5.08; 95% CI, 2.56–10.08). Using a cut-off point of 6, sensitivity and specificity were 64.3% and 72.9%, respectively. Exploratory factor analysis identified 3 dimensions of the scale, with poor fit for the 1-dimensional construct using confirmatory factory analysis. Conclusions: The MMAS-8 had satisfactory reliability and validity and thus might be suitable for assessment and counseling regarding medication adherence among adults with hypertension in a busy clinical setting in Korea.
Background: Annually, about 400 cases of sudden unexpected death are attributed to cancer in Tokyo, Japan. These individuals may have been undiagnosed, or their medical conditions may not have been carefully evaluated before death. We examined medical consultations, cancer diagnoses, and economic status of all cancer deaths investigated by medical examiners in 2009. Methods: Among cases handled by the Tokyo Medical Examiner’s Office in 2009 (N = 12 493), records for all cases of cancer death (n = 400) were reviewed to determine the extent of medical care provided, diagnosis before death, and economic status of the decedent. Results: Most of the decedents (n = 232; 58%) had received a diagnosis of terminal/advanced cancer during a medical consultation. Most did not receive such medical consultations at home, despite their very weak physical condition. However, nearly one quarter of decedents (24%; 95/400) had not received a cancer diagnosis before death. The proportions of decedents who had been indigent, received no medical consulting, and had colon cancer were significantly higher among undiagnosed cases than among diagnosed cases. Indigent persons were the largest subgroup (n = 19; 43%) among those who had never received a medical consultation (n = 44). In addition, the proportion of those who had discontinued or received no medical consultation was higher among indigent persons than among non-indigent persons. Conclusions: The quality of medical services for cancer patients could be improved by educating general practitioners about terminal care, expanding efforts to monitor and diagnose cancer, especially among indigent patients, and increasing participation rates for colorectal cancer screening.
Background: Data regarding the effects of tea, coffee, and milk on the risk of colorectal cancer are inconsistent. We investigated associations of tea, coffee, and milk consumption with colorectal cancer risk and attempted to determine if these exposures were differentially associated with the risks of proximal colon, distal colon, and rectal cancers. Methods: Data from 854 incident cases and 948 controls were analyzed in a case-control study of colorectal cancer in Western Australia during 2005–07. Multivariable logistic regression was used to analyze the associations of black tea (with and without milk), green tea, herbal tea, hot coffee, iced coffee, and milk with colorectal cancer. Results: Consumption of 1 or more cups of herbal tea per week was associated with a significantly decreased risk of distal colon cancer (adjusted odds ratio, 0.37; 95% CI, 0.16–0.82; PTrend = 0.044), and consumption of 1 or more cups of iced coffee per week was associated with increased risk of rectal cancer (adjusted odds ratio, 1.52; 95% CI, 0.91–2.54; PTrend = 0.004). Neither herbal tea nor iced coffee was associated with the risk of proximal colon cancer. Hot coffee was associated with a possible increased risk of distal colon cancer. Black tea (with or without milk), green tea, decaffeinated coffee, and milk were not significantly associated with colorectal cancer risk. Conclusions: Consumption of herbal tea was associated with reduced risk of distal colon cancer, and consumption of iced coffee was associated with increased rectal cancer risk.
Background: The prevalence of alcohol consumption among Thais is high, around 30%. We quantified the relationship between alcohol drinking and mortality in a rural population in the most populous region of Thailand. Methods: The data were from the Khon Kaen Cohort Study. About 24 000 Thai adults were enrolled between 1990 and 2001, and follow-up for vital status continued until March 16, 2012. Mortality data were obtained from the Bureau of Policy and Strategy, Ministry of the Interior, Thailand. A Cox proportional hazards model was used to analyze the association between alcohol drinking and death, controlling for age, education level, and smoking, and floating absolute risk was used to estimate the 95% confidence intervals of hazard ratios. Results: In total, 18 457 participants (5829 men and 12 628 women) were recruited, of whom 3155 died (1375 men and 1780 women) during a median follow-up period of 13.6 years. Although alcohol drinking was common (64% of men and 25% of women), the amounts consumed were very low (average, 4.3 g/day in men and 0.8 g/day in women). As compared with never drinkers, mortality risk was lower among current drinkers and higher among ex-drinkers. Current drinking was not associated with mortality from cancer or diseases of the circulatory system, although ex-drinkers appeared to have a higher risk of death from the latter. Conclusions: The leading causes of mortality were not associated with current alcohol drinking at the low consumption levels observed in this population.
Background: When using the change-in-estimate criterion, a cutoff of 10% is commonly used to identify confounders. However, the appropriateness of this cutoff has never been evaluated. This study investigated cutoffs required under different conditions. Methods: Four simulations were performed to select cutoffs that achieved a significance level of 5% and a power of 80%, using linear regression and logistic regression. A total of 10 000 simulations were run to obtain the percentage differences of the 4 fitted regression coefficients (with and without adjustment). Results: In linear regression, larger effect size, larger sample size, and lower standard deviation of the error term led to a lower cutoff point at a 5% significance level. In contrast, larger effect size and a lower exposure–confounder correlation led to a lower cutoff point at 80% power. In logistic regression, a lower odds ratio and larger sample size led to a lower cutoff point at a 5% significance level, while a lower odds ratio, larger sample size, and lower exposure–confounder correlation yielded a lower cutoff point at 80% power. Conclusions: Cutoff points for the change-in-estimate criterion varied according to the effect size of the exposure–outcome relationship, sample size, standard deviation of the regression error, and exposure–confounder correlation.