Socioeconomic status and the course and consequences of chronic pain

This article consists of a brief overview of the relationship between chronic pain (CP) and socioeconomic status (SES), a review of recent studies conducted in 2010 and then a discussion of the psychological meaning of the findings in light of wider understanding about SES and health [Bonathan C, Unpublished Data]. Chronic or persistent pain affects people across their lifespan, with significant impact on all aspects of life. It is associated with economic and social disadvantage [1], loss of work [2] and reduced work effectiveness [3].

The risk factors for CP are similar to those for many chronic health conditions [4], but while sex and age are routinely considered, less attention is paid to unhealthy behaviors, such as smoking [5], and to other markers of SES. The biopsychosocial model describes interacting factors in the onset, experience and outcome of CP. It pays particular attention to the role of emotional states, health beliefs and pain-related behaviors [6]. The social aspects are arguably the least explored [3], and this article focuses on an important aspect of the social environment; SES. There are two dimensions: economic (financial and material wealth) and social (such as education and community standing). Socioeconomic factors, such as lower levels of education, low income and unemployment are associated with higher rates of chronic and disabling pain [7]. However, the nature of the relationship between chronic pain and SES is unclear.

Review methods

Given that a thorough review was conducted in 1998 [8], we searched for papers examining the link between chronic pain and SES dated from 1998 to the present time. Databases such as PsycINFO, MEDLINE and Embase were used. Searches for references of relevant papers, combining terms, such as chronic and persistent pain, and all categories of SES were carried out. Of the 143 studies retrieved, 75 measured only partial aspects of SES and/or did not explore its relationship with chronic pain, 39 described treatment outcome, and 22 fell short of the inclusion criteria in other ways. This left seven studies, to which six were added from reference lists of reviews. Adjusting and rerunning the search with terms used in these studies (e.g., ‘disabling pain’) produced no new studies.

Studies were found in journals addressing epidemiology and social science [4], rheumatology and rehabilitation [4], pain [3] and general medicine [2], and which were conducted in Scandinavia or the UK. They focused on pain incidence, prevalence, severity, disability and adjustment, most often for musculoskeletal pain and in community settings. Sample sizes ranged from some 1000 to 10,000, with an outlier at nearly 40,000 [9]. All but one [9], which trawled health records, were recruited by postal survey. All used self-report questionnaires, supplemented in two studies by interview-administered measures and in one by an existing data set.

Does the recent evidence support an association between CP & SES?

Six cross-sectional studies examined the relationship between CP and education, occupational class, employment type, immigration status, housing tenure, car ownership, household income and (in a database from 1968) husband’s occupation. Two studies included area-level measures, such as the Townsend Index (a composite of home ownership, car ownership, unemployment and overcrowding) [10,11]. Both individual- and area-level factors have an impact on health and both can be targets for interventions [12]. Pain definition varied, with pain duration either assessed directly or forming part of a compound index.

The studies showed relative consistency in relationship of pain to markers of occupational class and income [13–15]. Disability, rather than experience of CP, was associated with social disadvantage in middle age [10], and with deprivation [11]. SES was more strongly associated with CP for men than for women. Overall these studies show that individuals of lower SES are more likely to have CP and also to have more severe, disabling pain and more comorbid chronic health conditions.

What factors mediate an association between CP & SES?

Two cross-sectional studies explored mediation [9,16]. Macfarlane et al. found a relationship of pain with adult and childhood SES (although the latter was less strong), which was partly explained by adverse life events and psychological distress, but they did not examine educational variables [16]. Hagen et al. examined an occupationally disabled back pain population and found a strong effect for education as a ‘protective’ factor, even after controlling for the 50% variance in pain attributable to occupational class, working conditions and lifestyle factors [9].

These studies are supplemented by two prospective studies that took both area- and individual-level measurements of SES at baseline [17,18]. A total of 10% of patients developed CP over 1 year and 15% of patients developed CP over 3 years. Both found an effect of SES on the development of CP but, for Davies et al., this effect disappeared after controlling for psychological factors, such as anxiety and depression, illness beliefs, symptoms and threatening life events [17]. The study by Jordan et al. identified an effect of neighborhood general health on the likelihood of developing pain, and suggested that part of this was attributable to shared poor services, such as healthcare [18].

The impact of SES on physiology

The association of CP with a variety of factors suggests that there may be an overarching explanation, and that factors, such as educational level, are merely part of a more complex picture. Marmot identifies a social gradient in health for conditions such as atherosclerosis, coronary heart disease, high blood pressure, diabetes, back pain and depression [19]. He calls this ‘status syndrome’ as it follows SES. Even after allowing for a greater prevalence of ‘risky behaviors’ in populations with lower SES, there is a marked gradient in life expectancies.

