Multiple sclerosis (MS) is a neurological disorder with variable disease course and diverse symptomatology. Demyelination in the central nervous system (CNS) causes symptoms such as fatigue, weakness and impairment in mobility, sensation, and vision.

Pain, including headaches, lower back pain, neuropathic pain and painful spasms, is a common symptom in people with MS (PwMS), with a wide prevalence range between 30 and 85% (1–4). In the general population, prevalence estimates of 20–30% are commonly reported (5, 6), although prevalence estimates from older populations report up to 72% (7). However, definitions and sample characteristics vary between studies, making it difficult to directly compare prevalence and severity of pain. One case–control study in PwMS found a similar prevalence of pain to controls [Odds Ratio (OR) 0.99–1.13], but showed that PwMS had more severe pain than controls, comparable to people with arthritis (8). Another study showed that both pain severity and interference was greater in PwMS than controls, but these differences were only significant in some, mostly older age groups (9). PwMS more often require pharmacological treatment (10) and pain can significantly interfere with daily activities (11) and employment (12) and affect quality of life (QOL) (13–15).

Damage in the central or peripheral nervous system, inflammatory, and musculoskeletal mechanisms, such as immobilization of parts of the body, may all contribute to pain (4, 16). Exogenous contributors include MS drugs, such as interferon beta, dimethyl-fumarate and long-term steroid treatment (4, 17), although no associations between MS drugs and pain, were found in a large North American study (14).

In MS, pain seems to be part of a complex interplay between other common symptoms or comorbidities, such as fatigue, anxiety, and depression (18, 19), although at least one study suggests the associations between anxiety, depression, and pain is present only in women (10). Bio-psycho-social models, recognizing psychological and environmental factors, have recently been proposed to disentangle the causal pathway between these factors. Different outcome measures have been assessed using comprehensive models, including a recent study showing that pain affects participation in life through fatigue and physical functioning (20). Modifiable lifestyle factors associated with MS onset and progression, such as vitamin D, smoking, stress, diet, physical activity, and body mass index (BMI) (21), have, however, not been studied in relation to pain to great extent.

Our study aimed to determine the associations between substantial pain and modifiable lifestyle factors and to explore the relationship between substantial pain, depression, anxiety, and fatigue in a large international sample of PwMS.

The sample characteristics of the 2,362 participants who completed data items for the substantial pain variable are shown in Table 1, with a full list of country of residence available in a previous publication (22). Unadjusted univariable analysis showed that several socio-demographic variables (Table 1), and health outcomes (Table 2) were associated with substantial pain.

In addition, we explored the association between several modifiable lifestyle factors as exposures and substantial pain as the outcome in a multivariable regression model adjusted for age, gender, duration of MS and level of disability. Table S2 in Supplementary Material shows the lifestyle factors for the sample. Participants who smoked (compared with those who never smoked) or were obese (compared with those with normal BMI) were twice as likely to report substantial pain. Those with moderate or high alcohol use (compared with low) and those with moderate levels of physical activity (compared with low) were 1.4–1.5 times less likely to report substantial pain, while those reporting high physical activity were 1.7 times less likely to report substantial pain; every 10 point increase in healthy diet score was associated with 1.2 times lower odds of substantial pain (Table 3).

Pain is common in PwMS and contributes significantly to disability (11, 12) and worse QOL (13, 14). Our results showed that those with substantial pain were more likely to be diagnosed longer ago and had a higher level of disability and more relapses. Most studies report associations between the likelihood of experiencing pain and pain intensity with longer disease duration, more progressive forms of MS and level of disability (14, 23, 24), while some do not (3, 25, 26). Inconsistent findings of associations between pain and relapses have also been reported, with some showing associations of pain with higher relapse rate (23), while others report increased relapses associated with lower prevalence of pain (26). In univariable analyses, we found a reduced risk of pain in people with relapsing remitting MS (RRMS), older age and female gender, in line with others (14). Age, gender and years since diagnosis were, however, no longer significantly associated with substantial pain in a multivariable regression adjusting for modifiable lifestyle factors, while disability level was still strongly related to pain.

