Marital Status, Health and Mortality: The Role of Living Arrangement

1
Marital Status, Health and Mortality The Role of
Living Arrangement

• Paul Boyle, Peteke Feijten and Gillian Raab
• University of St Andrews,
• School of Geography Geosciences
• Longitudinal Studies Centre - Scotland

2
Health differences between the married and
unmarried

• Unmarried people are less healthy and more likely
  to die than their married counterparts.
• This has been found for almost 150 years and in
  many countries
• France Farr, 1858
• Country-comparison by Hu Goldman, 1990
• USA Gove, 1973 Waite, 1995 Lillard Panis,
  1996 Kaplan, 2006
• UK Maxwell Harding, 1998 Breeze et al., 1999
  Gardner Oswald, 2004
• Unmarried usually meant single, but nowadays
  many unmarried people are in consensual unions.

3
Why are married people healthier?

• Selection
• Healthy people are more likely to marry and stay
  married than unhealthy people
• Causality
• Married people have healthier behaviour because
  they
• are cared for and corrected by their partner
• feel the obligation of being a healthy
  spouse/parent
• receive support from their partner in dealing
  with difficult situations
• Married people have on average better material
  well-being (income, assets and wealth)
• Married people have a more satisfying sex life

4
Marital status versus living arrangement

• Are the differences found between married and
  unmarried people due to marital status, or merely
  due to the fact that married people have someone
  to live with?
• Research question
• How does living arrangement affect health
• and death risk?

5
Hypotheses

• Those who live alone are more likely to die than
  those who live with others.
• Unmarried adults (never married, divorced or
  widowed) who live with others are no more likely
  to die than married adults who live with others.
• Living with other adults is more protective than
  living with children.
• Living arrangement in the past influences current
  health.

6
Data

• Longitudinal Study of England and Wales
• Sample LS members enumerated in census 1971,
  1981, 1991 and 2001 aged 20-64 in the census
  year (20-74 in 2001), not living in communal
  establishments, who are not lost in follow up.
• Type of data used
• census data on individual and household
  characteristics
• death records

7
Dependent variables
8
Covariates

• Gender (separate analyses for men and women)
• Age
• Living arrangement
• Marital status
• Living arrangement history
• Social class
• Economic activity status
• Highest qualification
• Tenure
• Car access
• Urban/rural indicator of place of residence

9
Definitions

• Marital status
• never married
• married (includes re-married and separated)
• divorced
• widowed
• Living arrangement
• alone
• with adults
• with children
• with adults and children
• Living arrangement history
• continually living with others
• alone then with others
• with others then alone
• alone with others alone again
• with others alone with others again

10
Method

• Hypotheses 1-3
• Logistic regression of death (0survived, 1died)
  in 10-year post census period, with individual
  and household characteristics in census year as
  covariates
• Hypothesis 4
• Logistic regression of poor health in 2001
  (0good/fair health, 1poor health), with
  individual and household characteristics in 2001
  and living arrangement history as covariates

11
H1 Those who live alone are more likely to die
than those who live with others.
12
H1 Those who live alone are more likely to die
than those who live with others.
13
H2 Unmarried adults (never married, divorced or
widowed) who live with others are no more likely
to die than married adults who live with others.
Source ONS Longitudinal Study of England and
Wales
14
H2 Unmarried adults (never married, divorced or
widowed) who live with others are no more likely
to die than married adults who live with others.
Source ONS Longitudinal Study of England and
Wales
15
H3 Living with other adults is more protective
than living with children.
Source ONS Longitudinal Study of England and
Wales
16
H3 Living with other adults is more protective
than living with children.
Source ONS Longitudinal Study of England and
Wales
17
H4 Living arrangement in the past influences
current health.
Source ONS Longitudinal Study of England and
Wales
18
H4 Living arrangement in the past influences
current health.
Source ONS Longitudinal Study of England and
Wales
19
Summary

• Our findings confirm results from previous
  marital status studies.
• In addition to that
• We found that over the whole 1971-2001 period,
  living with others is associated with better
  health and lower death risk for men and women.
• Yet, living arrangement cannot fully account for
  the protective effect of marriage for men,
  because married men living with others have a
  lower death risk than unmarried men living with
  others.
• When we control for background characteristics
  (mainly socio-economic), effects of living
  arrangement and marital status disappear for
  women.
• We found no support for our hypothesis that
  living arrangement history affects current health.

20
Discussion

• The category Unmarried, living with others
  increasingly consists of unmarried cohabitors. Is
  this similar to being married in the way it
  relates to health?
• What we found are associations. More
  sophisticated modelling could be used to
  distinguish selection from more causal mechanisms
  (e.g., simultaneous equation modelling like in
  Lillard Panis, 1996)

21
Acknowledgements

• The permission of the Office for National
  Statistics to use the Longitudinal Study is
  gratefully acknowledged, as is the help provided
  by staff of the Centre for Longitudinal Study
  Information User Support (CeLSIUS). CeLSIUS is
  supported by the ESRC Census of Population
  Programme (Award Ref H 507 25 5179). The authors
  alone are responsible for the interpretation of
  the data. The clearance number of this
  presentation is 30056A.
• The presentation of this research was made
  possible by an Overseas Conference Grant of the
  British Academy, whose support is gratefully
  acknowledged (grant nr OCG-47356).