Scottish Longitudinal Study
Development & Support Unit
Factors driving geographic variation in mortality in Scotland vs. England and Wales: the contribution of residential mobility and relative deprivation
Laura A Kelly (University of Pennsylvania, USA)
Michel Guillot (University of Pennsylvania, USA)
Samuel H Preston (University of Pennsylvania, USA)
Hans-Peter Kohler (University of Pennsylvania, USA)
13 February 2015
Scottish mortality falls below international standards for developed countries. Mortality in Scotland and other developed countries has steadily declined over the last century. However, a widening gap in life expectancy between Scotland, England and Wales, and other high-longevity countries emerged around 1950 and accelerated after 1980 (McCartney et al. 2012a; McCartney et al. 2012b).
This proposal builds upon previous work (currently under review) by the author documenting the contribution of smoking to Scotland’s excess mortality relative to England and Wales and other developed countries between 1950 and 2009, using two indirect estimation methods to evaluate the contribution of smoking to Scotland’s poor life expectancy showing (Peto et al. 1992; Preston, Crimmins and Wilmoth 2010b). The analysis found strong evidence that smoking-attributable mortality is the primary driver of Scotland’s large and widening adult mortality disadvantage regionally within the United Kingdom and relative to other developed countries.
The underlying reasons for Scotland’s relatively higher smoking prevalence, lung-cancer mortality, and years of life sacrificed to smoking remain uncertain. Smoking behavior can be viewed as a “proximate” cause through which deprivation operates to affect health. Such a possibility is consistent with the widespread observations that, within developed countries, people of lower education or income are more likely to smoke (Huisman, Kunst and Mackenbach 2005). This relationship between the socioeconomic gradient and smoking behaviors are also present in Scotland. However, Scotland has reduced or eliminated its economic disadvantages relative to England and Wales, but not its longevity disadvantage.
The health-selective migration hypothesis proposes that geographic health variation may arise in part due to selective residential mobility between those who move or do not move. In Scotland, previous research has provided mixed evidence for the explanatory power of deprivation on the relationship between population mobility and mortality, depending on the time-period and the geographic level of analysis. In West Scotland (Exeter et al. 2009) or Scotland as a whole (Exeter et al. 2005), area-level deprivation explains the majority of the negative relationship between mortality and population change measured over a 20 year period using broad geographical regions that were designed to be consistent areas throughout time (CATTS) over subsequent census periods (Exeter et al. 2009). Brown et al. used small area population mobility with solely the previous year prior to the 2001 Scottish census (Brown and Leyland 2010) to compare the mortality (all-cause and cause-specific) rates between areas, by population mobility and area deprivation.
This project asks whether factors driving temporal geographic mortality variation are unique to Scotland and supportive of the so-called “Scottish effect” on mortality. By using corresponding longitudinal studies for Scotland (Scottish Longitudinal Study) and England and Wales (ONS Longitudinal Study), the factors driving mortality heterogeneity can be estimated for each region. This project aims to systematically analyze the relationship between area-level relative deprivation, population mobility, and mortality variation in each region (ONS LS using GOR and for the SLS a smaller geography of Council area or Health Board).
- What is the extent of residential mobility flows into or out of a set of local areas (as defined above) within the United Kingdom? Evidence of mobility between areas of different health profiles would support the health-selective migration hypothesis.
- Does residential mobility between local areas associate with migration-dependent health or socioeconomic characteristics? Such associations would support the health-selective migration hypothesis.
- If the health-selective migration hypothesis is supported, could we quantify the migration-attributable mortality profiles between local areas? In other words, if migration had not occurred how would mortality profiles compare across local areas? Decompose the contribution of internal vs. external migrants to local area mortality profiles.
- Document the relative contribution of broad cause-of-death mortality, by age and sex.
- Does health-selective migration operate differently in Scotland vs. England and Wales?
- At the individual level, does socioeconomic status predict the risk of population mobility or health/mortality? Does population mobility affect the subsequent socioeconomic status or health/mortality?
At the area-level, geographic mortality variation will be estimated using age-standardized death rates by cause of death for men and women separately. Causes deemed “insignificant” to public health mortality variation – those causes with deaths counts <500 for either sex (cumulative across all ages, across time period) – will be dropped from the analysis. Latent variables responsible for mortality variation are identified through a factor analysis conducted on a death-rate-by-area matrix. ICD-10 coding will be used.
