Sleeper cell
Higher-Status Occupations and Breast Cancer: A Life-Course Stress Approach
Tetyana Pudrovska et al.
Social Science & Medicine, forthcoming
Abstract:
Using the 1957-2011 data from 3,682 White non-Hispanic women (297 incident breast cancer cases) in the Wisconsin Longitudinal Study, United States, we explore the effect of occupation in 1975 (at age 36) on breast cancer incidence up to age 72. Our study is motivated by the paradoxical association between higher-status occupations and elevated breast cancer risk, which presents a challenge to the consistent health advantage of higher social class. We found that women in professional occupations had 72%-122% and women in managerial occupations had 57%-89% higher risk of a breast cancer diagnosis than housewives and women in lower-status occupations. We explored an estrogen-related pathway (reproductive history, health behaviors, and life-course estrogen cycle) as well as a social stress pathway (occupational experiences) as potential explanations for the effect of higher-status occupations. The elevated risk of breast cancer among professional women was partly explained by estrogen-related variables but remained large and statistically significant. The association between managerial occupations and breast cancer incidence was fully explained by job authority defined as control over others' work. Exercising job authority was related to higher breast cancer risk (HR = 1.57, 95% CI: 1.12, 2.18), especially with longer duration of holding the professional/managerial job. We suggest that the assertion of job authority by women in the 1970s involved stressful interpersonal experiences that may have promoted breast cancer development via prolonged dysregulation of the glucocorticoid system and exposure of the breast tissue to adverse effects of chronically elevated cortisol. Our study emphasizes complex biosocial pathways through which women's gendered occupational experiences become embodied and drive forward physiological repercussions.
----------------------
Sleeping at the Limits: The Changing Prevalence of Short and Long Sleep Durations in 10 Countries
Yu Sun Bin, Nathaniel Marshall & Nick Glozier
American Journal of Epidemiology, 15 April 2013, Pages 826-833
Abstract:
Short (≤6 hours) and long (>9 hours) sleep durations are risk factors for mortality and morbidity. To investigate whether the prevalences of short and long sleep durations have increased from the 1970s to the 2000s, we analyzed data from repeated cross-sectional surveys of 10 industrialized countries (38 nationally representative time-use surveys; n = 328,018 adults). Logistic regression models for each country were used to determine changes in the prevalence of short and long sleep durations over time, controlling for sampling differences in gender, age, number of weekend days included, and season of data collection. Over the periods covered by data, the prevalence of short sleep duration increased in Italy (adjusted odds ratio = 2.64, 95% confidence interval (CI): 2.41, 2.89) and Norway (adjusted odds ratio = 2.33, 95% CI: 1.77, 3.08) but decreased in Sweden, the United Kingdom, and the United States. The prevalence of long sleep duration increased in Australia (adjusted odds ratio = 1.14, 95% CI: 1.05, 1.25), Finland (adjusted odds ratio = 1.30, 95% CI: 1.14, 1.48), Sweden (adjusted odds ratio = 1.51, 95% CI: 1.35, 1.69), the United Kingdom (adjusted odds ratio = 2.03, 95% CI: 1.68, 2.46), and the United States (adjusted odds ratio = 1.50, 95% CI: 1.36, 1.65) but decreased in Canada and Italy. No changes were observed in Germany or the Netherlands. Limited increases in short sleep duration challenge the claim of increasingly sleep-deprived societies. Long sleep duration is more widespread than is short sleep duration. It has become more prevalent and thus should not be overlooked as a potential contributor to ill health.
----------------------
Richard Shaw & Kate Pickett
American Journal of Public Health, June 2013, Pages 1052-1057
Objectives: In the United States, Hispanic mothers have birth outcomes comparable to those of White mothers despite lower socioeconomic status. The contextual effects of Hispanic neighborhoods may partially explain this "Hispanic paradox." We investigated whether this benefit extends to other ethnic groups.
Methods: We used multilevel logistic regression to investigate whether the county-level percentage of Hispanic residents is associated with infant mortality, low birth weight, preterm delivery, and smoking during pregnancy in 581 151 Black and 2 274 247 White non-Hispanic mothers from the US Linked Birth and Infant Death Data Set, 2000.
Results: For White and Black mothers, relative to living in counties with 0.00%-0.99% of Hispanic residents, living in counties with 50.00% or more of Hispanic residents was associated with an 80.00% reduction in the odds of smoking, an infant mortality reduction of approximately one third, and a modest reduction in the risks of preterm delivery and low birth weight.
