The European Journal of Public Health Advance Access originally published online on January 31, 2007
The European Journal of Public Health 2007 17(5):486-491; doi:10.1093/eurpub/ckl269
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Health Inequalities |
Health inequalities with the National Statistics-Socioeconomic classification: disease risk factors and health in the 1958 British birth cohort
Kate Atherton and Chris PowerCentre for Paediatric Epidemiology and Biostatistics, Institute of Child Health, 30 Guilford Street, London, WC1N 1EH, UK
Correspondence: Dr Kate Atherton, Centre for Paediatric Epidemiology and Biostatistics, Institute of Child Health, 30 Guilford Street, London, WC1N 1EH, UK, tel.: +44-207-813-8393, fax: +44-207-905-2381, email: k.atherton{at}ich.ucl.ac.uk
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionConclusionsAcknowledgementsReferences |
|---|
Background: Health inequalities using the new National Statistics socioeconomic classification (NS-SEC) have so far been assessed using only general measures of health, with little known about inequality for specific health outcomes. Preliminary analyses show that self-employed workers, distinguished for the first time by NS-SEC, show increased mortality risk in the last 5 years of working life. We examined health inequalities for multiple disease risk factors and health outcomes, with particular reference to cardiorespiratory risk in the self-employed. Methods: 8952 participants in the 1958 British birth cohort with information on adult occupation and disease risk factors at 45 years. Systolic and diastolic blood pressure, body mass index, glycosylated haemoglobin, total and high density lipoprotein (HDL) cholesterol, triglycerides, fibrinogen, C-reactive protein, tissue plasminogen activator (t-PA), von Willebrand factor, total immunoglobulin E (IgE), one-second forced expiratory volume, 4 kHz hearing threshold, visual impairment, depressive symptoms, anxiety, chronic widespread pain and self-rated health were measured. Results: Routine workers had poorer health than professional workers for most outcomes examined, except HDL cholesterol, triglycerides, t-PA and IgE in men; total cholesterol and IgE in women. Patterns of inequality varied depending on the outcome but rarely showed linear trend across the classes. Relative to professionals, own account workers (self-employed) did not show consistently increased levels of cardiorespiratory risk markers. Conclusions: Health inequalities are seen with NS-SEC across diverse outcomes for men and women. In mid-life, self-employed workers do not have an adverse cardiorespiratory risk profile.
Keywords: gender, health inequalities, NS-SEC, self-employment, social class
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionsAcknowledgementsReferences |
|---|
Since its inception over 100 years ago the Registrar General's (RG) social classification has been key to the understanding of social inequalities in health in the UK.1,2 Inequalities across the RG classification have been demonstrated for diverse health outcomes and disease risk factors, consistently and over time.3,4 However, demand for a measure with a clearer conceptual basis led, in 2001, to the National Statistics socioeconomic classification (NS-SEC) replacing the RG classification as the official measure of social class in the UK.5 Like RG, NS-SEC is an occupational-based measure however it has an explicit theoretical basis in the Goldthorpe Class schema2,6 which divides the working population into employers (who buy labour) employees (who sell their own labour) and self-employed workers (who neither buy nor sell their labour).
Our understanding of social inequalities in the UK is so closely associated with the RG classification that it is essential to fully characterize the impact of the switch to NS-SEC. Health inequalities were demonstrated with NS-SEC first in validation studies7 and subsequently in other studies, although so far these have been limited to general measures of health or mortality.8–14 Little is known about inequality for specific risk factors or health outcomes with NS-SEC.
The NS-SEC classifies self-employed workers and managers of small companies as a distinct group (termed own account workers), previously these individuals were distributed across the RG classes. Mortality rates for male own account workers have been shown to be more akin to those of managerial and professional groups than intermediate and routine occupations in most age groups,15,16 but rise dramatically in the final 5 years of working life.16 This rise was seen for deaths from ischaemic heart disease, stroke and cancer as well as all-cause mortality. To date, there has been no investigation of whether the later increase in mortality rate is predated by an increase in disease risk factors.
