Vol. 20, No. 2 Printed in Great Britain

International Journal of Epidemiology © International Epidemiological Association 1991

Socioeconomic Differentials in Cancer among Men GEORGE DAVEY SMITH. DAVID LEON, MARTIN J SHIPLEY AND GEOFFREY ROSE

The relationship between cancer rates and socioeconomic position has been investigated for many years, primarily as a source of aetiological hypotheses. The early studies which treated cancer as a single condition1"2 soon gave way to site-specific investigations3"5 which were notable in providing some of the earliest systematic evidence that environmental factors were important in the aetiology of a number of cancers. In the 1950s cancer incidence rates were used for the first time in explicit attempts to throw light on the still obscure aetiology of a range of malignancies.67 In addition to the underlying role of specific environmental factors, it was also postulated that lower socioeconomic position may go with an increased predisposition to cancer in general .8 This latter suggestion is in line with proposals that there is heightened general susceptibility to disease in particular groups,9'10 including those of low socioeconomic position.11"13 Examination of differentials in site-specific cancer rates is necessary to evaluate these hypotheses. This paper begins with a report of new analyses of Department of Epidemiology and Population Sciences, London School of Hygiene and Tropical Medicine, Keppel Street, London WC1E7HT, UK. The data relating to the Registrar General's Decennial Supplement on Occupational Mortality and the OPCS Longitudinal Study are Crown Copyright.

339

differentials in cancer mortality according to employment grade in the Whitehall study of London civil servants.14 In this study the differentials in all cause and in total cancer mortality have been larger than the equivalent social class differences in the British population,13 which may reflect the fact that the use of civil service employment grade produces groups which are more homogeneous with respect to aspects of material circumstances than those generated by the Registrar General's social classes. The differentials in site-specific cancer mortality in the Whitehall study are compared to the socioeconomic differentials in two other data sets; cancer mortality in the latest Registrar General's Decennial Supplement on occupational mortality16 and cancer registrations in the Office of Population Censuses and Surveys Longitudinal Study.1718 SUBJECTS AND METHODS 77K Whitehall Study In the Whitehall study 18 403 male civil servants aged 40-64 were examined between 1967 and 1969.M A questionnaire was completed regarding age, civil service employment grade and smoking habits. In two departments, with a total of 873 subjects, employment grades were not comparable with the rest of the sample. Thus 17 530 subjects remain for this study. In

Downloaded from http://ije.oxfordjournals.org/ at New York University on June 30, 2015

Davey Smith G (Department of Epidemiology and Population Sciences, London School of Hygiene and Tropical Medicine, Keppel Street, London WC1E 7HT, UK), Leon D, Shipley MJ and Rose G. Socioeconomic differentials in cancer among men. International Journal of Epidemiology 1991; 20: 339-345. The relationship between cancer and socioeconomic position is examined for men using data from three sources—the Whitehall Study of London civil servants, the OPCS Longitudinal Study and the Registrar General's Decennial Supplement Mortality from, or registration for, malignant neoplasms was higher overall in lower socioeconomic groups. There was considerable variation in the strength, and to a lesser extent direction, of the association of specific cancer sites and socioeconomic position within each of the studies. However, between the studies the relationships between socioeconomic position and the particular cancers were very similar. The similarity in results, taken in conjunction with the differences in design and methods of the three studies, makes it very unlikely that these consistent associations are due to artefacts. The heterogeneity in relationships between specific cancer sites and socioeconomic position suggests that no single factor—such as differences in general susceptibility or differences in smoking behaviour—can account for these associations. However socioeconomic differentials displayed by a particular malignancy do offer clues to its aetiology, and provide an indication of the scope that exists for reducing the burden of cancer within a population.

340

INTERNATIONAL JOURNAL OF EPIDEMIOLOGY

TABLE 1 The Registrar General's social classes Examples of occupations in class

Social class I

Professional occupations

II IIIN

IV

Intermediate occupations Skilled occupations (non-manual) Skilled occupations (manual) Partly skilled occupations

V

Unskilled occupations

HIM

Doctors, engineers, lawyers, surveyors Teachen, nurses Cashiers, police officers, commercial travellers Metal workers, butchers, ambulancemen Barmaids, bus conductors, machine tool operators Labourers, office cleaners, porters

