Search tips
Search criteria

Results 1-25 (157)

Clipboard (0)

Select a Filter Below

Year of Publication
more »
1.  The Effect of Adverse Housing and Neighborhood Conditions on the Development of Diabetes Mellitus among Middle-aged African Americans 
American journal of epidemiology  2007;166(4):379-387.
The authors examined the associations of observed neighborhood (block face) and housing conditions with the incidence of diabetes by using data from 644 subjects in the African-American Health Study (St. Louis area, Missouri). They also investigated five mediating pathways (health behavior, psychosocial, health status, access to medical care, and sociodemographic characteristics) if significant associations were identified. The external appearance of the block the subjects lived on and housing conditions were rated as excellent, good, fair, or poor. Subjects reported about neighborhood desirability. Self-reported diabetes was obtained at baseline and 3 years later. Of 644 subjects without self-reported diabetes, 10.3% reported having diabetes at the 3-year follow-up. Every housing condition rated as fair-poor was associated with an increased risk of diabetes, with odds ratios ranging from 2.53 (95% confidence interval: 1.47, 4.34 for physical condition inside the building) to 1.78 (95% confidence interval: 1.03, 3.07 for cleanliness inside the building) in unadjusted analyses. No association was found between any of the block face conditions or perceived neighborhood conditions and incident diabetes. The odds ratios for the five housing conditions were unaffected when adjusted for the mediating pathways. Poor housing conditions appear to be an independent contributor to the risk of incident diabetes in urban, middle-aged African Americans.
PMCID: PMC4519088  PMID: 17625220
African Americans; aging; diabetes mellitus; housing; questionnaires; residence characteristics
2.  A Multiethnic Cohort in Hawaii and Los Angeles: Baseline Characteristics 
American journal of epidemiology  2000;151(4):346-357.
The authors describe the design and implementation of a large multiethnic cohort established to study diet and cancer in the United States. They detail the source of the subjects, sample size, questionnaire development, pilot work, and approaches to future analyses. The cohort consists of 215,251 adult men and women (age 45–75 years at baseline) living in Hawaii and in California (primarily Los Angeles County) with the following ethnic distribution: African-American (16.3%), Latino (22.0%), Japanese-American (26.4%), Native Hawaiian (6.5%), White (22.9%), and other ancestry (5.8%). From 1993 to 1996, participants entered the cohort by completing a 26-page, self-administered mail questionnaire that elicited a quantitative food frequency history, along with demographic and other information. Response rates ranged from 20% in Latinos to 49% in Japanese-Americans. As expected, both within and among ethnic groups, the questionnaire data show substantial variations in dietary intakes (nutrients as well as foods) and in the distributions of non-dietary risk factors (including smoking, alcohol consumption, obesity, and physical activity). When compared with corresponding ethnic-specific cancer incidence rates, the findings provide tentative support for several current dietary hypotheses. As sufficient numbers of cancer cases are identified through surveillance of the cohort, dietary and other hypotheses will be tested in prospective analyses.
PMCID: PMC4482109  PMID: 10695593
alcohol drinking; cohort studies; diet; ethnic groups; obesity; physical fitness; prospective studies; smoking
3.  Calibration of the Dietary Questionnaire for a Multiethnic Cohort in Hawaii and Los Angeles 
American journal of epidemiology  2000;151(4):358-370.
The performance of the dietary questionnaire used in a multiethnic cohort study in Hawaii and Los Angeles was assessed in a calibration substudy that compared diet reported from the questionnaire with three 24-hour dietary recalls. For the calibration substudy, subjects from each of eight subgroups defined by sex and ethnic group (African-American, Japanese-American, Latino, and White) were chosen randomly from among the cohort members, and each participant’s previous day’s diet was assessed by telephone recall on three occasions over approximately 2 months. After completing the three 24-hour recalls, each calibration subject was sent a second questionnaire; 1,606 persons completed three recalls and a second questionnaire (127 to 267 per ethnic-sex group). This report describes correlation coefficients and calibration slopes for the relation between the 24-hour recalls and second questionnaire values for a selected set of macro- and micronutrients, as absolute intakes, nutrient densities, and calorie-adjusted nutrients. In all subgroups, estimates of the correlation between the questionnaire and 24-hour recalls were greater after energy adjustment (average correlations ranged from 0.57–0.74 for nutrient densities and from 0.55–0.74 for calorie-adjusted nutrients) than when absolute nutrient values were used (average range 0.26–0.57). For absolute nutrient intakes, the correlations were greatest for Whites, somewhat lower for Japanese-Americans and Latinos, and lowest for African-Americans. After energy adjustment, the difference between subgroups were diminished, and the correlations were generally highly satisfactory.
