REVIEW ARTICLE

Is high socioeconomic status a risk factor for multiple sclerosis? A systematic review R. Gouldena,b, T. Ibrahimc and C. Wolfsona,c Department of Epidemiology, Biostatistics and Occupational Health, McGill University, Montreal, QC; bNewcastle Medical School, Newcastle University, Newcastle, UK; and cDepartment of Medicine, McGill University, Montreal, QC, Canada

Keywords:

epidemiology, multiple sclerosis, risk factors, socioeconomic status, systematic review Received 10 April 2014 Accepted 29 August 2014 European Journal of Neurology 2015, 22: 899–911 doi:10.1111/ene.12586

High socioeconomic status (SES) is generally associated with better health outcomes, but some research has linked it with an increased risk of multiple sclerosis (MS). The evidence for this association is inconsistent and has not previously been systematically reviewed. A systematic review of cohort and case–control studies in any language was conducted looking at the association between MS and SES. MEDLINE and EMBASE were searched for articles in all languages published up until 23 August 2013. Twenty-one studies from 13 countries were included in the review. Heterogeneity of study settings precluded carrying out a meta-analysis, and a qualitative synthesis was performed instead. Five studies, all from more unequal countries, reported an association between high SES and MS. Thirteen studies reported no evidence of an association, and three studies reported an association with low SES. These 16 studies largely came from more egalitarian countries. The evidence for an association between high SES and increased MS risk is inconsistent but with some indication of a stronger effect in countries and time periods with higher inequality. Firm conclusions are hampered by the failure of most studies to control for other important risk factors for MS.

Background Multiple sclerosis (MS) is a chronic disease of the central nervous system driven by inflammation and neurodegeneration. Whilst its precise aetiology remains unknown, several risk factors have been reported, including female sex, low sunlight exposure, low vitamin D levels, Epstein Barr virus infection, and smoking [1]. Although less widely researched, numerous studies have posited a link between socioeconomic status (SES) and MS. Whilst many diseases are associated with low SES, MS is one of a smaller set of conditions linked to high social class [2–5]. The relationship, however, is unclear and poorly characterized. Internationally, MS occurs with greater frequency in high income nations [6]. Within countries, however, some studies find MS occurs more frequently amongst high SES groups, whilst other studies find no social Correspondence: R. Goulden, Department of Epidemiology, Biostatistics and Occupational Health, McGill University, 1020 Pine Avenue West, Montreal, QC, Canada H3A 1A2 (tel.: +1 438 937 9917; fax: +1 514 934 4458; e-mail: [email protected]).

© 2014 EAN

gradient, or even the opposite [7,8]. Much recent and ongoing research includes SES as a potential confounding factor without a clear understanding of the relationship. The aim of this systematic review is to determine whether SES is related to the risk of developing MS.

Methods Search strategy and study selection

A search was performed in MEDLINE and EMBASE, both via Ovid, for articles in any language published up to 23 August 2013. Subject headings and titles/abstracts were searched for a combination of MS- and SES-related terms, outlined in Table 1. Studies were included if they were cohort or case– control studies, with MS occurrence as the outcome and SES as the exposure. Descriptive and ecological studies were excluded as our interest was in establishing the causal nature, if any, of the relationship between SES and MS, and this would not be possible

899

EUROPEAN JOURNAL OF NEUROLOGY

a

900

R. GOULDEN, T. IBRAHIM AND C. WOLFSON

Table 1 Search terms used

Search field MEDLINE MeSH headings EMBASE Emtree headings Title and abstract (used in both)

MS-related terms

SES-related terms

exp multiple sclerosis

exp socioeconomic factors

exp multiple sclerosis ‘multiple sclerosis’ or ‘disseminated sclerosis’

exp social status or exp socioeconomics socioeconomic* or socio-economic* or SES or social status or social class or income or occupation* or employment or unemploy* or education* or wealth or affluen* or poverty or depriv* or residence or neighborhood

exp, ‘exploded’ search, with all lower branches of the term searched. *wildcard character.

in such uncontrolled studies. Measures of SES had to apply to a time before disease onset. They could relate to income, education level, occupation or neighbourhood characteristics, either of the subjects or their parents. Two reviewers (RG, TI) first screened the title and abstract of all articles retrieved by the search. The full text of selected studies was further assessed for eligibility and final inclusion in the study. The reference lists of these studies, and review articles, were manually searched to identify further relevant articles. Data extraction

The following data were extracted from each study: location, study design, case and control population, SES measure used, potential confounders examined, and the measure of association with MS. Estimated odds ratios (ORs) or relative risks (RRs), and their confidence intervals, were extracted if they were reported; otherwise they were calculated if there were sufficient data available. Data were extracted by one author (RG) and reviewed by another (TI). Quality assessment

Quality assessment criteria that were specific to the features of our question were developed using an iterative process. Two authors (RG, CW) independently assessed a sample of five articles using the Newcastle Ottawa Scale [9] and discussed any discrepancies in scoring. This process was used as a basis to outline the features of an ‘ideal’ study to which each article would be compared. Studies were assessed against this

standard by one author (RG), with a second author (CW) reviewing the process; disagreements were resolved through consensus. Studies were evaluated on their susceptibility to the three main categories of bias – selection, information, and confounding – and the appropriateness of their analytical methods. An overall summary score was not assigned as it was felt that study quality is a multidimensional and qualitative concept. Rather, each specific criterion was evaluated with both a score (A to C, with A being the highest) and a descriptive explanation for this rating.

Results Literature search

The search strategy retrieved 1517 articles from MEDLINE and 2382 from EMBASE (Fig. 1). Of these articles, 107 were selected for full text review, of which 17 met the inclusion criteria. Ten articles identified from reference lists had their full text reviewed, of which three were included. One study that was only available as a conference abstract during the initial search was subsequently published as a full article and was also included [10]. This gave a total of 21 articles, of which 18 were in English and three in Spanish. The main reasons for excluding articles whose full text was reviewed were that SES was included only as a confounder (n = 32), the study was a case series or ecological study (n = 25), they were review articles or commentaries (n = 20), or their measure of SES referred to a time point after disease onset (n = 10). The 21 included studies are summarized in Table 2. The studies came from 13 different countries, published across 47 years (1967–2014), and ranging in size from a study of 16 cases and 56 controls to a cohort study of over 1.5 million individuals. Due to the geographical and temporal diversity of the included studies, and the wide range of SES measures used, a qualitative synthesis was performed. Studies reporting an association with high SES

Four case–control studies [12–15] and one nested case– control study [16] reported a statistically significant association between high SES and increased MS risk. The largest of these studies, and the only one using prospectively collected data, was nested in a cohort of US army veterans of the Korean and Second World Wars [16]. As conscription was in use during these conflicts, the study population was probably representative of the US male population but less so of women, who comprised

Is high socioeconomic status a risk factor for multiple sclerosis? A systematic review.

High socioeconomic status (SES) is generally associated with better health outcomes, but some research has linked it with an increased risk of multipl...
247KB Sizes 0 Downloads 9 Views