The 1958 British birth cohort is a nationwide sample, which is well characterized with respect to health inequalities17–20; participants have recently completed a biomedical survey at age 44–45 years.19 The cohort provides an ideal study sample in which to examine health inequalities with NS-SEC for several specific measures in an economically active population prior to the onset of life shortening chronic disease. Our aims were 2-fold: first, to establish whether social inequalities exist across NS-SEC for a range of specific measures of health and disease risk in men and women, and to examine the pattern of inequality. Secondly, to examine the health of own account workers; we postulated that this group would have a relatively high prevalence of cardiorespiratory risk factors at this premorbid life stage, prior to the subsequent increase in mortality reported elsewhere.15
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionsAcknowledgementsReferences |
|---|
Study population
Participants were originally enrolled in the Perinatal Mortality Survey of all born in England, Scotland and Wales, during 1 week in March 1958 (n = 17 638),21,22 with follow-up throughout childhood and in adulthood at age 23, 33, 42 and 45 years.19 Immigrants with the same birth dates were recruited to the study up to age 16 years (n = 920), to give a total of 18 558 cohort members. At age 45 years, 12 069 participants still in contact with the study were invited to a clinical examination undertaken in their home by a trained nurse. A total of 9377 participants were examined, of whom 8952 were economically active at either age 42 or 33 years and could be assigned an NS-SEC class. Ethical approval for the medical examination of the British 1958 Birth Cohort was obtained from South East multi-centre Research Ethics Committee (ref: 01/1/44).
Biomedical outcomes
Details of the biomedical assessment have been given elsewhere.23 Briefly, mean systolic and diastolic blood pressures were calculated from a maximum of three readings; body mass index (BMI) was calculated in kg/m2 from standardized measurements of height and weight. Non-fasted venous blood samples were analysed for glycosylated haemoglobin (HbA1c), total and high-density lipoprotein (HDL) cholesterol, triglycerides, fibrinogen, C-reactive protein (CRP), tissue plasminogen activator (t-PA), von Willebrand factor (vWF), and total immunoglobulin E (IgE) levels. Lung function was assessed by spirometry to give forced expiratory volume (FEV1); hearing threshold at 4 kHz was taken as the reading in the better ear. Visual acuity was assessed using a logMAR scale with visual impairment defined as logMAR > 0.2 in the better eye. Depressive and anxiety symptoms were assessed by nurse-administered Clinical Interview Schedule24; participants were scored as symptomatic if they had two or more symptoms. Participants shaded areas of pain on a manikin from which chronic widespread pain was assessed according to American College of Rheumatology criteria,25 they also rated their general health as excellent, good, fair or poor: fair and poor categories were combined and are hereafter termed poor self-rated health.
Occupational classifications
Both RG social class and NS-SEC were derived from information on current or most recent job at 42 years; if occupation at 42 years was not given, information from the 33-year survey was used (males: n = 360; females: n = 774). Six RG categories were used: professional (I), managerial/technical (II), other non-manual (IIInm), skilled manual (IIIm), partly skilled (IV) and unskilled manual (V). NS-SEC class was derived using the full derivation method based on 1990 Standard Operational Classification (SOC) codes.26 Operational categories were collapsed into five classes: managerial/professional (1), intermediate (2), small employers/own account workers (3), lower supervisory/technical (4) and semi-routine/routine (5).
Data analysis
Observed mean values or proportions were presented by NS-SEC class for each outcome. To test for inequalities we compared semi-routine/routine workers (NS-SEC 5) with managerial/professional workers (NS-SEC 1) using linear regression [with 95% confidence intervals (CI)] for continuous outcomes and logistic regression [to give odds ratios (OR) with 95% CI] for categorical outcomes. To test whether own account workers have poorer health we compared NS-SEC 3 with NS-SEC 1 workers. Because men and women occupy different types of occupations in general, models were constructed separately for men and women. HDL cholesterol, triglycerides, fibrinogen, CRP, t-PA, vWF, total IgE and hearing threshold were handled using a natural log transformation; geometric means are presented and percentage change in outcome calculated.27 For HbA1c, geometric means are presented but analyses use untransformed data, with robust estimators to account for heteroscedascity that could not be eliminated by transformation.28 Where appropriate, analyses were adjusted for factors that might affect the measurement of outcome (treatment for hypertension or diabetes, use of respiratory inhaler, recent food consumption, recent chest infection, flooring, air temperature, background noise, time or month of interview, laboratory batch and delay in receipt of blood sample).
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionsAcknowledgementsReferences |
|---|
Table 1 shows the composition of the five NS-SEC classes, with respect to RG social class. A relatively small proportion of the population are in the extreme RG classes (10.7% of men and 7.6% of women fall into I or V), whereas there is better distribution across NS-SEC. The managerial/professional NS-SEC class is the most prevalent group for men and women, although relatively high proportions of women also have semi-routine or routine occupations. At the extremes, RG and NS-SEC match closely; all of RG I and the majority of II are categorized as NS-SEC 1, likewise classes IV and V are mostly categorized as NS-SEC 5. There is also reasonable comparability for the intermediary classes such that the majority of IIInm can be found in NS-SEC 2 and IIIm in NS-SEC 4. However, as expected some individuals classified as non-manual workers by RG appear as routine workers in NS-SEC. NS-SEC 3 (own account workers) consists of individuals from across RG, except for class I. For men, most NS-SEC 3 workers come from RG IIIm and II. For women, the composition of NS-SEC 3 is more evenly dispersed, although class II predominates.