Conclusions: The health benefits of living in Hispanic areas appear to bridge ethnic divides, resulting in better birth outcomes even for those of non-Hispanic origin.
----------------------
Texting While Driving and Other Risky Motor Vehicle Behaviors Among US High School Students
Emily O'Malley Olsen, Ruth Shults & Danice Eaton
Pediatrics, forthcoming
Objective: To assess the prevalence of texting/e-mailing while driving (TWD) and association of TWD with other risky motor vehicle (MV) behaviors among US high school students.
Methods: Data were used from the Centers for Disease Control and Prevention's 2011 national Youth Risk Behavior Survey, which assessed TWD during the 30 days before the survey among 8505 students aged ≥16 years from a nationally representative sample of US high school students. TWD frequency was coded into dichotomous and polychotomous variables. Logistic regression assessed the relationship between TWD and other risky driving behaviors, controlling for age, race/ethnicity, and sex.
Results: The prevalence of TWD on ≥1 days during the 30 days before the survey was 44.5% (95% confidence interval: 40.8%-48.2%). Students who engaged in TWD were more likely than their non-TWD counterparts to not always wear their seatbelt (prevalence ratio; 95% confidence interval: 1.16; 1.07-1.26), ride with a driver who had been drinking alcohol (1.74; 1.57-1.93), and drink alcohol and drive (5.33; 4.32-6.59). These other risky MV behaviors were most likely to occur among students who frequently engaged in TWD.
Conclusions: Nearly half of US high school students aged ≥16 years report TWD during the past 30 days; these students are more likely to engage in additional risky MV behaviors. This suggests there is a subgroup of students who may place themselves, their passengers, and others on the road at elevated risk for a crash-related injury or fatality by engaging in multiple risky MV behaviors.
----------------------
Cellphone bans and fatal motor vehicle crash rates in the United States
Siew Hoon Lim & Junwook Chi
Journal of Public Health Policy, May 2013, Pages 197-212
Abstract:
A number of states in the United States have laws restricting drivers from using cellphones. Using state-level panel data, we examined the effect of cellphone laws on fatal crashes in the United States between 2000 and 2010. Our results show that there is insufficient power to detect a reduction in overall fatal crash rates scaled by vehicle miles and population estimates. Cellphone bans, however, have significantly reduced the fatal crash rates of drivers in certain age cohorts. The effect was most pronounced among drivers between 18 and 34 years of age. We did not find any significant effect among drivers in the 55 and older age cohorts.
----------------------
Car Ratings Take a Back Seat to Vehicle Type: Outcomes of SUV vs. Passenger Car Crashes
Dietrich Jehle et al.
University at Buffalo Working Paper, May 2013
Background: Car safety ratings are routinely utilized in making automobile purchase decisions. These one to five star ratings are based on crash test data comparing vehicles of similar type, size and weight.
Objectives: In this study, we hypothesized that car safety ratings are less important than other vehicle factors in predicting outcomes of head-on crashes between SUVs and standard passenger vehicles.
Methods: A retrospective study was conducted on the drivers in severe head-on motor vehicle crashes entered into the FARS (Fatality Analysis Reporting System) database between 1995 and 2010. This database includes all motor vehicle crashes in United States that resulted in a death within 30 days. Only two car head-on crashes between a highly (safety) rated passenger car and a more poorly rated SUV were included in the initial analysis. Outcomes of all SUV vs. passenger car and passenger car vs. passenger car head-on crashes were then compared by safety rating. In all analyses, crashes were excluded if either vehicle was older than model year 1995, incomplete information was available on specific vehicle types, or vehicle crash ratings were not available for both vehicles. These paired crash results were entered into a univariate logistic regression model with driver death as the outcome of interest.
Results: There were 83,251 vehicles of any type involved in head-on crashes in the database. In head-on crashes where the front driver passenger car crash rating was superior to the SUV's front driver crash rating, the odds of death was 4.03 times higher for the driver of the passenger car (95% CI: 3.04 - 5.35). Ignoring crash ratings, the odds of death was 7.59 times higher for the car driver than the SUV driver in all head-on crashes (95% CI: 6.75 - 8.52). In passenger vs. passenger car head-on crashes, a lower car safety rating was associated with a 1.47 times higher odds of death (95% CI: 1.26-1.71).
