Low Occupational Social Class Research Articles

The Political Sociology of Voting Participation in Southern Africa: A Multilevel Study of Regional and Social Class Predictors in 11 Countries

Voting participation constitutes and legitimatizes electoral democracy. However, research has been unable to predict and explain the lack of participation across political contexts. The present study aims to predict voting participation based on regional predictors using 11 countries from the Afrobarometer (2019). For the analysis, the study uses modern measurement models and Bayesian multilevel logit models (binary and categorical). The regional analysis suggests that a greater frequency of experiencing bribery predicts lower voting probabilities. Regions with citizens who are more highly educated predict lower voting probabilities for deciding not to vote. Regions with higher levels of skilled craft workers predict lower voting probabilities. Compared to those in the lower occupational social class, citizens in the higher occupational social class have a lower voting probability. The study concludes by offering implications for the relational approach to contentious politics and democratization.

Persistent social inequality in low life satisfaction among adolescents in Denmark 2002–2018

PurposeThe aims were to examine social inequality in low life satisfaction among Danish adolescents and whether the social inequality changes over time. MethodsThe study population was 20,328 11-, 13- and 15-year-olds from the Danish Health Behaviour in School-aged Children (HBSC) study 2002, 2006, 2010, 2014 and 2018 with data about life satisfaction and occupational social class (OSC). Low life satisfaction was defined as values under 6 on the Cantril Ladder. ResultsThe proportion with low life satisfaction was 10.5% in high, 11.7% in medium and 14.7% in low OSC (chi2-test, p < 0.0001). The OR (95% CI) for low life satisfaction was 1.25 (1.13–1.39) in medium and 1.62 (1.43–1.83) in low compared to high OSC. There was no upwards or downwards trend in low life satisfaction over time and the absolute and relative social inequality in low life satisfaction did not change from 2002 to 2018. ConclusionThe study shows that the political ambition on reducing social inequality in health and wellbeing has not been met in relation to low life satisfaction among Danish adolescents.

Trends in social inequality in overweight and obesity among adolescents in Denmark 1998-2018.

The aim was to analyze trends in overweight and obesity in relation to socioeconomic position among Danish adolescents in the 20-year period 1998-2018. The study used data on self-reported height and weight and parents' occupational social class (OSC) from 11-, 13- and 15-year-old schoolchildren in 1998, 2002, 2006, 2010, 2014 and 2018, n = 22,177. The analyses included absolute social inequality in overweight/obesity (prevalence difference between low and high OSC) and relative social inequality (OR for overweight/obesity). In the total sample, the prevalence of overweight and obesity was 9.7% and 1.4%, respectively, with significantly higher prevalence in low than high OSC. There were significantly increasing trends in both overweight and obesity 1998-2018 in low OSC and no significant increase in high OSC. The OR for overweight was 1.59 (1.42-1.74) in middle and 2.16 (1.89-2.46) in low OSC, OR for obesity 1.74 (1.29-2.34) in middle and 2.97 (2.15-4.11) in low OSC. Associations were not modified by survey year. There was a persistent absolute and relative social inequality in overweight and obesity 1998-2018 among Danish adolescents.

Intake of sugar sweetened soft drinks among adolescents: Trends and social inequality in Denmark 2002-2018.

Intake of sugar sweetened soft drinks (SSSD) has decreased among adolescents, but trends in social inequality in SSSD intake are unknown. Examine trends in social inequality in SSSD intake among adolescents in Denmark during 2002-2018. Five Health Behaviour in School-aged Children surveys with data on SSSD intake and parents' occupational social class (OSC) from nationally representative samples of 11, 13 and 15 year olds, n =20,112. The overall prevalence of daily SSSD intake decreased from 10.1% in 2002 to 6.4% in 2018. The prevalence decreased in both high OSC (from 8% to 5%) and middle OSC (from 10% to 6%) but remained around 12% in low OSC. The odds ratio (OR) estimates of low compared with high OSC increased over the years around an overall OR of 2.01 (1.74-2.34). Danish adolescents' SSSD intake decreased during 2002-2018 and was higher the lower the parents' OSC. Thus, social inequality increased during 2002-2018.

Trends in socioeconomic differences in daily smoking among 15-year-old Danes 1991-2014.