|
Inequalities between NS-SEC 1 and 5 were demonstrated for most outcomes studied such that health was poorer in NS-SEC 5, relative to NS-SEC 1 (Tables 2 and 3). In both men and women there was inequality for blood pressure, BMI, HbA1c, fibrinogen, CRP, vWF, FEV1, hearing threshold, visual impairment, depressive symptoms, anxiety symptoms, chronic widespread pain and self-rated health. Additionally, total cholesterol showed inequality in men; HDL cholesterol, triglycerides and t-PA in women. Apart from the general trend of poorer outcomes in NS-SEC 5 relative to NS-SEC 1, there was no consistent gradient in inequality across the outcomes.
|
|
For male own account workers (NS-SEC 3) systolic blood pressure and levels of HbA1c were higher and hearing at 4 kHz was poorer than NS-SEC 1; however NS-SEC 3 males also had the highest levels of beneficial HDL cholesterol (Table 2). In contrast, female NS-SEC 3 workers had lower levels of HbA1c, fibrinogen and vWF than NS-SEC 1 (Table 3). This group also had the lowest hearing threshold of all NS-SEC classes, indicating their better hearing, although the confidence intervals for this estimate included zero. Both men and women in NS-SEC 3 had increased odds of chronic widespread pain (OR = 2.2, 95% CI: 1.6, 2.9 and OR = 1.9, 95% CI: 1.3, 2.8, respectively).
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionsAcknowledgementsReferences |
|---|
Main findings of this study
We demonstrated poorer health amongst routine workers relative to professionals using NS-SEC, for both men and women, across multiple measures of health and disease risk. To our knowledge this is the first study to examine health inequalities with NS-SEC and such a wide range of health indicators. There was no consistent pattern of inequality, that is trend across NS-SEC classes, rather it varied according to the outcome. With respect to own account workers, there was little evidence to suggest that this group had a poorer cardiorespiratory risk profile at this premorbid age.
Strengths and weaknesses of this study
Currently the prevalence of chronic disease in the cohort is low, therefore this group can generally be considered to be in a premorbid state.29 In contrast to previous studies, multiple and, for the most part, objective measures of health and disease risk were available, allowing comparison of health inequalities with NS-SEC across a range of specific outcomes. The study sample provides nationwide coverage, and whilst attrition is inevitable in a study that follows people for long periods of time, the sample of 8952 individuals (55.8% of the resident surviving cohort at 45 years) is broadly representative of this age group in the general population. For example, 24.2% of the cohort were in semi-routine or routine occupations, compared with 25.5% of the 2001 England and Wales census population (age 25–44 years).30
A total of 1134 participants (12.6%) were assigned an NS-SEC class on the basis of occupation at age 33 years, mostly due to economic inactivity (homemaker, sickness or disability, or unemployment) at age 42 years, whereas a minority (2.3%) had not participated in the 42-year survey. Our use of 33-year occupational data is consistent with the recommendation that the economically inactive are classified by their most recent occupation, an approach found to be appropriate even for the long-term non-employed.5 When analyses were repeated excluding those who had not participated at the 42-year survey results were little changed by exclusion of this small group. As we might expect, however, for some health measures, inequalities were reduced, such that the confidence intervals spanned unity (anxiety in NS-SEC 5 for men and women and fibrinogen in NS-SEC 3 for women).
NS-SEC is designed to be derived from 2000 SOC coding of occupation, but for the cohort only 1990 SOC codes were available, from which an interim version was derived. A study comparing both versions of NS-SEC showed 95% agreement.31 The distribution of individuals across NS-SEC classes relative to RG in our study is comparable to that reported elsewhere, using NS-SEC derived from 2000 SOC.32
Comparison with other studies
We found inequalities between extreme classes of NS-SEC across several measures of disease risk and physical and mental functioning. Previous studies investigating health inequalities with NS-SEC have only considered general, non-specific measures of health (such as self-rated health) or mortality as outcomes. Our study based on the 1958 cohort shows that in general, outcomes for which health inequalities were absent with NS-SEC also showed a lack of inequality with RG social class.23 NS-SEC is explicitly based on work relations theory, however it would be misleading to interpret NS-SEC solely as a work-related measure since occupation is closely linked to other dimensions of socioeconomic position, such as material circumstances.2
The pattern of inequality with NS-SEC varied depending on the outcome, but was rarely one of consistent incremental decrease in health across the classes. Previous studies reported health to deteriorate incrementally across NS-SEC using general measures of health.8,11–13 Whereas in our study poor self-rated health did not increase incrementally across the classes, largely because own account workers reported relatively good health. Differences between studies may be due to the differing age structure of the study samples. Longitudinal analysis of a mixed age sample has shown own account workers to have better rated health than intermediate occupation workers in the first decades of working life.11,12 Estimates from a mixed age sample may give different results than a sample at a specific age.