culation of rate ratios and confidence intervals (CI) used Cox's proportional hazards regression model,19 with age as a continuous variable and grade as a factor. To validate the presentation of summary measures of the grade effect, formal tests of heterogeneity over age and calendar period were carried out using standard methods. The Decennial Supplement Data from microfiche published as part of the 1981 Registrar General's Decennial Supplement on Occupational Mortality in Great Britain16 were used. Death certificates were coded according to the Ninth Revision of the International Classification of Diseases (ICD). For the sites considered here the ICD code groups remain the same as described above, except for unspecified primary (which are ICD 195 and ICD 199 only in ICD-9, codes 196-198 having been dropped) and haematopoietic malignancies (ICD 200-208). Inspection of the bridge coding exercise,20 in which a sample of deaths were coded according to both ICD-8 and ICD-9, reveals that these changes had no important influence on classification of these groups. Mortality rates were calculated by social class based on occupation (Table 1) for men aged 20-64. The denominators comprised the number of observed deaths in Great Britain in four peri-censal years (19791980, 1982-1983) classified into social class groups based on occupation stated on the death certificate, while the numerators were estimated from the distribution of the population of Great Britain by social class derived from occupation at the 1981 Census. Standardized mortality ratios (SMR) were calculated for nonmanual and manual workers by aggregating observed and expected deaths for social classes I, II, IIIN and IIIM, IV, V respectively. The magnitudes of the manual to non-manual differentials in cancer mortality were estimated by taking the ratios of the corresponding SMRs. The Longitudinal Study The design and methods used in the Office of Population Censuses and Survey's Longitudinal Study have been detailed elsewhere."•18-21 In summary, the study population is a 1% sample of those enumerated at the 1971 Census of England and Wales. The sample was defined as all those individuals whose date of birth, stated at Census, fell on one of four dates of birth spread evenly throughout the year. Nearly 97% (513 072) of these individuals were traced and flagged at the national Health Service Central Register, providing the basis for the linkage of cancer registrations from the National Cancer Registration Scheme to the Census information on study members.

Downloaded from http://ije.oxfordjournals.org/ at New York University on June 30, 2015

some of the analyses, these men are divided into four employment grades: administrative, professional or executive, clerical, and 'other grades' (men in messenger and other unskilled manual jobs). In other analyses, a dichotomy is made between a 'high' group, composed of the administrative and the professional and executive grades, and a 'low' group consisting of the remainder. Smoking has been categorized according to cigarette use as 'current smoker', 'ex-smoker' and 'never smoker'. The 409 men who only smoked pipes or cigars have been excluded from the analyses that involve smoking status. Records from over 99% of subjects were traced and flagged at the National Health Service Central Registry. Death certificates coded according to the Eighth Revision of the International Classification of Diseases (ICD) have been obtained, and this mortality follow-up to 31 January 1987 provides the basis for this analysis. For ten subjects cause of death was unknown: these have been excluded from cause-specific mortality analyses. Mortality due to malignant neoplasms (ICD codes 140-209) has been divided into the following main groups: oesophagus (ICD code 150), stomach (151), colon (153), rectum (154), pancreas (157), trachea, bronchus and lung (162—referred to as lung cancer), prostate (185), bladder (188), brain (191), unspecified primary (195-199), haematopoietic (200-207) and other (all other codes). Three sites with a small number of deaths were included in the 'other' group for the main analyses, but were also examined separately since they have previously been seen to show associations with socioeconomic position. These were liver and gallbladder (155-156—referred to as liver), kidney (189) and malignant melanoma (172). Mortality rates have been calculated using personyears at risk. These rates have been standardized for age at entry by the direct method, using five-year age bands and the total population as the standard. Cal-

341

SOCIOECONOMIC DIFFERENTIALS IN CANCER AMONG MEN TABLE 2

Definition of 'high' and 'low' socioeconomic categories and standard population rates used in the Longitudinal Study (LS) Housing tenure

Household access to cars

Access to household amenities

Own social class

Educational qualifications

'High' socioeconomic category

Owner occupied

Access to one or more cars

Sole use of all amenities

'Low' socioeconomic category

Local authority

No access to cars

One or more amenities missing

Non-manual social classes I-IIIN Manual social classes IIIM-V

Standard population*

All LS males

All LS males

All LS males

Higher/professional qualifications or A-levels No higher/ professional qualifications or A-levels All LS men aged 18-70 at census

AU LS men aged 15 and over at census

"Source of standard rates used in calculation of standardized registration ratios.

study byfivesocioeconomic measures; housing tenure, household access to cars, household access to domestic amenities, social class and educational qualifications. Person-years at risk were accumulated for each level of each measure, taking into account age atriskand year of follow-up. Indirect standardization was used to esti-

TABLE 3 Age-adjusted mortality rates per 1000 person-years (number of deaths) by employment grade in Whitehall study

All causes* All cancers Specific cancers Oesophagus Stomach Colon Rectum Pancreas Lung Prostate Bladder Brain Unspecified Haematopoietic Other Non-cancer

Administrators

Professional or executive

Clerical

Other

X1 for trend

Low/High grade rate ratio

9.89 (141) 3.14 (47)

12.84 (2322) 3.91 (713)

18.00 (905) 5.30 (265)

20.81 (663) 6 24 (212)

137 60 (p-5) 0.55 (p= 0.5) 71.79 (p-CO.001) 2.81 (p=0.1) 0.00 (pX>.5) 0.03 (p>0.5) 3.94 (p= 0.05) 0.16 (pX>-5) 1.52 (p=0.2) 91.77 (p

Socioeconomic differentials in cancer among men.

The relationship between cancer and socioeconomic position is examined for men using data from three sources--the Whitehall Study of London civil serv...
519KB Sizes 0 Downloads 0 Views