PMCID: PMC4482461  PMID: 10695594
calibration; diet surveys; epidemiologic methods; ethnic groups; nutrition surveys; questionnaires
4.  A Comparison of Risk Factors for Calcified Atherosclerotic Plaque in the Coronary, Carotid, and Abdominal Aortic Arteries 
American journal of epidemiology  2007;166(3):340-347.
The extent of shared risk factors for calcified atherosclerotic plaque (CAP) of the coronary, carotid, and abdominal aortic arteries is unknown. CAP was measured by computed tomography in 1,125 individuals in families affected with diabetes. Statistical methods adjusted for the lack of independence between observations. CAP scores were standardized, and tests of interaction were conducted to compare risk factor relations across vascular beds. The average age of the cohort was 61 years, and 84% had diabetes. The correlation in CAP scores across vascular beds ranged from 0.59 to 0.72. Age, albumin/creatinine ratio, hemoglobin A1c, diabetes, hypertension, and lipid-lowering therapy were correlated with quantity of CAP in all vascular beds (all p < 0.05); no differences in the strength of these relations were noted. In contrast, other significant correlates differed in the strength of their relations with CAP. The risk factor pack-years of smoking was most strongly correlated with CAP in the abdominal aorta (p < 0.005). Male gender, previous myocardial infarction, and coronary revascularization were most strongly correlated with CAP in the coronary arteries (p < 0.0001). In summary, CAPs of the coronary, carotid, and abdominal aortic arteries generally share common risk factors, even though several of these factors have a greater impact on CAP in one vascular bed than another.
PMCID: PMC4450101  PMID: 17493948
atherosclerosis; calcification; physiologic; diabetes mellitus; type 2; North Carolina; risk factors; siblings
5.  Chronic Venous Disease in an Ethnically Diverse Population The San Diego Population Study 
American journal of epidemiology  2003;158(5):448-456.
In a 1994–1998 cross-sectional study of a multiethnic sample of 2,211 men and women in San Diego, California, the authors estimated prevalence of the major manifestations of chronic venous disease: spider veins, varicose veins, trophic changes, and edema by visual inspection; superficial and deep functional disease (reflux or obstruction) by duplex ultrasonography; and venous thrombotic events based on history. Venous disease increased with age, and, compared with Hispanics, African Americans, and Asians, non-Hispanic Whites had more disease. Spider veins, varicose veins, superficial functional disease, and superficial thrombotic events were more common in women than men (odds ratio (OR) = 5.4, OR = 2.2, OR = 1.9, and OR = 1.9, respectively; p < 0.05), but trophic changes and deep functional disease were less common in women (OR = 0.7 for both; p < 0.05). Visible (varicose veins or trophic changes) and functional (superficial or deep) disease were closely linked; 92.0% of legs were concordant and 8.0% discordant. For legs evidencing both trophic changes and deep functional disease, the age-adjusted prevalences of edema, superficial events, and deep events were 48.2%, 11.3%, and 24.6%, respectively, compared with 1.7%, 0.6%, and 1.3% for legs visibly and functionally normal. However, visible disease did not invariably predict functional disease, or vice versa, and venous thrombotic events occurred in the absence of either.
PMCID: PMC4285442  PMID: 12936900
cross-sectional studies; diagnostic imaging; ethnic groups; population; thrombosis; ultrasonics; veins
American journal of epidemiology  2014;179(11):1301-1311.