Conclusion: Vehicle type (passenger vehicle vs. SUV) is a much more important predictor of death than crash safety ratings in SUV vs. passenger vehicle head-on crashes. The increased safety of SUVs in head-on crashes with passenger vehicles should be taken into consideration when purchasing a car.
----------------------
Sepideh Modrek & Mark Cullen
Social Science & Medicine, forthcoming
Abstract:
While the negative effects of unemployment have been well studied, the consequences of layoffs and downsizing for those who remain employed are less well understood. This study used human resources and health claimsdata from a large multi-site fully insured aluminum company to explore the health consequences of downsizing on the remaining workforce. We exploit the variation in the timing and intensity of layoff to categorize 30 plants as high or low layoff plants. Next, we select a stably employed cohort of workers with history of health insurance going back to 2006 to 1) describe the selection process into layoff and 2) explore the association between the severity of plant level layoffs and the incidence of four chronic conditions in the remaining workforce. We examined four health outcomes: incident hypertension, diabetes, asthma/COPD and depression for a cohort of approximately 13,000 employees. Results suggest that there was an increased risk of developing hypertension for workers that remain at the plants with the highest level of layoffs, and increased risk of developing diabetes for salaried workers that remain at the plants with the highest level of layoffs. The hypertension results were robust to a several specification tests. In addition, the study design selected only healthy workers, therefore our results are likely to be a lower bound and suggest that adverse health consequences of the current recession may affect a broader proportion of the population than previously expected.
----------------------
Richard Steckel
American Journal of Human Biology, forthcoming
Objectives: The paper tests the thrifty phenotype hypothesis, according to which nonharmonious growth trajectories are costly for adult health.
Methods: The American surge in the prevalence of type 2 diabetes is concentrated in the South, a region characterized by a long history of poverty followed by rapid economic growth beginning in the 1960s. Civil rights legislation further accelerated income growth for African-Americans in the region. The paper investigates the hypothesis by using per capita income at the state level as a proxy for net nutritional conditions.
Results: Regressions at the state level explain 56% of the variation in the prevalence rate of type 2 diabetes in 2009 using two explanatory variables: the ratio of per capita income in 1980 to that in 1950 and the share of the population that was African-American. The paper discusses ways that rapid economic growth may have translated into weight gain and type 2 diabetes.
Conclusions: If the thrifty phenotype hypothesis is correct, future rates in the prevalence of type 2 diabetes are predictable based on income history. The forecast for rapidly developing countries such as India and China are ominous.
----------------------
For a Sociology of Expertise: The Social Origins of the Autism Epidemic
Gil Eyal
American Journal of Sociology, January 2013, Pages 863-907
Abstract:
This article endeavors to replace the sociology of professions with the more comprehensive and timely sociology of expertise. It suggests that we need to distinguish between experts and expertise as requiring two distinct modes of analysis that are not reducible to one another. It analyzes expertise as a network linking together agents, devices, concepts, and institutional and spatial arrangements. It also suggests rethinking how abstraction and power were analyzed in the sociology of professions. The utility of this approach is demonstrated by using it to explain the recent precipitous rise in autism diagnoses. This article shows that autism remained a rare disorder until the deinstitutionalization of mental retardation created a new institutional matrix within which a new set of actors - the parents of children with autism in alliance with psychologists and therapists - were able to forge an alternative network of expertise.
----------------------
Aggregation and The Estimated Effects of Local Economic Conditions on Health
Jason Lindo
NBER Working Paper, May 2013
Abstract:
This paper considers the relationship between local economic conditions and health with a focus on different approaches to geographic aggregation. After reviewing the tradeoffs associated with more- and less-disaggregated analyses - including an investigation of the migratory response to changing economic conditions - I update earlier state-level analyses of mortality and infant health and then consider how the estimated effects vary when the analysis is conducted at differing levels of geographic aggregation. This analysis reveals that more-disaggregated analyses severely understate the extent to which downturns are associated with improved health. Further investigation reveals that county economic conditions have an independent effect on mortality but that state and regional economic conditions are stronger predictors. I also leverage county-level data to explore heterogeneity in the link between county economic conditions and health across states, demonstrating that local downturns lead to the greatest improvements in health in low-income states.
----------------------
Childhood Infections and Adult Height in Monozygotic Twin Pairs
Amie Hwang et al.
