ARTICLE IN PRESS

Respiratory Medicine (2006) 100, 1282–1285

0954-6111/$ - sdoi:10.1016/j.r

$The studyNeither agenci�Correspond

Hospital, BirmiE-mail addr

SHORT COMMUNICATION

Age, sex, material deprivation and respiratorymortality$

Rachel Jordana,b, Neville Verlanderc, Babatunde Olowokurea,b,�,Jeremy I. Hawkerd

aHealth Protection Agency, Regional Surveillance Unit (West Midlands), 2nd Floor Lincoln House,Heartlands Hospital, BirminghamB9 5SS, UKbHealth Protection Research and Development Unit, Department of Public Health and Epidemiology,University of Birmingham, Edgbaston, Birmingham B15 2TT, UKcStatistics Unit, Statistics, Modelling and Economics Department, Health Protection Agency Centre forInfections, 61 Colindale Avenue, London NW9 5EQ, UKdLocal and Regional Services Division, Health Protection Agency, 9th Floor, Ladywood House,45 Stephenson Street, Birmingham B2 4DY, UK

Received 24 January 2005; accepted 15 October 2005

KEYWORDSRespiratoryinfection;Mortality;Deprivation;Ecological study;Age;Sex

ee front matter & 2005med.2005.10.014

was funded in part by tes influenced the methoing author. Health Protengham, UK, B9 5SS. Teless: babatunde.olowok

Summary The aim of this study was to examine the effects of age, sex and socialdeprivation on mortality rates for respiratory infection. An ecological study wasundertaken, using official public health mortality data and population census datafor the West Midlands health region, UK. Postcodes at the time of death were used toassign Townsend deprivation scores and the resulting deprivation quintile. Poissonregression analysis was used to estimate the association between respiratorymortality, deprivation quintile, age and sex. In most age groups there was astatistically significant trend of increasing mortality with increasing deprivation. Therelative risk for the most deprived was highest in the 45–64 year age-group(RR ¼ 4.4, 95% CI 4.0, 4.8). However, the absolute risks were greater in those aged75–84 years (RR ¼ 1.3, 95% CI 1.3, 1.4) where the annual death rate was 669 per100,000. Consistently higher mortality rates were seen in males. These resultssuggest that the risk of mortality from respiratory infection varies by sex andgenerally increases with increasing age and deprivation quintile. The identifiedassociation between deprivation and mortality from respiratory infections isconsistent with the effect of deprivation on many other diseases. Addressing thesocial determinants of ill health may help to reduce the high burden of respiratory

Elsevier Ltd. All rights reserved.

he Health Protection Agency and also by the West Midlands Public Health Development Fund.ds or analysis of the study.ction Agency, Regional Surveillance Unit (West Midlands), 2nd Floor Lincoln House, Heartlands.: +121 773 7077; fax: +121 773 1407.ure@hpa.org.uk (B. Olowokure).

ARTICLE IN PRESS

Age, sex, material deprivation and respiratory mortality 1283

mortality in the UK. However, individual level studies and examination of other areasare needed to explain the mechanisms by which deprivation increases the risk ofmortality from respiratory infection, and thereby identify target groups for effectiveinterventions.& 2005 Elsevier Ltd. All rights reserved.

Introduction

Research has consistently shown that those in lowersocioeconomic groups experience poorer healththan those in more affluent socioeconomicgroups.1–7 Although this relationship is extensivelydocumented in a number of age-groups, particu-larly the young, it is comparatively less welldescribed in those aged 65 years or more.

Respiratory infections are a common cause ofmortality and every year in the UK about 35,000people, mainly those aged 465 years, die frompneumonia and other respiratory infections. This isnearly twice the European Union (EU) averageand constitutes over 7% of total elderly deaths inthe UK.8,9

Given the high interest and importance of socio-economic risk factors in contributing to ill health,and the impact of respiratory diseases on mortality,there is remarkably little recent data available onthe potential association between social depriva-tion and mortality from respiratory infection.Furthermore, few studies have examined differ-ences across widely different age-groups.

This study uses 1991 census data linked tomortality data for respiratory infections from1983 to 1999 to examine the effects of age, sexand social deprivation on mortality rates forrespiratory infection in the West Midlands healthregion, UK.