The International Agency for Research on Cancer controversially has classified formaldehyde as causing nasopharyngeal carcinoma and myeloid leukaemia. To provide further information on this question, we extended follow-up of 14,008 chemical workers at six factories in England and Wales, covering the period 1941-2012. Mortality was compared with national death rates, and associations with incident upper airways cancer and leukaemia were explored in nested case-control analyses. Excess deaths were observed from cancers of the oesophagus (100 v 93.1 expected), stomach (182 v 141.4), rectum (107 v 86.8), liver (35 v 26.9) and lung (813 v 645.8), but none of these tumours exhibited a clear exposure-response relationship. Nested case-control analyses of 115 men with upper airways cancer (including one nasopharyngeal cancer), 92 with leukaemia, and 45 with myeloid leukaemia indicated no elevations of risk in the highest exposure category (high exposure for ≥1 year). When the two highest exposure categories were combined the odds ratio for myeloid leukaemia was 1.26 (95%confidence interval: 0.39, 4.08). Our results provide no support for a hazard of myeloid leukaemia, nasopharyngeal carcinoma or other upper airways tumours from formaldehyde, and indicate that any excess risk of these cancers, even from relatively high exposures, is at most small.
PMCID: PMC4189094  PMID: 24714728
Cancer; chemical industry; formaldehyde; mortality; myeloid leukaemia; nasopharyngeal cancer
7.  Gang Exposure and Pregnancy Incidence among Female Adolescents in San Francisco: Evidence for the Need to Integrate Reproductive Health with Violence Prevention Efforts 
American journal of epidemiology  2008;167(9):1102-1109.
Among a cohort of 237 sexually active females aged 14–19 recruited from community venues in a predominantly Latino neighborhood in San Francisco we examined the relationship between gang exposure and pregnancy incidence over two years of follow-up. Using discrete-time survival analysis we investigated whether individual and partner gang membership were associated with pregnancy incidence and determined whether partnership characteristics, contraceptive behaviors and pregnancy intentions mediated the relationship between gang membership and pregnancy. Pregnancy incidence was determined by urine-based testing and self-report. Seventy-seven percent of participants were Latinas, with one in five born outside the U.S. One-quarter (27.4%) became pregnant over follow-up. Participants’ gang membership had no significant effect on pregnancy incidence (Hazard Ratio (HR)=1.25; 95% confidence interval (CI): 0.54, 3.45); however, having partners who were in gangs was associated with pregnancy (HR=1.90; 95% CI: 1.09, 3.32). Perceived male partner’s pregnancy intentions and having a partner in detention each mediated the effect of partner’s gang membership on pregnancy risk. Increased pregnancy incidence among young women with gang-involved partners highlights the importance of integrating reproductive health prevention into programs for gang-involved youth. In addition, high pregnancy rates indicate a heightened risk for sexually transmitted infections.
PMCID: PMC4226274  PMID: 18308693
pregnancy; violence; sexual partners; sexually transmitted diseases; Hispanic Americans; adolescent
8.  Prevalence of and Risk Factors for Subclinical Cardiovascular Disease in Selected US Hispanic Ethnic Groups 
American journal of epidemiology  2008;167(8):962-969.
In this study, the authors determined the prevalence and extent of cardiovascular disease (CVD) risk factors and subclinical CVD in four US Hispanic subgroups, as well as associations between the CVD risk factors and subclinical CVD in these groups. Participants were 1,437 Hispanic men and women enrolled in the Multi-Ethnic Study of Atherosclerosis in 2000–2002. Fifty-six percent were Mexican-American, 12% were Dominican-American, 14% were Puerto Rican-American, and 18% were Other Hispanic-American. All participants underwent clinical examinations for coronary artery calcium, thoracic aortic calcium, carotid intimal-medial thickness, ankle-brachial index, left ventricular mass, and left ventricular size. Mexican Americans had the highest levels of coronary artery calcium, thoracic aortic calcium, and carotid intimal-medial thickness, while Puerto Rican Americans had the highest prevalence of an ankle-brachial index less than 1.0 and levels of left ventricular mass. The magnitudes of the associations between coronary artery calcium and age, sex, and body mass index were similar across all Hispanic subgroups. However, there were differences in the magnitude and significance of the associations between coronary artery calcium and hypertension, hypercholesterolemia, and cigarette smoking among the different Hispanic subgroups. This finding was also present for the other subclinical CVD measures. These results suggest a differential relationship between risk factors and either prevalence or extent of subclinical disease by Hispanic subgroup.
PMCID: PMC4107279  PMID: 18283034
atherosclerosis; cardiovascular diseases; ethnic groups; Hispanic Americans; risk factors
9.  Long-term Plasma Lipid Changes Associated with a First Birth The Coronary Artery Risk Development in Young Adults Study 
American journal of epidemiology  2004;159(11):1028-1039.