American Journal of Epidemiology, forthcoming
Abstract:
Adult height is determined by genetics and childhood nutrition, but childhood infections may also play a role. Monozygotic twins are genetically matched and offer an advantage when identifying environmental determinants. In 2005-2007, we examined the association of childhood infections with adult height in 140 height-discordant monozygotic twin pairs from the California Twin Program. To obtain information on childhood infections and growth, we interviewed the mothers of monozygotic twins who differed in self-reported adult height by at least 1-inch (2.5 cm). Within-pair differences in the relative frequency of childhood infections were highly correlated, especially within age groups. A conditional logistic regression analysis demonstrated that more reported episodes of febrile illness occurred in the twin with shorter stature (odds ratio = 2.00, 95% confidence interval: 1.18, 3.40). The association was strongest for differences in the relative frequency of infection during the toddler years (ages 1-5: odds ratio = 3.34, 95% confidence interval: 1.47, 7.59) and was similar when restricted to twin pairs of equal birth length. The association was not explained by differential nutritional status. Measures of childhood infection were associated with height difference in monozygotic twin pairs, independent of genome, birth length, and available measures of diet.
----------------------
Blood Cholesterol Trends 2001-2011 in the United States: Analysis of 105 Million Patient Records
Harvey Kaufman et al.
PLoS ONE, May 2013
Objectives: We report annual trends in low density lipoprotein cholesterol (LDL-C) from an in-care patient population of nearly 105 million adults across the United States (U.S.), from 2001 through 2011.
Background: Average blood cholesterol values have declined in the U.S. since at least 1960. The National Health and Nutrition Examination Survey (NHANES) reported declining blood cholesterol values from 1999 through 2010. In the absence of more recent published data, we examined LDL-C values from a single clinical laboratory database to determine whether these values continued to decline through 2011.
Methods and Results: We extracted almost 247 million LDL-C results from nearly 105 million adults who received diagnostic testing from a single national clinical laboratory. Annual age-adjusted mean LDL-C values were calculated, and analyzed by gender. Piecewise regression analysis of the total study population indicates a breakpoint, or change in slope, in the years following 2008 (F = 163.13; p<0.05). Between 2001 and 2008, the average rate of annual decline was -2.05 mg/dL (95% CI [-2.35, -1.75]). After 2008, mean LDL-C levels flattened out, with a slope not statistically different from zero (slope = -0.10 mg/dL/year; 95% CI [-1.46, 1.26]). This stabilization was observed in both genders and all age ranges, and was also reflected in the percentage of results in low- and high-risk categories.
Conclusions: The trends reported suggest historical progress in decreasing LDL-C levels, observed from 2001-2008, may have stalled in recent years. Further research is needed to determine the cause of the observed trends and develop new strategies to reduce lipid-based cardiovascular risk further.
----------------------
Amar Hamoudi & Jennifer Beam Dowd
American Journal of Public Health, June 2013, Pages 1039-1045
Objectives: We examined the impact of the dramatic increases in housing prices in the United States in the 1990s and early 2000s on physical health outcomes among a representative sample of middle-aged and older Americans.
Methods: Using a quasi-experimental design, we exploited geographic and time variation in housing prices using third-party valuation estimates of median single-family detached houses from 1988 to 2007 in each of 2400 zip codes combined with Health and Retirement Study data from 1992 to 2006 to test the impact of housing appreciation on physical health outcomes.
Results: Respondents living in communities in which home values appreciated more rapidly had fewer functional limitations, performed better on interviewer-administered physical tasks, and had smaller waist circumference.
Conclusions: Our results indicate that increases in housing wealth were associated with better health outcomes for homeowners in late middle age and older. The recent sharp decline in housing values for this group may likewise be expected to have important implications for health and should be examined as data become available.
----------------------
Disability, Earnings, Income and Consumption
Bruce Meyer & Wallace Mok
NBER Working Paper, March 2013
Abstract:
Using longitudinal data for 1968-2009 for male household heads, we determine the prevalence of pre-retirement age disability and its association with a wide range of outcomes, including earnings, income, and consumption. We then employ some of these quantities in the optimal social insurance framework of Chetty (2006) to study current compensation for the disabled. Six of our findings stand out. First, disability rates are high. We divide the disabled along two dimensions based on the persistence and severity of their work-limiting condition. We estimate that a person reaching age 50 has a 36 percent chance of having been disabled at least temporarily once during his working years, and a 9 percent chance that he has begun a chronic and severe disability. Second, the economic consequences of disability are frequently profound. Ten years after disability onset, a person with a chronic and severe disability on average experiences a 79 percent decline in earnings, a 35 percent decline in after-tax income, a 24 percent decline in food and housing consumption and a 22 percent decline in food consumption. Third, economic circumstances differ sharply across disability groups. The outcome decline for the chronically and severely disabled is often more than twice as large as that for the average disabled head. Fourth, our findings show the partial and incomplete roles that individual savings, family support and social insurance play in reducing the consumption drop that follows disability. Fifth, time use and detailed consumption data further indicate that disability is associated with a decline in well-being. Sixth, using the quantities we have estimated, we provide the range of behavioral elasticities and preference parameters consistent with current disability compensation being optimal within the Chetty framework.
