It’s no secret that in the field of pain, we have been living with an unhappy marriage for a while. The specifics of the biomedical disease model is clashed against the inclusive, nebulous biopsychosocial model to explain the things we see. All the while, our patients are stuck in the middle. We are able to specify cause and effect simply and efficiently with the reductionist nature of biomedicine. Comparatively, when our eyes are opened to the vast array of psychological and social variables - previously unaccounted for in healthcare - we are stripped of our hard-won certainty. This uneasy partnership causes clinicians to paralyse in one particular area; socioeconomic status (S.E.S). Research, aside from confirming what we already know - that we have a problem - circumvents the issue, placates a deep-seated aversion to asking hard questions and enables clinicians to continue holding onto biases. The hard questions being how exactly does low S.E.S lead to pain and disability? And how does S.E.S affect our patients? Through this blog I will unpack how our current pain management environment has created paralysing socioeconomic soup and why it is harming our patients. I will explain that research has enabled this inaction and that we need to begin to get specific to rise out of this quicksand, even if our models continue to disagree.
The great socioeconomic pain and disability gradient is one of the mysteries of our time. Social factors are claimed to be the biggest drivers of health inequality (Ref). The biomedical model tends to remove these factors from the equation. Whereas one of the defining features of the biopsychosocial framework is that it includes them. Social determinants are the circumstances in which people are born, grow up, live, work and age, and the systems put in place to deal with illness (Ref). The problem is that, at the individual level, our patients experience poor outcomes as a result of these elements and we feel powerless to take action. When patients come from socially disadvantaged backgrounds they may be treated unfairly or receive different care (Ref, Ref). For all the inclusivity of the biopsychosocial framework it may not transcend into action on the social side of our patient’s lives. We lack clarity and direction.
The role of research
This may have something to do with research failing to define our terms and tell us anything meaningful. S.E.S is a broad term and represents an attempt to capture individual's or group access to the basic resources required to achieve and maintain good health (Ref). Typically S.E.S is commonly conceptualised with the main constructs of income, wealth, work status and education but often becomes conflated with other aspects that determine minority groups (Ref). This broad concept is often measured with composite or proxy scores, within themselves contain fantastically complex constructs; the entire measurement premise lacks precision and reliability (Ref, Ref). An important composite constructs that come readily to mind in S.E.S analysis are wealth and income. The variability of this both within and between countries is mind boggling, as epidemiologist Michael Marmot puts it,
“of course, money could only appear to matter. It may be that poor people have worse health not because they have insufficient money but for some other reason” (Ref).
There are serious questions about the validity of even using this as a marker for S.E.S as there is no standardised way of measuring it (Ref). While some constructs do not change over time - like gender, sexuality and race - others do. Measurement of these constructs is often done once or between two short time intervals, although there are exceptions (Ref). These factors aggregate and interact with each other to produce poor health outcomes among vulnerable populations, but we have a very limited idea about how this happens (Ref).
Research into pain and S.E.S typically assumes the modus operandi of performing large cohort or cross-sectional studies (Ref, Ref, Ref, Ref, Ref, Ref). Although these are potentially useful for defining a problem, but these types of studies are implicitly limp and void of any actionable components; they demonstrate association only. Here are some excerpts from recent reviews;
“This unexplained socio-economic gradient contributes to the confirmation of the social component in a bio-psycho-social model of pain” (Ref).
“However, the nature of the relationship between chronic pain and SES is unclear.” (Ref)
We need to see causal links and researchers seem to be unwilling to ask this simple question. Enabling a stand-off between clinicians about what is and isn’t important for our patients.
So what do we know
What we need to know is how does S.E.S work in pain and disability? Are there causal links between elements of S.E.S and pain? What about disability? If so, which ones? Unfortunately, both pain and disability are generally lumped together in the research and it leaves the picture blurry. And when we attempt to separate specific constructs of S.E.S and determine relationships to pain or disability, we get further confusion.
S.E.S and Pain
Beyond the obvious fact that within countries there is a gradient of health inequity wherein lower S.E.S correlates with increased low back pain, when we look closely at the data there is a different story being told. A recent large cross-sectional analysis (n=27,552) in Europe found ‘wealthier’ countries actually had higher rates of reported pain (Ref). This trend is seen in another very large population survey (SAGE study n = 30,146) across 6 countries ranging from poorer countries like Ghana and India to wealthier countries like Russian Federation and China (Ref). Those poorer countries had lower rates of low back pain reporting (Ref). This observational effect is obviously confounded by unknown sociocultural nuances in reporting of pain and there are severe limitations in the singular timepoint. But still, the question has to be asked what is it about the wealth of a nation that increases pain reporting? What are we actually measuring here?
Although cross-sectional studies can’t ask causal questions, with cohort studies you can get closer. Lallukka et al (2013) reports on cohort data from 2231 young Finns (The Young Finns Study) followed for 20 years. This somewhat controls for the confounding differences in reporting across nations. They at least manage to postulate reasons why lower socioeconomic position as a child continuing into adulthood is associated with worse low back pain (notably there were no differences found with non specific low back pain). The same relationship hasn’t been replicated from other large longitudinal cohorts, so we really have to place caution on these results (Ref). They propose some potential mediating factors to poor outcomes; these are physical factors (biomechanical pathway), health behaviours (mainly via a metabolic pathway) and psychosocial factors (perception of symptoms, coping).
S.E.S and the causal pathway to chronic pain
Christensen et al (2006), in a cross-sectional study, attempt to pin the tail on some other variables that may mediate the pathway from low S.E.S to chronic pain. And importantly, these factors are already known to moderate the relationship between pain and disability. They showed people in lower socio-economic positions report lower levels of problem-solving (but only among women) and higher levels of avoidant behaviours (only men) (Ref). A little like throwing wet spaghetti at the wall and seeing what sticks, these data are reported with the familiar cautionary tagline of ‘association only’. It may fit some biases (not my biases), but these kinds of explanations just cannot be justified through our available evidence yet.
S.E.S and disability
The plausibility of lower socioeconomic status causing increased disability would seem to make some sense. And we have research that demonstrates this may be the case (Ref). Patients with lower S.E.S report higher perceived levels of disability in an Austrian population survey (n=15,474) (Ref). When parsed out for specific elements of S.E.S, education was more associated with perceived disability then followed by income. But we have no idea why low rates of education could be causing perceived disability.
This selective review, although admittedly scraping the surface of the dearth of literature, demonstrates two important points; we have to seriously doubt that low S.E.S can lead to pain and although disability may be more intuitively linked to aspects of S.E.S, the evidence can’t tell us much about how this happens. We are left to posit stories into the data, which confirms our bias and may actually insult our patients. If you explained to your next patient that they were likely to end up with a worse outcome because they probably don’t cope as well as their richer counterparts, you’d better believe they would feel insulted. So we seem to do nothing instead.
Asking causal questions
You may be thinking, this is population level cohort data, we just can’t ask these causal questions from these data. That’s where you’re wrong. Making causal inferences from population level observational data is a burgeoning field in public health and has been around since the 1950s (Ref, Ref). Because, well, it needs to be. No longer will researchers be limp in their approach to these large problems. With better designed cohorts we should be, in fact we need to be, asking these causal questions.
Although it seems like socioeconomic constructs are one big paralysing soup, one which is out of the control of clinicians. We don’t have to be paralysed in the face of this area; we are lucky to have the specifics of biomedicine and we should obviously include these variables as our biopsychosocial model tells us. We need to be specific with S.E.S so that we can take action but first we need to know why and how. Both at the individual and group levels. If we don’t, patients are going to continue to suffer.