Aim: To examine trends in socioeconomic differences in daily smoking among 15-year-old Danes between 1991-2014, using occupational social class as indicator of socioeconomic status. Methods: The study included 15-year-olds participating in seven Danish Health Behaviour in School-aged Children studies between 1991-2014, n = 8,641. The analyses focused on absolute socioeconomic differences (prevalence difference between low and high occupational social class) and relative socioeconomic differences communicated by odds ratio for daily smoking. Results: The prevalence of daily smoking declined from 18.6% in 1991 to 4.5% in 2014. Across all surveys, the prevalence was 8.9% in high, 12.8% in middle and 16.5% in low occupational social classes (p < 0.0001). The absolute socioeconomic differences increased from 1991 to 2006 and declined thereafter. Across all survey years, the odds ratio (95% confidence interval) for daily smoking was 1.40 (1.19-1.65) in middle and 1.90 (1.56-2.32) in low versus high occupational social classes. The statistical interaction between occupational social class and survey year was significant (p = 0.0404), suggesting increasing relative socioeconomic differences from 1991 to 2014. Conclusions: There was a substantial decline in daily smoking among 15-year-olds between 1991-2014 in all occupational social class groups. The prevalence of daily smoking was highest in the low occupational social class during the entire period. The absolute socioeconomic differences in daily smoking increased between 1991-2006 and declined thereafter. The relative socioeconomic differences increased over 1991-2014. Studies of change in socioeconomic differences over time should address both absolute and relative socioeconomic differences as they may result in different conclusions and because important improvement in prevalence patterns may be disguised by exclusive focus on changes in relative socioeconomic differences.

Short report: persistent social inequality in poor self-rated health among adolescents in Denmark 1991-2014.

The aim was to examine trends in social inequality in poor self-rated health (SRH) among adolescents in Denmark 1991-2014. The analysis included 18 996 11-15-year-old school children from the cross-sectional Health Behaviour in School-aged Children studies in 1991, 2002, 2006, 2010 and 2014. Across the five surveys, the prevalence of poor SRH was 14.2%, remaining almost unchanged from 1991 to 2014. The proportion with poor SRH was 12.2% in high, 14.3% in middle and 17.6% in low occupational social class. This social inequality in poor SRH was persistent during the entire study period, both in terms of absolute and relative social inequality.

Trends in social inequality in loneliness among adolescents 1991-2014.

Loneliness and social inequality in health are important public health concerns. We examined (i) trends in loneliness among adolescents from 1991 to 2014 in Denmark and (ii) trends in social inequality in loneliness. Study population: 11-15-year olds in random samples of schools in 1991, 1994, 1998, 2006 and 2014, n = 19 096. Loneliness was measured by a single item and social background by parents' occupational social class (OSC). We calculated absolute (%) differences in loneliness between high and low OSC and relative differences by odds ratio for loneliness. Across all surveys, 6.3% reported feeling lonely. The prevalence increased from 4.4% in 1991 to 7.2% in 2014. The prevalence of loneliness in high, middle and low OSC was 5.8, 5.9 and 8.0%. The increase in loneliness was more pronounced in higher than lower OSC, resulting in a decreasing absolute social inequality in loneliness. The statistical interaction between OSC and survey year was significant, P = 0.0176, i.e. the relative social inequality in loneliness also decreased from 1991 to 2014. The prevalence of loneliness increased from 1991 to 2014. The social inequality in loneliness decreased in both absolute and relative terms because of a rising prevalence of loneliness among children from high OSC.

Persistent social inequality in medicine use for headache among adolescents in Denmark 1991-2014.

Socioeconomic variation in adolescents' medicine use behaviour is an understudied issue. To examine the association between socioeconomic background and medicine use for headache among adolescents, and how this association changes over time. Data stem from the Danish part of the international Health Behaviour in School-aged Children (HBSC) study 1991-2014 with data about parents' occupational social class (OSC) and self-reported medicine use for headache, n = 26,685. This study examined absolute social inequality (percent difference between high and low OSC) and relative social inequality (odds ratio for medicine use by OSC). In total, 40.5% used medicine for headache in the past month. There was a significant increase from 32.3% in 1991 to 42.8% in 2002 (test for trend, p < 0.0001) and very little change 2002-2014. This pattern was similar in high, middle and low OSC. The prevalence of medicine use for headache in high, middle and low OSC was 36.2%, 41.5% and 44.8% (p < 0.0001). The OR (95% CI) for medicine use was 1.25 (1.18-1.324) in middle and 1.43 (1.33-1.54) in low OSC. Medicine use for headache increased 1991-2002 and remained stable 2002-2014. There was increasing medicine use for headache with decreasing OSC; this social inequality was persistent 1991-2014.