----------------------
Predictors of health behaviors after the economic downturn: A longitudinal study
Jonathan Macy, Laurie Chassin & Clark Presson
Social Science & Medicine, July 2013, Pages 8-15
Abstract:
Economic declines and their associated stress, shortage of financial resources, and changes in available time can impair health behaviors. This study tested the association between change in working hours, change in employment status, and financial strain and health behaviors measured after the 2008 recession after controlling for pre-recession levels of the health behaviors. The moderating influences of demographic factors and pre-recession levels of the health behaviors on the association between change in working hours and employment status and financial strain and the health behaviors were also tested. Participants (N = 3984) were from a longitudinal study of a U.S. Midwestern community-based sample. Regression analyses tested the unique relations between change in hours worked per week, change in employment status, and financial strain and five health behaviors over and above demographic factors and pre-recession levels of the same behavior. Models included predictor by covariate interactions. Participants who reported higher levels of financial strain engaged in lower levels of all but one of the five health behaviors, but there were no significant main effects of a change in the number of hours worked per week or change in employment status. Significant interactions revealed moderation of these relations by demographic characteristics, but findings differed across health behaviors. Financial strain negatively affected engagement in multiple healthy behaviors. Promoting the maintenance of healthy behaviors for disease prevention is an important public health goal during times of economic decline.
----------------------
The Effect of Health on Earnings: Quasi-Experimental Evidence from Commuting Accidents
Martin Hallaa & Martina Zweimüller
Labour Economics, forthcoming
Abstract:
This paper interprets accidents occurring on the way to and from work as negative health shocks to identify the causal effect of health on labor market outcomes. We argue that in our sample of exactly matched injured and non-injured workers, these health shocks (predominantly impairments in the musculoskeletal system) are quasi-randomly assigned. A fixed-effects difference-in-differences approach estimates a negative and persistent effect on subsequent employment and earnings. After initial periods with ahigher incidence of sick leave, injured workers are more likely to be unemployed, and a growing share of them leave the labor market via disability retirement. Injured workers who manage to stay in employment incur persistent earnings losses. The effectsare somewhat stronger for sub-groups of workers who are typically less attached to the labor market.
----------------------
An Investigation into the Effect of Type I and Type II Diabetes Duration on Employment and Wages
Travis Minor
Economics & Human Biology, forthcoming
Abstract:
Using data from the National Longitudinal Survey of Youth 1979, the current study examines the effect of type I and type II diabetes on employment status and wages. The results suggest that both the probability of employment and wages are negatively related to the number of years since the initial diagnosis of diabetes. Moreover, the effect of diabetes duration on the probability of employment appears to be nonlinear, peaking around 16 years for females and 10 years for males. A similar negative effect on wages is found only in male diabetics. Finally, the results suggest that failure to distinguish between type I and type II diabetics may lead to some counterintuitive results.
----------------------
Family income and child health in the UK
Bénédicte Apouey & Pierre-Yves Geoffard
Journal of Health Economics, July 2013, Pages 715-727
Abstract:
Recent studies examining the relationship between family income and child health in the UK have produced mixed findings. We re-examine the income gradient in child general health and its evolution with child age in this country, using a very large sample of British children. We find that there is no correlation between income and child general health at ages 0-1, that the gradient emerges around age 2 and is constant from age 2 to age 17. In addition, we show that the gradient remains large and significant when we reduce the endogeneity of income. Furthermore, our results indicate that the gradient in general health reflects a greater prevalence of chronic conditions among low-income children and a greater severity of these conditions. Taken together, these findings suggest that income does matter for child health in the UK and may play a role in the intergenerational transmission of socioeconomic status.