Methods

Anonymized mortality data for the West MidlandsHealth Region, from 1983 to 1999, were obtainedfrom the official public health mortality files of theOffice for National Statistics. Deaths caused byrespiratory infections were identified using pre-viously published aggregations of respiratory ICD-9codes that included acute respiratory infections(460–466), pneumonia (480–486, 770) and influenza(487).10 Between 1984 and 1993 there was a changein interpretation of World Health Organization rule3,11 referring to the assignment of underlying causeof death. This resulted in a sudden artificialreduction in the number of deaths attributed tocertain causes, particularly pneumonia. To adjustfor this artificial reduction, we multiplied the

number of deaths between 1984 and 1993 by ICDcode-specific conversion factors.11

Postcode at time of death was used to identifycensus enumeration district (ED) and assign aTownsend deprivation score to each. This area-based score is derived from four 1991 censusvariables (the proportion of households that: areowner-occupied; have no car; are overcrowded,and; have unemployed residents). The more de-prived an area, the higher the score.12 EDs werethen ranked by deprivation score and divided intofive approximately equal deprivation quintiles(from 1 [least deprived] to 5 [most deprived]).

Data were analysed in STATA 5.0 using Poissonregression analysis with log population as an offsetto take account of variation in population size, andthe variables age, sex and Townsend quintile ascovariates. The models were tested for significanceand adjusted for over-dispersion or under-disper-sion (that is, more or less variation than expectedfor a Poisson model). Significant interaction(Po0.001) was found between age, sex and Town-send quintile. The effect of deprivation wascalculated within each age group separately. Forsome of the age groups there was also a significantinteraction between quintile and sex. We havepresented the results for the quintile relative risksaveraged across the sexes.

Results

During the 17-year study period, 90,448 deathswere attributed to respiratory infections. Of these87,479 (96.7%) could be assigned a Townsend score.

Table 1 shows death rates and relative risks ofdeath for Townsend quintile score by age-group,adjusted for sex. Death rates increase markedlywith increasing age in adults, with highest rates inthose agedX85 years at 36 per 1000 population peryear and consistently higher rates in males. Withthe exception of the oldest age-group, a statisti-cally significant positive mortality trend withincreasing deprivation can be seen. The relativerisk for the most deprived is highest in the 45–64year age-group (RR ¼ 4.4, 95% CI 4.0, 4.8), where30% of deaths are in the most deprived quintile. Inall age-groups 475 years the relative risks aregreater than two. For the 15–44 year age-group,

ARTICLE IN PRESS

Table

1Adjusted

relative

risksof

dea

thfrom

respiratoryinfection,

WestMidland

s19

83-99,

byTownsen

dquintilescorean

dag

e-grou

p

Townsen

dquintile

Relative

risk

ofdea

th(95%

CI)

0–4yr

5–14

yr15

–44

yr45

–64

yr65

–74

yr75

–84

yr85

+yr

1(lea

stdep

rive

d)

11

11

11

12

1.3(1.2,1.5)

1.8(1.5,2.2)

1.1(1.0,1.3)

1.5(1.3,1.7)

1.2(1.1,1.3)

1.03

(0.97,

1.09

)0.96

(0.91,

1.02

)3

1.2(1.0,1.4)

2.2(1.8,2.6)

1.5(1.3,1.7)

1.9(1.7,2.2)

1.4(1.3,1.5)

1.07

(1.01,

1.13

)0.97

(0.92,

1.03

)4

1.9(1.6,2.2)

1.9(1.6,2.2)

2.0(1.8,2.3)

2.8(2.5,3.1)

1.8(1.7,2.0)

1.15

(1.09,

1.22

)0.89

(0.84,

0.95

)5(m

ostdep

rive

d)

2.4(2.1,2.8)

2.6(2.2,3.0)

3.0(2.7,3.4)

4.4(4.0,4.8)

2.2(2.1,2.4)

1.3(1.3,1.4)

0.88

(0.83,

0.94

)

P-va

lue(trend

)o0.00

1o0.00

1o0.00

1o0.00

1o0.00

1o0.00

1o0.00

1

RR(fem

ale)

0.78

0.93

0.54

0.70

0.68

0.73

0.96

(95%

CI)

(0.73,

0.84

)(0.83,

1.01

)(0.51,

0.58

)(0.66,

0.74

)(0.65,

0.71

)(0.70,

0.75

)(0.92,

1.01

)

Totaldea

ths(198

3–99

)55

796

851

3938

11,072

29,918

41,037

Population(199

1)35

1,03

765

2,02

52,18

4,50

41,15

4,74

546

2,27

926

3,09

166

,591

Ann

ualdea

thrate

(per

100,00

0)Male

100.9

324

167

803

4166

Female

80.8

216

117

587

3441

Total

90.9

2.3

2014

166

936

25

R. Jordan et al.1284

ARTICLE IN PRESS

Age, sex, material deprivation and respiratory mortality 1285

males have twice the mortality risk comparedwith females.

Discussion

The results of this study suggest that the risk ofmortality from respiratory infection varies by sexand generally increases with increasing age anddeprivation quintile.