Previous studies have reported declines in high density lipoprotein (HDL) cholesterol 1–2 years after pregnancy. In 1986–1996, the authors prospectively examined the association between childbearing and changes in fasting plasma lipids (low density lipoprotein, HDL, and total cholesterol; triglycerides) among 1,952 US women (980 Black, 972 White) in the Coronary Artery Risk Development in Young Adults study. Repeated-measures multiple linear regression was used to examine lipid changes over three time intervals (baseline to years 5, 7, and 10) in time-dependent follow-up groups: P0 (0 pregnancies), P1 (≥1 miscarriages/abortions), B1 (1 birth), and B2 (≥2 births). Means stratified by race and baseline parity (nulliparous or parous) were fully adjusted for study center, time, height, baseline diet, and other baseline and time-dependent covariates (age, smoking, education, weight, waist circumference, alcohol intake, oral contraceptive use, physical activity, short pregnancies). For both races, fully adjusted HDL cholesterol declines of −3 to −4 mg/dl were associated with a first birth versus no pregnancies during follow-up (p < 0.001). Higher-order births were not associated with greater declines in HDL cholesterol (B2 similar to B1, no association among women parous at baseline). In Whites, total and low density lipoprotein cholesterol declines were associated with follow-up births. HDL cholesterol declines of −3 to −4 mg/dl after a first birth persisted during the 10 years of follow-up independent of weight, central adiposity, and selected behavior changes.
PMCID: PMC4107869  PMID: 15155287
ethnic groups; lipids; lipoproteins; HDL cholesterol; parity; pregnancy
10.  Effectiveness of Highly Active Antiretroviral Therapy among Injection Drug Users with Late-Stage Human Immunodeficiency Virus Infection 
American journal of epidemiology  2005;161(11):999-1012.
Highly active antiretroviral therapy (HAART) has been shown to be effective in different populations, but data among injection drug users are limited. Human immunodeficiency virus-infected injection drug users recruited into the Acquired Immunodeficiency Syndrome Link to Intravenous Experiences (ALIVE) Study as early as 1988 were tested semiannually to identify their first CD4-positive T-lymphocyte cell count below 200/μl; they were followed for mortality through 2002. Visits were categorized into the pre-HAART (before mid-1996) and the HAART eras and further categorized by HAART use. Survival analysis with staggered entry was used to evaluate the effect of HAART on acquired immunodeficiency syndrome-related mortality, adjusting for other medications and demographic, clinical, and behavioral factors. Among 665 participants, 258 died during 2,402 person-years of follow-up. Compared with survival in the pre-HAART era, survival in the HAART era was shown by multivariate analysis to be improved for both those who did and did not receive HAART (relative hazards = 0.06 and 0.33, respectively; p < 0.001). Inferences were unchanged after restricting analyses to data starting with 1993 and considerations of lead-time bias and human immunodeficiency viral load. The annual CD4-positive T-lymphocyte cell decline was less in untreated HAART-era participants than in pre-HAART-era participants (— 10/μl vs. —37/μl, respectively), suggesting that indications for treatment may have contributed to improved survival and that analyses restricted to the HAART era probably underestimate HAART effectiveness.
PMCID: PMC4078731  PMID: 15901620
antiretroviral therapy; highly active; HIV; substance abuse; intravenous; substance-related disorders; survival; treatment outcome
11.  Personal Use of Hair Dye and the Risk of Certain Subtypes of Non-Hodgkin Lymphoma 
American journal of epidemiology  2008;167(11):1321-1331.
Personal use of hair dye has been inconsistently linked to risk of non-Hodgkin lymphoma (NHL), perhaps because of small samples or a lack of detailed information on personal hair-dye use in previous studies. This study included 4,461 NHL cases and 5,799 controls from the International Lymphoma Epidemiology Consortium 1988–2003. Increased risk of NHL (odds ratio (OR) = 1.3, 95% confidence interval (CI): 1.1, 1.4) associated with hair-dye use was observed among women who began using hair dye before 1980. Analyses by NHL subtype showed increased risk for follicular lymphoma (FL) and chronic lymphocytic leukemia/small lymphocytic lymphoma (CLL/SLL) but not for other NHL subtypes. The increased risks of FL (OR = 1.4, 95% CI: 1.1, 1.9) and CLL/SLL (OR = 1.5, 95% CI: 1.1, 2.0) were mainly observed among women who started using hair dyes before 1980. For women who began using hair dye in 1980 or afterward, increased FL risk was limited to users of dark-colored dyes (OR = 1.5, 95% CI: 1.1, 2.0). These results indicate that personal hair-dye use may play a role in risks of FL and CLL/SLL in women who started use before 1980 and that increased risk of FL among women who started use during or after 1980 cannot be excluded.