----------------------
Martin Binder & Alex Coad
Social Science & Medicine, June 2013, Pages 155-167
Abstract:
Bad health decreases individuals' happiness, but few studies measure the impact of specific illnesses. We apply matching estimators to examine how changes in different (objective) conditions of bad health affect subjective well-being for a sample of 100,278 observations from the British Household Panel Survey (BHPS) database (1996-2006). The strongest effect is for alcohol and drug abuse, followed by anxiety, depression and other mental illnesses, stroke and cancer. Adaptation to health impairments varies across health impairments. There is also a puzzling asymmetry: strong adverse reactions to deteriorations in health appear alongside weak increases in well-being after health improvements. In conclusion, our analysis offers a more detailed account of how bad health influences happiness than accounts focusing on how bad self-assessed health affects individual well-being.
----------------------
Does More Schooling Improve Health Outcomes and Health Related Behaviors? Evidence from U.K. Twins
Vikesh Amin, Jere Behrman & Tim Spector
Economics of Education Review, forthcoming
Abstract:
Several recent studies using instrumental variables based on changes in compulsory school leaving age laws have estimated the causal effect of schooling on health outcomes and health-related behaviors in the U.K. Despite using the same identification strategy and similar datasets, no consensus has been reached. We contribute to the literature by providing results for the U.K. using a different research design and a different dataset. Specifically, we estimate the effect of schooling on health outcomes (obesity and physical health) and health-related behaviors (smoking, alcohol consumption and exercise) for women through within-MZ twins estimates using the TwinsUK database. For physical health, alcohol consumption and exercise, the within-MZ twins estimates are uninformative about whether there is a causal effect. However, we find (1) that the significant association between schooling and smoking status is due to unobserved endowments that are correlated with schooling and smoking (2) there is some indication that more schooling reduces the body mass index for women, even once these unobserved endowments have been controlled for.
----------------------
Income Inequalities in Unhealthy Life Styles in England and Spain
Joan Costa-Font, Cristina Hernández-Quevedo & Dolores Jiménez Rubio
Economics & Human Biology, forthcoming
Abstract:
Health inequalities in developed societies are persistent. Arguably, the rising inequalities in unhealthy lifestyles might underpin these inequality patterns, yet supportive empirical evidence is scarce. We examine the patterns of inequality in unhealthy lifestyles in England and Spain, two countries that exhibit rising obesity levels with a high prevalence of smoking and alcohol use. This study is unique in that it draws from health survey data spanning over a period in which major contextual and policy changes have taken place. We document persistent income-related inequalities in obesity and smoking; both unhealthy lifestyles appear to be disproportionately concentrated among the relatively poor in recent decades. In contrast, alcohol use appears to be concentrated among richer individuals in both periods and countries examined.
----------------------
Peter Muennig et al.
PLoS ONE, March 2013
Abstract:
We set out to examine the material, psychological, and sociological pathways mediating the income gradient in health and mortality. We used the 2008 General Social Survey-National Death Index dataset (N = 26,870), which contains three decades of social survey data in the US linked to thirty years of mortality follow-up. We grouped a large number of variables into 3 domains: material, psychological, and sociological using factor analysis. We then employed discrete-time hazard models to examine the extent to which these three domains mediated the income-mortality association among men and women. Overall, the gradient was weaker for females than for males. While psychological and material factors explained mortality hazards among females, hazards among males were explained only by social capital. Poor health significantly predicted both income and mortality, particularly among females, suggesting a strong role for reverse causation. We also find that many traditional associations between income and mortality are absent in this dataset, such as perceived social status.
----------------------
Markus Schafer
Social Science & Medicine, June 2013, Pages 176-184
Abstract:
Building from the premise that network ties influence why and how people seek health care, this study examines whether different types of close relations predict two distinct, but overlapping forms of care utilization. To that end, I examine the use of conventional care and complementary and alternative medicine (CAM). Analyses are conducted with a national sample of older American adults aged 57-85 in 2005/2006 (n = 3005). I find that partnered men who are very likely to discuss health with a partner had a greater number of physician visits in the past year, net of their health status and other relevant factors. Having children with whom health is likely to be discussed was also associated with more visits, as was the presence of non-kin ties. On the other hand, the use of complementary and alternative medicine was predicted not by spousal or other kin-based relationships, but only by having non-kin ties with whom a respondent could discuss health. Results suggest that understanding the relational undercurrents of care utilization requires attention to diverse forms of social relations and to diverse expressions of care participation.