Persistent social inequality in low intake of vegetables among adolescents, 2002-2014.

To examine the trend in social inequality in low intake of vegetables among adolescents in Denmark from 2002 to 2014 using occupational social class (OSC) as socio-economic indicator. Repeated cross-sectional school surveys including four waves of data collection in 2002-2014. The analyses focused on absolute social inequality (difference between high and low OSC in low vegetable intake) as well as relative social inequality (OR for low vegetable intake by OSC). The nationally representative Health Behaviour in School-aged Children (HBSC) study in Denmark. The study population was 11-15-year olds (n 17 243). Low intake of vegetables was defined as less than weekly intake measured by food frequency items. OSC was measured by student reports of parents' occupation. The proportion of participants who reported eating vegetables less than once weekly was 8·9 %, with a notable decrease from 11·9 % in 2002 to 5·9 % in 2014. The OR (95 % CI) for less than weekly vegetable intake was 2·28 (1·98, 2·63) in the middle compared with high OSC and 3·12 (2·67, 3·66) in the low compared with high OSC. The absolute social inequality in low vegetable intake decreased from 2002 to 2014 but the relative social inequality remained unchanged. The study underscores that it is important to address socio-economic factors in future efforts to promote vegetable intake among adolescents. The statistical analyses of social inequality in vegetable intake demonstrate that it is important to address both absolute and relative social inequality as these two phenomena may develop differently.

Short communication: Persistent socio-economic inequality in frequent headache among Danish adolescents from 1991 to 2014.

The association between socio-economic status (SES) and headache among adolescents is an understudied issue, and no study has examined whether such an association changes over time. The aim was to examine trends in socio-economic inequality in frequent headache among 11- to 15-year-olds in Denmark from 1991 to 2014, using occupational social class (OSC) as indicator of SES. The study applies data from the Danish part of the international Health Behaviour in School-aged Children (HBSC) study. HBSC includes nationally representative samples of 11-, 13- and 15-year-olds. This study combines data from seven data survey years from 1991 to 2014, participation rate 88.6%, n=31,102. We report absolute inequality as per cent difference in frequent headache between high and low OSC and relative inequality as odds ratio for frequent headache by OSC. In the entire study population, 10.4% reported frequent headache. There was a significant increase in frequent headache from 8.0% in 1991 to 12.9% in 2014, test for trend, p<0.0001. This increasing trend was significant in all OSCs. The prevalence of frequent headache was significantly higher in low than high OSC, OR=1.50 (95% CI: 1.34-1.67). This socio-economic inequality in frequent headache was persistent from 1991 to 2014. There was a significant and persistent socio-economic inequality, i.e. increasing prevalence of frequent headache with decreasing OSC. The association between socio-economic position and headache did not significantly change over time, i.e. the statistical interaction between OSC and survey year was insignificant. The prevalence of frequent headache among adolescents increases with decreasing SES. This socio-economic inequality has been persistent among adolescents in Denmark from 1991 to 2014. Clinicians should be aware of this social inequality.

Increasing prevalence of emotional symptoms in higher socioeconomic strata: Trend study among Danish schoolchildren 1991-2014.

Aims: The aims of this study were: (a) to examine trends in daily emotional symptoms among 11- to 15-year-olds from 1991 to 2014 in Denmark, and (b) to examine trends in social inequality in daily emotional symptoms, that is, whether the differences in prevalence between adolescents with parents of varying occupational social class changed over time. Methods: We combined seven comparable cross-sectional Health Behaviour in School-aged Children surveys (N=31,169). Daily emotional symptoms were measured by the HBSC Symptom Check List and occupational social class (OSC) by students' reports about parents' occupation. We calculated absolute (per cent) differences in emotional symptoms between high and low OSC and relative differences by odds ratio for emotional symptoms by parents' OSC. Results: Eight per cent reported at least one daily emotional symptoms, with an increasing trend from 1991 to 2014 (p<0.001). The prevalence in high, middle and low OSC was 6.2%, 7.4% and 10.6% (p<0.0001). From 1991 to 2014, there was an increase in the prevalence of daily emotional symptoms in high (p<0.0001) and middle (p<0.0001) but not low OSC (p=0.4404). This resulted in a diminishing absolute social inequality in emotional symptoms. The statistical interaction between OSC and survey year was significant (p=0.0023) and suggests a diminishing relative social inequality in emotional symptoms from 1991 to 2014. Conclusions: There was an increasing prevalence of daily emotional symptoms from 1991 to 2014 and a diminishing social inequality in prevalence of daily emotional symptoms in terms of both absolute and relative social inequality.