PMCID: PMC4025953  PMID: 18408225
case-control studies; hair dyes; lymphoma; non-Hodgkin
12.  Occupational Exposure to Pesticides and Risk of Adult Brain Tumors 
American journal of epidemiology  2008;167(8):976-985.
The authors examined incident glioma and meningioma risk associated with occupational exposure to insecticides and herbicides in a hospital-based, case-control study of brain cancer. Cases were 462 glioma and 195 meningioma patients diagnosed between 1994 and 1998 in three US hospitals. Controls were 765 patients admitted to the same hospitals for nonmalignant conditions. Occupational histories were collected during personal interviews. Exposure to pesticides was estimated by use of a questionnaire, combined with pesticide measurement data abstracted from published sources. Using logistic regression models, the authors found no association between insecticide and herbicide exposures and risk for glioma and meningioma. There was no association between glioma and exposure to insecticides or herbicides, in men or women. Women who reported ever using herbicides had a significantly increased risk for meningioma compared with women who never used herbicides (odds ratio = 2.4, 95% confidence interval: 1.4, 4.3), and there were significant trends of increasing risk with increasing years of herbicide exposure (p = 0.01) and increasing cumulative exposure (p = 0.01). There was no association between meningioma and herbicide or insecticide exposure among men. These findings highlight the need to go beyond job title to elucidate potential carcinogenic exposures within different occupations.
PMCID: PMC3967588  PMID: 18299277
13.  Comparison of Random Forest and Parametric Imputation Models for Imputing Missing Data Using MICE: A CALIBER Study 
American journal of epidemiology  2014;179(6):764-774.
Multivariate imputation by chained equations (MICE) is commonly used for imputing missing data in epidemiologic research. The “true” imputation model may contain nonlinearities which are not included in default imputation models. Random forest imputation is a machine learning technique which can accommodate nonlinearities and interactions and does not require a particular regression model to be specified. We compared parametric MICE with a random forest-based MICE algorithm in 2 simulation studies. The first study used 1,000 random samples of 2,000 persons drawn from the 10,128 stable angina patients in the CALIBER database (Cardiovascular Disease Research using Linked Bespoke Studies and Electronic Records; 2001–2010) with complete data on all covariates. Variables were artificially made “missing at random,” and the bias and efficiency of parameter estimates obtained using different imputation methods were compared. Both MICE methods produced unbiased estimates of (log) hazard ratios, but random forest was more efficient and produced narrower confidence intervals. The second study used simulated data in which the partially observed variable depended on the fully observed variables in a nonlinear way. Parameter estimates were less biased using random forest MICE, and confidence interval coverage was better. This suggests that random forest imputation may be useful for imputing complex epidemiologic data sets in which some patients have missing data.
PMCID: PMC3939843  PMID: 24589914
angina, stable; imputation; missing data; missingness at random; regression trees; simulation; survival
14.  Comparison of Random Forest and Parametric Imputation Models for Imputing Missing Data Using MICE: A CALIBER Study 
American Journal of Epidemiology  2014;179(6):764-774.
Multivariate imputation by chained equations (MICE) is commonly used for imputing missing data in epidemiologic research. The “true” imputation model may contain nonlinearities which are not included in default imputation models. Random forest imputation is a machine learning technique which can accommodate nonlinearities and interactions and does not require a particular regression model to be specified. We compared parametric MICE with a random forest-based MICE algorithm in 2 simulation studies. The first study used 1,000 random samples of 2,000 persons drawn from the 10,128 stable angina patients in the CALIBER database (Cardiovascular Disease Research using Linked Bespoke Studies and Electronic Records; 2001–2010) with complete data on all covariates. Variables were artificially made “missing at random,” and the bias and efficiency of parameter estimates obtained using different imputation methods were compared. Both MICE methods produced unbiased estimates of (log) hazard ratios, but random forest was more efficient and produced narrower confidence intervals. The second study used simulated data in which the partially observed variable depended on the fully observed variables in a nonlinear way. Parameter estimates were less biased using random forest MICE, and confidence interval coverage was better. This suggests that random forest imputation may be useful for imputing complex epidemiologic data sets in which some patients have missing data.