We found little evidence to suggest that own account workers had a poorer cardiorespiratory risk profile compared with other classes. NS-SEC 3 males showed only higher systolic blood pressure and levels of HbA1c than professional/managerial workers, whilst females had better health on a number of cardiorespiratory risk factors. A previous study showed that NS-SEC 3 males experienced a dramatic increase in both all cause and, specifically, cardiovascular disease mortality in the final few years of working life,15 although a longitudinal analysis of self-rated health, a powerful predictor of mortality, did not find a contemporaneous decrease in health amongst this group. The average age of the 1958 cohort at the clinical examination was 45 years, 15–20 years younger than the age at which the increase in mortality was observed; even so, one might expect to see elevated levels of risk factors at this earlier age, prior to the onset of major life-shortening chronic disease. It may be that poor health in the final decades of working life forces individuals from other NS-SEC classes to take up self-employment, either for the benefits of more flexible working conditions or because they are otherwise unemployable. Whilst it remains important to monitor cardiorespiratory risk in this group as they approach retirement future work may also wish to consider whether social mobility can explain the mortality increase.
Own account workers did, nevertheless, report some functional limitations: males had the greatest hearing loss of any NS-SEC class, and both males and females reported the highest prevalence of chronic widespread pain. Own account workers have been previously shown to be at greater risk of developing limiting long-standing illness relative to higher managerial/professional groups, although semi-routine/routine classes had the greatest risk.10 As this measure included any physical or mental limitation it is possible that a specific susceptibility to chronic widespread pain was diluted.
|
Conclusions |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionsAcknowledgementsReferences |
|---|
Health inequalities are evident with NS-SEC across a variety of measures of health and disease risk. This finding is not surprising, given the close overlap with RG categories, particularly in the extreme classes. The varied pattern of inequality seen across NS-SEC classes suggests that the aetiological role of the social constructs captured by NS-SEC is not uniform across health outcomes. At this premorbid age own account workers do not show an adverse cardiorespiratory profile. Whilst this result is preliminary it is possible that health selection effects may provide an alternative explanation for the later increase in mortality.
|
Acknowledgements |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionsAcknowledgementsReferences |
|---|
We are especially grateful to the study participants in the 2002–04 biomedical follow-up and to the nurses, office and laboratory staff who contributed to the successful completion of the nationwide fieldwork. Peter Shepherd provided assistance with SOC occupational data; Leah Li provided statistical support. The biomedical examination and related statistical analyses were funded by Medical Research Council grant G0000934, awarded under the Health of the Public initiative. Research at the Institute of Child Health and Great Ormond Street Hospital for Children NHS Trust benefits from research and development funding received from the NHS Executive.
Key points
|
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionsAcknowledgementsReferences |
|---|
1 Szreter SRS. The genesis of the Registrar-General's social classification of occupations. British J Sociol (1984) 35(4):522–46.[CrossRef]
2 Galobardes B, Shaw M, Lawlor DA, et al. Indicators of socioeconomic position (part 2). J Epidemiol Commun Health (2006) 60(2):95–101.
3 Acheson D. Independent inquiry into inequalities of health (1998) London: The Stationery Office.
4 Department of Health and Social Security. Inequalities in health: report of a research working group (Black report). (1980) London: DHSS.
5 Rose D, O'Reilly K. The ESRC review of Government Social Classifications. (1998) London & Swindon: Office for National Statistics; Economic and Social Research Council.
6 Rose D, Pevalin DJ. The NS-SEC explained. In: A researcher's guide to the National Statistics socio-economic classification (2003) London: Sage Publications Ltd. 28–43.
7 Rose D, Pevalin DJ. A researcher's guide to the National Statistics socio-economic classification (2003) London: Sage Publications Ltd.
8 Craig P, Forbes J. Social position and health: are old and new occupational classifications interchangeable? J Biosoc Sci (2005) 37(1):89–106.[CrossRef][Web of Science][Medline]
9 Macintyre S, McKay L, Der G, Hiscock R. Socio-economic position and health: what you observe depends on how you measure it. J Public Health Med (2003) 25(4):288–94.