Loneliness and objectively measured physical capability in middle-aged adults.

Loneliness is associated with poor functional ability in older people. Little is known about this association in the middle-aged. The aim is to investigate if perceived loneliness is associated with lower physical capability among middle-aged men and women and if the associations of loneliness with physical capability interact with socioeconomic position and cohabitation status. 5224 participants from Copenhagen Aging and Midlife Biobank (CAMB) aged 49-62 years (mean age 54) were included. Handgrip strength (measured by a dynamometer) and maximal number of chair rises in 30 s was recorded. Multivariate linear regression analyses were adjusted for age, occupational social class, cohabitation status, morbidity and personality traits. No association was found between loneliness and physical capability. For example estimates for handgrip strength in 'often' lonely men and women compared with the 'not lonely' were 1.2 kg (95% CI - 0.5;2.9)/1.0 kg (-0.7;2.6). Low occupational social class was associated with poorer physical capability, and living alone was associated with poorer handgrip strength in men [-2.4 kg (95% CI - 3.2;-1.5)] and poorer chair rise test in women [-0.8 rises (95% CI - 1.6;-0.1)]. There was no support for interactions. In contrast to earlier studies among older people, no association between loneliness and physical capability was found in this cohort of middle-aged men and women. Loneliness may not yet have resulted in detectable differences in physical capability in this age group. Further research is needed to clarify if, and at what point in the life course loneliness begins to affect physical capability.

Prevalence and incidence of sarcopenia in the very old: findings from the Newcastle 85+ Study.

IntroductionRecognition that an older person has sarcopenia is important because this condition is linked to a range of adverse outcomes. Sarcopenia becomes increasingly common with age, and yet there are few data concerning its descriptive epidemiology in the very old (aged 85 years and above). Our aims were to describe risk factors for sarcopenia and estimate its prevalence and incidence in a British sample of the very old.MethodsWe used data from two waves (2006/07 and 2009/10) of the Newcastle 85+ Study, a cohort born in 1921 and registered with a Newcastle/North Tyneside general practice. We assessed sarcopenia status using the European Working Group on Sarcopenia in Older People (EWGSOP) definition. Grip strength was measured using a Takei digital dynamometer (Takei Scientific Instruments Ltd., Niigata, Japan), gait speed was calculated from the Timed Up and Go test, and lean mass was estimated using a Tanita‐305 body fat analyzer. We used logistic regression to examine associations between risk factors for prevalent sarcopenia at baseline and incident sarcopenia at follow‐up.ResultsEuropean Working Group on Sarcopenia in Older People sarcopenia was present in 21% of participants at baseline [149/719 participants, mean age 85.5 (0.4) years]. Many participants had either slow gait speed or weak grip strength (74.3%), and hence measurement of muscle mass was frequently indicated by the EWGSOP definition. Incidence data were available for 302 participants, and the incident rate was 3.7 cases per 100 person years at risk. Low Standardized Mini‐Mental State Examination, lower occupational social class, and shorter duration of education were associated with sarcopenia at baseline, while low muscle mass was associated with incident sarcopenia. Low body mass index (BMI) was a risk factor for both in a graded fashion, with each unit decrease associated with increased odds of prevalent [odds ratio (OR) 1.29, 95% confidence interval (CI): 1.21, 1.37] and incident (OR 1.20, 95% CI: 1.08, 1.33) sarcopenia.ConclusionsTo our knowledge, this is the first study to describe prevalence and incidence of EWGSOP sarcopenia in the very old. Low BMI was a risk factor for both current and future sarcopenia; indeed, there was some evidence that low BMI may be a reasonable proxy for low lean mass. Overall, the high prevalence of sarcopenia among the very old suggests that this group should be a focus for future research.

Role of psychosocial work factors in the relation between becoming a caregiver and changes in health behaviour: results from the Whitehall II cohort study

BackgroundThe present study tested the effects of becoming a caregiver combined with adverse working conditions on changes in health behaviours.MethodsParticipants were 5419 British civil servants from the Whitehall II cohort...