PMCID: PMC3939843  PMID: 24589914
angina, stable; imputation; missing data; missingness at random; regression trees; simulation; survival
15.  Maternal Smoking, Alcohol Consumption, and Caffeine Consumption During Pregnancy in Relation to a Son's Risk of Persistent Cryptorchidism: A Prospective Study in the Child Health and Development Studies Cohort, 1959–1967 
American journal of epidemiology  2007;167(3):10.1093/aje/kwm311.
The Child Health and Development Studies is a ≥40-year follow-up of 20,754 pregnancies occurring between 1959 and 1967 in California. There were 84 cases of undescended testes at birth persisting to at least age 2 years among 7,574 liveborn sons whose mothers were interviewed in early pregnancy. Cases were matched to three controls on birth year and race. Compared with mothers of controls, mothers of cryptorchid boys consumed more caffeine during pregnancy (odds ratio = 1.4, 95% confidence interval: 1.1, 1.9 for an interquartile range equivalent to three cups of coffee per day) but were not more likely to smoke or drink alcohol when all behaviors were considered together. Other maternal and perinatal risk factors were not significantly associated with persistent cryptorchidism and did not confound the association with caffeine.
PMCID: PMC3864885  PMID: 18024986
alcohol drinking; caffeine; cryptorchidism; pregnancy; prospective studies; risk factors; smoking
16.  Timing of Menarche and First Full-Term Birth in Relation to Breast Cancer Risk 
American journal of epidemiology  2007;167(2):230-239.
Ages at menarche and first birth are established risk factors for breast cancer. The interval between these ages may also affect risk, since the breast is more susceptible to carcinogenic insults during this period than during the parous period. However, few investigators have studied this relation. Using logistic regression, the authors evaluated associations between the timing of reproductive events and breast cancer risk among 4,013 cases and 4,069 controls enrolled in a multicenter, population-based US case-control study of White and African-American women (1994–1998). For White, parous premenopausal and postmenopausal women, those who had an interval of ≥16 years between the ages of menarche and first birth had 1.5-fold (95% confidence interval (CI): 1.0, 2.2) and 1.4-fold (95% CI: 1.1, 1.8) increased risks of breast cancer, respectively, in comparison with those who had ≤5 years between these ages. Adjusting for age at first birth altered these risk estimates somewhat, to odds ratios of 1.5 (95% CI: 0.8, 2.9) and 1.0 (95% CI: 0.6, 1.5), respectively. These associations were stronger for lobular and hormone-receptor-positive tumors but were absent among premenopausal African-American women. The authors conclude that the interval between age at menarche and age at first birth is associated with the risk of hormonally sensitive types of breast cancer, particularly among White women.
PMCID: PMC3804121  PMID: 17965112
breast neoplasms; histology; menarche; menopause; pregnancy; premenopause; receptors, estrogen; receptors, progesterone
17.  International physical activity comparisons: where to go from here? 
American journal of epidemiology  2010;171(10):1065-1068.
PMCID: PMC3696728  PMID: 20406761
18.  Relation between Psychiatric Syndromes and Behaviorally Defined Sexual Orientation in a Sample of the US Population 
American journal of epidemiology  2000;151(5):516-523.
Most surveys of the prevalence of psychiatric disorders among lesbians and gay men find no increased risk in comparison with heterosexuals. However, the majority of this work has relied on convenience samples drawn from the visible lesbian and gay community. The authors examined differences in 1-year prevalence of six psychiatric syndromes among sexually active individuals in the 1996 National Household Survey of Drug Abuse who reported either exclusive heterosexuality (n = 9,714) or having any same-gender sex partners (n = 194) in the prior year. Although nearly three quarters of homosexually active individuals did not meet criteria for any of the six syndromes assessed, in multivariate logistic regression analyses, homosexually active men were more likely than other men to evidence major depression and panic attack syndromes. In contrast, homosexually active women were more likely than other women to be classified with alcohol or drug dependency syndromes. Both men and women reporting any same-gender sex partners were more likely than others to have used mental health services in the year prior to interview. These findings suggest a small increased risk among homosexually active populations in 1-year psychiatric morbidity and use of mental health care services.