10 Bartley M, Sacker A, Clarke P. Employment status, employment conditions, and limiting illness: prospective evidence from the British household panel survey 1991-2001. J Epidemiol Commun Health (2004) 58(6):501–6.
11 Drever F, Doran T, Whitehead M. Exploring the relation between class, gender, and self rated general health using the new socioeconomic classification. A study using data from the 2001 census. J Epidemiol Commun Health (2004) 58(7):590–6.
12 Sacker A, Clarke P, Wiggins RD, Bartley M. Social dynamics of health inequalities: a growth curve analysis of aging and self assessed health in the British household panel survey 1991–2001. J Epidemiol Commun Health (2005) 59(6):495–501.
13 Chandola T, Bartley M, Wiggins R, Schofield P. Social inequalities in health by individual and household measures of social position in a cohort of healthy people. J Epidemiol Commun Health (2003) 57(1):56–62.
14 Chandola T, Jenkinson C. The new UK National Statistics Socio-Economic Classification (NS-SEC); investigating social class differences in self-reported health status. J Public Health Med (2000) 22(2):182–90.
15 Fitzpatrick J. Examining mortality rates by the NS-SEC using death registration data and the 1991 census. In: A researcher's guide to the National Statistics socio-economic classification (2003) London: Sage Publications Ltd. 173–93.
16 Sacker A, Firth D, Fitzpatrick R, et al. Comparing health inequality in men and women: prospective study of mortality 1986-96. BMJ (2000) 320:1303–7.
17 Manor O, Matthews S, Power C. Health selection: the role of inter- and intra-generational mobility on social inequalities in health. Soc Sci Med (2003) 57(11):2217–27.[CrossRef][Web of Science][Medline]
18 Power C, Matthews S. Origins of health inequalities in a national population sample. Lancet (1997) 350(9091):1584–89.[CrossRef][Web of Science][Medline]
19 Power C, Elliott J. Cohort profile: 1958 British birth cohort (National Child Development Study). Int J Epidemiol (2006) 35(1):34–41.
20 Power C, Manor O, Matthews S. The duration and timing of exposure: effects of socioeconomic environment on adult health. Am J Public Health (1999) 89(7):1059–65.
21 Butler NR, Bonham DG. Perinatal mortality: the first report of the British Perinatal Mortality Survey. (1963) Edinburgh: Livingstone.
22 Butler NR, Alberman E. Perinatal problems: the second report of the British Perinatal Mortality Survey. (1969) Edinburgh & London: Livingstone.
23 Power C, Atherton K, Strachan DP, et al. Lifecourse influences on health in British adults: effects of socioeconomic position in childhood and adulthood. In: Int. J Epidemiol.
24 Lewis G, Pelosi AJ, Araya R, Dunn G. Measuring psychiatric disorder in the community: a standardized assessment for use by lay interviewers. Psychol Med (1992) 22(2):465–86.[Web of Science][Medline]
25 Wolfe F, Smythe HA, Yunus MB, et al. The American College of Rheumatology 1990 criteria for the classification of Fibromyalgia. Report of the Multicenter Criteria Committee. Arthritis Rheum (1990) 33(2):160–72.[Web of Science][Medline]
26 Office for National Statistics. The National Statistics socio-economic classification: user manual (2005) Basingstoke: Palgrave Macmillan.
27 Cole TJ. Sympercents: symmetric percentage differences on the 100 log(e) scale simplify the presentation of log transformed data. Stat Med (2000) 19(22):3109–25.[CrossRef][Web of Science][Medline]
28 White H. A heteroskedasticity-consistent covariance matrix estimator and a direct test for heteroskedasticity. Econometrica (1980) 48(4):817–38.[CrossRef][Web of Science]
29 Thomas C, Hypponen E, Power C. Type 2 diabetes mellitus in midlife estimated from the Cambridge Risk Score and body mass index. Arch Intern Med (2006) 166(6):682–88.
30 Office for National Statistics. last accessed:1-11-2006. http://www.statistics.gov.uk/census2001/access_results.asp.
31 Donkin A, Lee YH, Toson B. Implications of cha nges in the UK social and occupational classifications in 2001 for vital statistics. Popul Trends (2002) 107):23–9.[Medline]
32 Heath A, Martin J, Beerten R. Old and new social class measures. In: A researcher's guide to the National Statistics socio-economic classification (2003) London: Sage Publications Ltd. 226–43.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  K Atherton, D J Berry, T Parsons, G J Macfarlane, C Power, and E Hypponen Vitamin D and chronic widespread pain in a white middle-aged British population: evidence from a cross-sectional population survey Ann Rheum Dis, June 1, 2009; 68(6): 817 - 822. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||