Does self-efficacy mediate the association between socioeconomic background and emotional symptoms among schoolchildren?

Emotional symptoms are widespread among adolescents with the highest prevalence among lower socioeconomic groups. Less is known about why and how to reduce this inequality but personal control, e.g., self-efficacy may be crucial. This study examines whether self-efficacy is a mediator in the association between occupational social class (OSC) and emotional symptoms. Data stem from the cross-sectional Health Behavior in School-aged Children-Methodology Development Survey 2012 (HBSC-MDS) conducted among 11-15-year old schoolchildren in two Danish municipalities. Participation rate was 76.8% of 5165 enrolled schoolchildren, n=3969. Low OSC is associated with higher odds of daily emotional symptoms and low selfefficacy. Schoolchildren with low self-efficacy have higher odds for daily emotional symptoms. We find a strong and statistically significant direct effect between low OSC and daily emotional symptoms (OR=1.55, 95%CI: 1.33; 1.84) and a borderline statistically significant indirect effect of self-efficacy [OR=1.17 (0.99; 1.38)]. Socioeconomic inequality in emotional symptoms exists. This inequality is partly explained by socioeconomic inequality in self-efficacy. Promotion of personal competences like self-efficacy may reduce emotional symptoms among all socioeconomic groups, thereby reducing socioeconomic inequalities in emotional symptoms.

Salud mental en menores españoles. Variables socioeducativas

Resumen Antecedentes. La investigación de la salud mental en los menores de edad supone una necesidad actual como medida de detección, análisis y prevención de posibles trastornos mentales. Objetivo Analizar la influencia de la clase social ocupacional de los padres y de la etapa educativa en la que se encuentran los menores, como factores de riesgo en la salud mental de los menores españoles. Método Se utilizaron los resultados obtenidos en la Encuesta Nacional de Salud Española 2006, con una muestra de 5812 menores de cuatro a 15 años. Se estimaron las medidas de asociación con respecto a las variables independientes midiendo la salud mental mediante el Cuestionario de Capacidades y Dificultades (SDQ-versión padres). Resultados La relación entre la salud mental y la variable clase social ocupacional 1 (más privilegiadas) (OR 0.256) y la clase social 2 (clase media) (OR 0.523) aportaron valores significativos (p0.05) con respecto a la E. Secundaria. Discusión y conclusión Se puede concluir que las clases sociales ocupacionales elevadas y medias son factores protectores con respecto a la salud mental de los menores en comparación con la pertenencia a una clase social ocupacional baja. Sin embargo, la pertenencia a la E. Infantil o a la E. Primaria no supone un factor de riesgo de salud mental en comparación con la pertenencia a la E. Secundaria.

Emotional symptoms among adolescents: epidemiological analysis of individual-, classroom- and school-level factors.

Large proportions of schoolchildren suffer from emotional symptoms and there are large variations across schools. It is unknown to what degree this variation is due to composition of schoolchildren in each school or to contextual factors. Objectives are to identify factors at individual, classroom and school levels associated with emotional symptoms. Data stem from the Danish contribution to the international Health Behaviour in School-aged Children study 2010 including 4922 schoolchildren aged 11-15-years from a random sample of schools and including data from school leaders. Emotional symptoms are defined as daily presence of at least one of four symptoms: feeling low, irritable or bad tempered, nervous and having difficulties falling asleep. Multilevel multivariable logistic regression analyses are applied to identify and quantify factors at individual, classroom and school level. Schoolchildren from low (odds ratio (OR) 1.70, 95% CI: 1.33-2.17) and medium (OR 1.50, 95% CI: 1.22-1.85) occupational social class (OSC), girls (OR 1.32, 95% CI: 1.13-1.56) and schoolchildren exposed to bullying (OR 3.82, 95% CI: 2.71-5.40), had increased odds for emotional symptoms. A negative classroom climate was associated with emotional symptoms (OR 1.29, 95% CI: 0.99-1.69) and so was being part of classrooms with a high prevalence of bullying (OR 1.28, 95% CI: 1.0-1.60). Female sex, low OSC, single parent family, exposure to bullying and a high prevalence of bullying within a class are all associated with emotional symptoms. Most variation across schools is explained by individual-level factors but psychosocial aspects of the classroom environment also play a role.

Commentary: Can maternal-paternal comparisons contribute to our understanding of maternal pre-pregnancy obesity and its association with offspring cognitive outcomes?