PMCID: PMC3698226  PMID: 10707921
health surveys; homosexuality; mental disorders; psychiatry; substance-related disorders
19.  Physical activity over the life course: whose behavior changes, when and why? 
American journal of epidemiology  2009;170(9):1078-1083.
Physical activity tends to decline from childhood into adulthood. The maintenance of high levels of physical activity throughout life is therefore an important public health objective. Relatively little is known about changes in physical activity behavior over the life course, the domains of physical activity in which they occur, the characteristics of those whose physical activity declines and the factors associated with such changes. In future, the incorporation of more accurate measures of physical activity in large population studies would help to establish more accurate estimates of associations in this area. Determinants of behavior change, including the effects of socioeconomic position and social mobility on physical activity and fitness, are likely to change constantly throughout life, but it is largely unknown which determinants are most important at each life stage, let alone whether and at what times those determinants change. Better evidence on determinants of behavior change throughout the life course would contribute greatly to our understanding of when and how to intervene in order to help create and sustain lifelong healthy behavior patterns in those who have most to gain from adopting them.
PMCID: PMC3672467  PMID: 19767350
Physical activity; determinants; behavior change; transition
20.  Coronary Artery Calcification in Japanese Men in Japan and Hawaii 
American journal of epidemiology  2007;166(11):1280-1287.
Explanations for the low prevalence of atherosclerosis in Japan versus United States are often confounded with genetic variation. To help remove such confounding, coronary artery calcification (CAC), a marker of subclinical atherosclerosis, was compared between Japanese men in Japan and Japanese men in Hawaii. Findings are based on risk factor and CAC measurements that were made from 2001 to 2005 in 311 men in Japan and 300 men in Hawaii. Men were aged 40 to 50 years and without cardiovascular disease. After age-adjustment, there was a 3-fold excess in the odds of prevalent CAC scores ≥10 in Hawaii versus Japan (relative odds [RO] = 3.2; 95% confidence interval [CI] = 2.1,4.9). While men in Hawaii had a generally poorer risk factor profile, men in Japan were 4-times more likely to smoke cigarettes (49.5 vs. 12.7%, p<0.001). In spite of marked risk factor differences between the samples, none of the risk factors provided an explanation for the low amounts of CAC in Japan. After risk factor adjustment, the RO of CAC scores ≥10 in Hawaii versus Japan was 4.0 (95% CI = 2.2,7.4). Further studies are needed to identify factors that offer protection against atherosclerosis in Japanese men in Japan.
PMCID: PMC3660555  PMID: 17728270
Atherosclerosis; cohort studies; coronary disease; Japan; men; risk factors
21.  Less subclinical atherosclerosis in Japanese men in Japan than in white men in the United States in the post World-War-II birth cohort 
American journal of epidemiology  2007;165(6):617-624.
Coronary heart disease (CHD) incidence and mortality remain very low in Japan despite major dietary changes and increases in CHD risk factors that should have resulted in substantial increase in CHD rates (Japanese paradox). Primary genetic effects are unlikely, given the substantial increase in CHD in migrant Japanese to the U.S. For men aged 40–49, levels of total cholesterol and blood pressure have been similar in Japan and the U.S. throughout their lifetime. The authors tested the hypothesis that levels of subclinical atherosclerosis, coronary artery calcification and intima-media thickness of the carotid artery (IMT), in men aged 40–49 are similar in Japan and the U.S. The authors conducted a population-based study of 493 randomly-selected men: 250 men in Kusatsu, Shiga, Japan, and 243 white men in Allegheny County, Pennsylvania, U.S. in 2002–2005. The Japanese had a less favorable profile of many risk factors than the whites. Prevalence ratio for the presence of coronary calcium score ≥10 in the Japanese compared to the whites was 0.52 (95% CI, 0.35, 0.76). Mean (SE) IMT was significantly lower in the Japanese (0.616 (0.005) versus 0.672 (0.005) mm, p<0.01). Both associations remained significant after adjusting for risk factors. The findings warrant further investigations.
PMCID: PMC3660737  PMID: 17244636
Atherosclerosis; epidemiology; men; risk factors
22.  [No title available] 
PMCID: PMC3636775  PMID: 15972941
23.  Telomere Length and Mortality: A Study of Leukocytes in Elderly Danish Twins 
American journal of epidemiology  2008;167(7):799-806.