Consistently higher mortality rates were seen inmales, with those aged 15–44 years having approxi-mately twice the mortality risk of females. Deathrates increased markedly with increasing age.Furthermore, within separate age-groups mortalityincreased with increasing deprivation and wassignificantly higher in the most deprived quintilecompared to the least deprived quintile. However,this relationship was not observed in the oldest agegroup and the trend appeared to be in the oppositedirection. This phenomenon of ‘attenuation’ hasbeen described previously.13 It is not clear why‘attenuation’ occurs, it may be that in the elderlypopulation factors such as degree of frailty,comorbidities and social support override anyeffect of deprivation. It may also be that area-level measures of deprivation are not as relevantfor the very elderly.

This study is subject to ecological bias, wherebywe cannot assume that associations at group levelwould also hold at the individual level. However,there is evidence to show that deprivation mea-sures at both area-level and individual-level makeindependent contributions to mortality.14 In addi-tion, the associations identified are both biologi-cally plausible and in line with the effect ofdeprivation on other infectious diseases andchronic diseases.1–7,15–17

It is possible that the inequalities identified in thisstudy are due primarily to confounding social orenvironmental factors. The design of this study doesnot permit examination of potential explanatoryvariables such as smoking, comorbidities (e.g.chronic lung disease), nursing home residency,housing, heating, cold exposure and vaccinationstatus. Individual level studies as well as examina-tion of other area characteristics are required toquantify the direct effect of deprivation, and clarifythe mechanisms by which deprivation increases therisk of mortality from respiratory infection, not onlyin the elderly, but also younger populations.

The identification of modifiable factors andimplementation of proven policy interventions suchas improved uptake of influenza and pneumococcalvaccine may reduce the overall impact of deathsfrom respiratory infections. However, there would

need to be strategic targeting of areas andindividuals in order to prevent identified inequal-ities increasing.

Conflict of interest and source of funding

None of the authors have any conflict of interest

References

1. Hawker JI, Olowokure B, Sufi F, Weinberg J, Gill N, WilsonRC. Social deprivation and hospital admission for respiratoryinfection: an ecological study. Respir Med 2003;97:1219–24.

2. Spencer N, Logan S, Scholey S, Gentle S. Deprivation andbronchiolitis. Arch Dis Child 1996;74:50–2.

3. Das S, Sabin C, Wade A, Allan S. Sociodemography of genitalco-infection with Neisseria gonorrhoeae and Chlamydiatrachomatis in Coventry, UK. Int J STD AIDS 2005;16:318–22.

4. Olowokure B, Hawker J, Weinberg J, Gill N, Sufi F.Deprivation and hospital admission for infectious intestinaldiseases. Lancet 1999;353:807–8.

5. Reading R, Colver A, Openshaw S, Jarvis S. Do interventionsthat improve immunisation uptake also reduce socialinequalities in uptake? Br Med J 1994;308:1142–4.

6. Olowokure B, Wilson RC, Hawker JI. Material deprivation,hospitalisation and adverse events temporally associatedwith immunisation. Vaccine 2004;22:1842–4.

7. Marsh GN, Channing DM. Deprivation and health in onegeneral practice. Br Med J 1986;292:1173–6.

8. European Respiratory Society and European Lung Founda-tion. European Lung White Book—the first comprehensivesurvey on respiratory health in Europe. Sheffield, UK:European Respiratory Society Journals; 2003.

9. Office for National Statistics. Mortality statistics: cause2002, Series DH2 No. 29. London: HMSO; 2003.

10. Wilson D, Bhopal R. Impact of infection on mortality andhospitalization in the North East of England. J Public HealthMed 1998;20:386–95.

11. OPCS. Mortality statistics: cause 1984, Series DH2 No. 11.London: HMSO; 1985.

12. Townsend P, Phillimore P, Beattie A. Health and deprivation:inequality in the north. London: Croome Helm; 1988.

13. Huisman M, Kunst AE, Anderson O, et al. Socioeconomicinequalities in mortality among elderly people in 11European populations. J Epidemiol Community Health 2004;58:468–75.

14. Davey Smith G, Hart C, Watt G, et al. Individual social class,area-based deprivation, cardiovascular disease risk factors,and mortality: the Renfrew and Paisley Study. J EpidemiolCommunity Health 1998;52:399–405.

15. Baker D, Middleton E. Cervical screening and healthinequality in England in the 1990s. J Epidemiol CommunityHealth 2003;57:417–23.

16. Payne JN, Coy J, Milner PC, Patterson S. Are deprivationindicators a proxy for morbidity? A comparison of theprevalence of arthritis, depression, dyspepsia, obesity andrespiratory symptoms with unemployment rates and Jarmanscores. J Public Health Med 1993;15:161–70.

17. Ebrahim S, Papacosta O, Wannamethee G, Adamson J,British Regional Heart Study. Social inequalities and dis-ability in older men: prospective findings from the Britishregional heart study. Soc Sci Med 2004;59:2109–20.