Marmot shows that status is the common risk factor by referring to research in primates who show similar gradients of atherosclerosis, despite not smoking or missing appointments with doctors [19]. Low-status results from lack of control over one’s life, greater unpredictability, lower degree of social support and threats to present status through loss of work or benefits. Chronic stress owing to low status causes chronically high levels of adrenalin (epinephrine) and cortisol. These inhibit insulin’s action, raising blood sugar levels and increasing risk of diabetes, while suppressing the immune system. The mechanisms which link status with cardiovascular disease are much better explored and elaborated than those that affect people with CP. We are some way from showing a causal chain from low SES to high levels of CP. We are surely justified, however, in suspecting that such links are there to be found.

How does SES impact on an individual’s ability to live with CP?

Three studies examined and found a relationship between SES and impact of pain [20–22]. Koster et al. reported a relationship of SES with pain disability, but not severity [21]. Brekke et al. found that those living in more deprived areas were less well educated, and (after controlling for pain intensity and disability) were more distressed and had poorer healthcare in which they were less involved [20]. Christensen et al. described participants in the lowest SES who were disabled by pain experiencing less support at work [22]. There were some common themes in findings: problem-solving was reported less by those of lower SES, and strategies for avoiding or minimising pain, including analgesic use, reported more than in higher SES. These findings are subject to the unreliability of self-report on scales.

Conclusion

The finding of an association between lower SES and higher rates of CP with greater disability is robust. In addition to being more likely to develop CP, individuals with lower SES are more disabled and distressed, even when controlling for confounding variables. The relationship at an individual level appeared substantially mediated by psychosocial factors, such as psychological wellbeing, lifestyle factors and working conditions. Large differences in definitions of CP mean that the populations examined may not be comparable. Psychological factors were largely defined in terms of anxiety and depression (scores on self-report questionnaires), producing findings consistent with much other pain literature [6].

Area-level variables were less routinely sampled than individual-level variables, although measures help identify determinants of health inequalities [12], and different markers of SES produce different levels of association with CP. Although studies reviewed here were not restricted to employees, they were conducted in developed countries with state-provided health and education services, so differences are unlikely to be attributable to absolute poverty or lack of healthcare. In addition, similar to the Whitehall study, only part of the differences can be accounted for by lifestyle factors [19].

Therefore, if not poverty or lifestyle, what else makes sense of the consistent association? Marmot [19] and Wilkinson and Pickett [23] propose that social comparison might be an important mechanism by which inequality leads to poorer health as it has already been shown to explain the link between poor health and low job status [24]. Social threats raise anxiety, which is implicated in development of and adjustment to CP [6]. The consequences of feeling socially inferior, living in a less socially cohesive neighborhood with a more imminent sense of threat [25], combined with poorer education and, therefore, poorer job opportunities are likely to interact with the psychological factors known to increase the risk of CP. The concept of control over one’s working and living environment is neglected in this work, although it may well bridge the gap between broader measures (e.g., working conditions) and individual state (e.g., anxiety or depression).

Financial restrictions also constrain coping strategies and since some coping is consistently associated with greater disability [26], this is a self-maintaining process. Educational level may also influence choice of strategies [27]. However, coping should not be considered in isolation from influences on health beliefs and behaviors. Healthcare provision can discriminate against those of lower SES by a less involving consulting style and more restricted decision-making [28], where more participatory consulting might foster acceptance of pain and exploration of self-management options [29].

While the health inequality literature strongly indicates the importance of relative rather than absolute SES, the wealthy may still be more able to purchase elements of protection, such as immediate access to healthcare or to rehabilitation after injury. Wealthier individuals can afford to work shorter hours, while white collar jobs allow more adjustments for those with disabilities to continue at work. This itself protects against development or worsening of other illness [30].

Much remains to be explored and understood at both individual and societal levels about the association between low SES and disabling CP. Overall, the findings of this article suggest opportunities for the prevention and treatment of CP at individual and population levels. Blyth et al. argue that intervening only with high-risk individuals overlooks the opportunity to reduce the overall burden of CP by preventing worsening of those at low and medium risk [3]. Addressing the disproportionate burden of pain in those of lower SES would require directing more resources to those who currently have less. Those at the lower end of the spectrum are more likely to develop CP, be more disabled by it and experience more serious impacts on their lives. This relationship is largely indirect, mediated by a wide range of psychosocial factors. It would be very helpful to see more pain studies using models from the social factors in health literature, in particular those related to control, (un)fairness [24], and social comparison; and to the associated cognitive and emotional processes and their physiological manifestations.