Leukocyte telomere length, representing the mean length of all telomeres in leukocytes, is ostensibly a bioindicator of human aging. The authors hypothesized that shorter telomeres might forecast imminent mortality in elderly people better than leukocyte telomere length. They performed mortality analysis in 548 same-sex Danish twins (274 pairs) aged 73–94 years, of whom 204 pairs experienced the death of one or both co-twins during 9–10 years of follow-up (1997–2007). From the terminal restriction fragment length (TRFL) distribution, the authors obtained the mean TRFL (mTRFL) and the mean values of the shorter 50% (mTRFL50) and shortest 25% (mTRFL25) of TRFLs in the distribution and computed the mode of TRFL (MTRFL). They analyzed the proportions of twin pairs in which the co-twin with the shorter telomeres died first. The proportions derived from the intrapair comparisons indicated that the shorter telomeres predicted the death of the first co-twin better than the mTRFL did (mTRFL: 0.56, 95% confidence interval (CI): 0.49, 0.63; mTRFL50: 0.59, 95% CI: 0.52, 0.66; mTRFL25: 0.59, 95% CI: 0.52, 0.66; MTRFL: 0.60, 95% CI: 0.53, 0.67). The telomere-mortality association was stronger in years 3–4 than in the rest of the follow-up period, and it grew stronger with increasing intrapair difference in all telomere parameters. Leukocyte telomere dynamics might help explain the boundaries of the human life span.
PMCID: PMC3631778  PMID: 18270372
aged; leukocytes; mortality; survival analysis; telomere; twins
24.  Invited Commentary: Lipoproteins and Dementia—Is It the Apolipoprotein A-I? 
American journal of epidemiology  2007;165(9):993-997.
Because of the aging of the population, dementia has become a major public health problem. There has been growing evidence for a possible association between lipids and dementia. A large body of literature has demonstrated multiple hypothesized biologic links between lipids and neurodegenerative or other biologic pathways connected to dementing processes. However, the epidemiologic associations have been conflicting: dyslipidemia at middle age, but not in later life, seems to be associated with higher dementia risk in some but not all studies. Results from the Honolulu-Asia Aging Study reported by Saczynski et al. (Am J Epidemiol 2007;165:000–00) suggest that lipoprotein constituents, such as apolipoprotein A-I, a major component of the high density lipoprotein, may be more informative in enlightening the association between lipids and dementia. In this commentary, the epidemiology and biology of apolipoprotein A-I in relation to dementia is reviewed.
PMCID: PMC3627057  PMID: 17298956
Alzheimer disease; apolipoproteins; dementia; lipids
25.  Misuse of the linear mixed model when evaluating risk factors of cognitive decline 
American Journal of Epidemiology  2011;174(9):1077-1088.
The linear mixed model (LMM), routinely used to describe change over time of outcomes and association with risk factors, assumes that a unit change in any predictor is associated with a constant change in the outcome. When used on psychometric tests, this assumption may not hold. Indeed, psychometric tests usually suffer from ceiling and/or floor effects, and curvilinearity (i.e. varying sensitivity to change).
This work aimed at determining the consequences of such a misspecification when evaluating predictors of cognitive decline. As an alternative to the LMM, two mixed models based on latent processes that handle discrete and bounded outcomes were considered. Models differences were illustrated using four psychometric tests from the cohort PAQUID. Type I error of the Wald test for risk factors regression parameters were then formally assessed in a simulation study. It demonstrated that type I errors in the LMM could be dramatically inflated for some tests so that spurious associations with risk factors were found. In particular confusion between effects on mean level and on change over time was highlighted. The authors thus recommend the use of the alternative mixed models when studying psychometric tests and more generally quantitative scales (quality of life, activities of daily living).
PMCID: PMC3551607  PMID: 21965187
Aged; Cognition Disorders; epidemiology; etiology; Dementia; epidemiology; etiology; Educational Status; France; epidemiology; Humans; Linear Models; Longitudinal Studies; Models, Statistical; Neuropsychological Tests; Psychometrics; Risk Factors; biostatistics; cognition; longitudinal studies; psychometrics; risk factors; statistical models

Results 1-25 (157)