In this issue of the IJE, Casas and colleagues 1 report associations of maternal pre-pregnancy obesity with lower infant cognitive development scores in two European cohorts, RHEA and INMA. This was observed after extensive confounder adjustment (in particular, multiple indicators of socioeconomic position) and results were replicated across the two cohorts, reflecting two important methodological strengths of the paper. This paper is timely as it contributes to a body of literature on maternal obesity and offspring cognitive ability that is currently sparse and relatively inconclusive. Although independent associations have been reported in several previous studies, inconsistent 2 and predominantly null 3 associations have also been observed when multiple cohorts have been investigated simultaneously. Nevertheless, the persisting associations observed independently of a comprehensive range of sociodemographic and postnatal factors, here and elsewhere, raise the question as to whether maternal obesity might adversely affect cognitive development in offspring via intrauterine mechanisms. In the presence of escalating rates of obesity worldwide, this association warrants further study since, if causal, this would clearly be an increasing public health concern. A crucial step in investigating the association of maternal obesity with offspring cognitive development, above and beyond replication of results, is to disentangle adjusted associations that are due to intrauterine mechanisms from those that reflect residual cofounding by postnatal, socioeconomic and familial factors not fully captured by the measured confounders. As can be seen from the results in the main paper and supplementary information, maternal obesity and lower infant cognitive scores were both associated with less favourable sociodemographic factors in this study (lower parental education, lower occupational social class, shorter breastfeeding duration etc.). Furthermore, confounder associations (namely, with infant cognitive scores) appeared stronger in RHEA, compared with INMA. Since the unadjusted association for maternal obesity and infant cognition was also substantially larger in RHEA (6.00, vs 3.37 in INMA), and this reduced markedly, resulting in similar effect sizes across the cohorts following confounder adjustment, confounding factors therefore appear to account for a substantial proportion of the observed relationship between maternal obesity and offspring cognition, at least at the unadjusted level. Given the known difficulties in fully capturing confounding, particularly socioeconomic,

Association between area-level socioeconomic deprivation and a cluster of behavioural risk factors: cross-sectional, population-based study

The Commission on Social Determinants of Health has urged governments across the world to promote health equity by reducing the gap between the most and least deprived individuals in society. Some of this gap can be bridged by promoting healthy lifestyles through targeted Public Health policy and interventions. Cross-sectional analyses of data on behavioural risk factors, individual socioeconomic factors and neighbourhood deprivation score collected from 26 290 adults aged over 16 years who participated in the 2008 East of England Lifestyle Survey. After adjustment for individual socioeconomic factors, across quintiles of increasing neighbourhood deprivation, participants were more likely to smoke and less likely to consume five portions of fruit and vegetables on five or more days of the week (least deprived versus most deprived quintile: odds ratios for not smoking 0.45 (0.41-0.50); and fruit and vegetable consumption 0.70 (0.64-0.76), P-trend <0.0001). Greater neighbourhood deprivation and lower occupational social class were independently associated with a lower summary healthy lifestyle score (both P-trend <0.0001). Public health interventions aimed at reducing health inequalities by targeting behavioural risk factors may focus in particular on reducing smoking and increasing fruit and vegetable consumption in more deprived communities.

Nonmarital Cohabitation Among Older Finnish Men and Women: Socioeconomic Characteristics and Forms of Union Dissolution

Nonmarital cohabitation has increased substantially among older adults. Our objective was to enhance understanding of cohabitation by comparing elderly cohabiters with the married according to socioeconomic status and union dissolution. We used population registration data on Finns aged 65 years and older living with a cohabiting partner or a married spouse in 1997 (n = 140,902). The participants were characterized according to various socioeconomic indicators and followed for separation, institutionalization, bereavement, and death until 2002. Elderly cohabitation almost doubled between 1990 and 2003, with 3.4% of men and 2.1% of women currently cohabiting. Low educational attainment, low occupational social class, and living in rented housing were associated with cohabiting rather than being married. Low income among men but high income among women predicted cohabitation. Cohabiting unions were more likely than marriages to end through separation, institutionalization, bereavement, and death, with the highest excess risk being for separation. These effects were mostly independent of socioeconomic factors. In socioeconomic terms, elderly cohabiters are mostly less privileged than the married. Nonmarital unions seem somewhat less protective than marriages against institutionalization and death. The substantially higher risk of separation also puts cohabiters at higher risk of losing potential care and support provided by coresident partners.