At the area-level, population mobility will determined for “in-movers,” “out-movers,” and “non-movers” as the percentage of the focal population that moved between local-area units using address at each Census compared with 1 year prior and each Census compared to the next Census (or death). First, potential associations between residential mobility and health limitations will be estimated by comparing health-induced limitations (long-term illness measure) for in-movers and out-movers relative to non-movers. Secondly, the likelihood of post-1991 mobility will be estimated as a function of 1991 health-induced limitations by a nested-logistic regression which will ultimately incorporate measures of socioeconomic characteristics of self and the household.
At the area-level, I will next estimate the contribution of differential population mobility and relative deprivation to the geographic distribution of the mortality of focal area populations. The geographic distribution of health (long-term illness) and mortality (all-cause and cause-specific) will be estimated for residence in 1991 and 2001. Age-standardized estimates of health and mortality prevalence of time- and sex-specific local area populations will be estimated across socioeconomic groups.
Brimblecombe, N., D. Dorling, and M. Shaw. 1999. "Mortality and migration in Britain, first results from the British Household Panel Survey." Soc Sci Med 49(7):981-988.
—. 2000. "Migration and geographical inequalities in health in Britain." Soc Sci Med 50(6):861-878.
Brown, D.and A.H. Leyland. 2010. "Scottish mortality rates 2000-2002 by deprivation and small area population mobility." Soc Sci Med 71(11):1951-1957.
Carstairs, V.and R. Morris. 1989. "Deprivation: explaining differences in mortality between Scotland and England and Wales." Bmj 299(6704):886-889.
Exeter, D.J., P.J. Boyle, Z. Feng, and M. Boyle. 2009. "Shrinking areas and mortality: an artefact of deprivation effects in the West of Scotland?" Health Place 15(1):399-401.
Exeter, D.J., Z. Feng, R. Flowerdew, and P.J. Boyle. 2005. "Shrinking areas and mortality: an artefact of deprivation effects?" J Epidemiol Community Health 59(11):924-926.
Hanlon, P., R. Lawder, D. Buchanan, A. Redpath, D. Walsh, R. Wood, M. Bain, D. Brewster, J. Chalmers, and D. Walsh. 2005. "Why is mortality higher in Scotland than in England and Wales? Decreasing influence of socioeconomic deprivation between 1981 and 2001 supports the existence of a ‘Scottish Effect’." Journal of Public Health 27(2):199-204.
Hanlon, P., D. Walsh, D. Buchanan, A. Redpath, M. Bain, D. Brewster, J. Cahalmers, R. Muir, M. Samlls, and J. Willis. 2001. Chasing the Scottish Effect: Why Scotland needs a step-change in health if it is to catch up with the rest of Europe: Common Services Agency for the Scottish Health Service, Information & Statistics Division.
Huisman, M., A.E. Kunst, and J.P. Mackenbach. 2005. "Educational inequalities in smoking among men and women aged 16 years and older in 11 European countries." Tob Control 14(2):106-113.
Human Mortality Database. 2014. "Human Mortality Database." Berkeley (USA) and Germany: University of California and Max Planck Institute for Demographic Research.
McCartney, G., C. Collins, D. Walsh, and G.D. Batty. 2012a. "Why the Scots die younger: synthesizing the evidence." Public Health 126(6):459-470.
McCartney, G., D. Walsh, B. Whyte, and C. Collins. 2012b. "Has Scotland always been the 'sick man' of Europe? An observational study from 1855 to 2006." Eur J Public Health 22(6):756-760.
Norman, P., P. Boyle, and P. Rees. 2005. "Selective migration, health and deprivation: a longitudinal analysis." Soc Sci Med 60(12):2755-2771.
Peto, R., A.D. Lopez, J. Boreham, M. Thun, and C. Heath, Jr. 1992. "Mortality from tobacco in developed countries: indirect estimation from national vital statistics." Lancet 339(8804):1268-1278.
Preston, S.H., E.M. Crimmins, and J.R. Wilmoth. 2010b. "Contribution of smoking to international differences in life expectancy." Pp. 105-131 in International Differences in Mortality at Older Ages: Dimensions and Sources., edited by S.H. Preston, E.M. Crimmins, and B. Cohen. Washington (DC). National Academies Press (US) National Academy of Sciences.
Tencza, C., A. Stokes, and S. Preston. 2014. "Factors responsible for mortality variation in the United States: A latent variable analysis." Demographic Research 31(2):27-70.
Walsh, D., M. Taulbut, and P. Hanlon. 2010. "The aftershock of deindustrialization—trends in mortality in Scotland and other parts of post-industrial Europe." The European Journal of Public Health 20(1):58-64.