Reduced mental health QOL, fatigue, depression, and anxiety had substantial adjusted associations with substantial pain, in line with previous literature (2, 15, 18, 20, 27). Pain is a multi-factorial symptom, with clear physical correlation in PwMS with CNS demyelination and degeneration (28), but also contributed to by psycho-social factors such as personality, depression, anxiety, and comorbid conditions, such as fatigue (18). The interplay between these factors determines the perception of pain. The presence of pain in its own right may, however, also lead to some of these comorbidities, as may poorer lifestyle behaviors, which in turn may be maintained due to the presence of pain (e.g., physical inactivity, and smoking).

Our study identified important associations between the experience of substantial pain and modifiable lifestyle risk factors. Current smokers were twice as likely to report substantial pain in our study, in line with other studies (29). Another study showed that current smoking increased the likelihood of pain in those reporting depression (30). The association between smoking and pain has been summarized as a positive feedback loop where smoking may be a risk factor for pain, while pain is a powerful motivator to continue smoking due to anti-nociceptive effects (31). This feedback loop may be especially hard to break for people experiencing depression or anxiety who may have increased pain sensitivity and negative affect when attempting smoking cessation (32). Clinicians should be aware of this interplay when managing MS symptoms. Promoting smoking cessation as part of a secondary and tertiary management strategy is, however, vital since smoking has been shown to play a significant role in disease progression (33), although this may be more difficult in people experiencing pain or depression.

Similarly, we showed a significant association between more physical activity and lower odds of experiencing substantial pain, in line with previous literature (34). While people may be less likely to engage in physical activity due to pain, there may well be a role for physical activity in pain management in PwMS (35). Physical activity is known to increase pain threshold and tolerance (36). In a study of older adults, those with chronic pain had lower levels of physical activity, and fear-avoidance believes played an important role in predicting future physical activity levels, over pain characteristics (37). Others have shown that psycho-social factors, including negative beliefs about pain and avoidance of activity are likely to influence pain severity and interference (38). Promoting physical activity among PwMS is of importance given the wide-ranging benefits of exercise in potentially modifying the course of the disease (39) and reducing fatigue and depression (40–42); however, consideration of psycho-social barriers to increasing physical activity is essential.

Poorer diet and obesity were associated with higher odds for substantial pain in our study. Little is published about this association in humans but there is some evidence from animal studies that a high-fat diet can contribute to pain (43, 44), and it is plausible that poor diet contributes to systemic inflammation and increased pain sensitivity (45). Poor diet may lead to becoming overweight or obese which in itself is associated with increased likelihood of pain (30, 46). Weight loss can improve pain symptoms (47, 48), although pain can also be a barrier to weight loss (49). A population-based cohort study showed that being overweight or obese was associated with reporting future pain, while pain increased the odds for future obesity, with smoking confounding this relationship (19). As with diet, chronic inflammation is likely implicated in this interplay between physical activity, weight, and pain (50). Obesity has become a focus in secondary prevention in MS, with clear associations with poorer health outcomes (51). Our findings suggest that moderate alcohol use need not be considered negative for PwMS, in line with previous research (52).

Based on our data, it is unclear whether these modifiable lifestyle factors directly contribute to development of substantial pain in MS, or indirectly through fatigue, anxiety, and/or depression; or whether pain is contributing to fatigue, depression and poorer lifestyle behaviors. Future studies could further investigate these associations by attempting to clarify temporality through longitudinal data, and investigation mediating and moderating factors involved. Future studies should also record the type and potentially the cause of pain, to further understand the underlying mechanism.

Depression, smoking, physical activity, and being overweight have previously been reported as modifiable risk factors to be considered in the prevention and management of pain in other populations (30, 53). It is likely that a comprehensive secondary and tertiary prevention strategy for pain (and health outcomes in general) in PwMS would include advice about lifestyle factors such as diet, weight, physical activity, and smoking (21). Rehabilitation services often address lifestyle factors that can minimize the impact of pain in PwMS (54). However, pain, psycho-social factors, and common symptoms or comorbidities such as depression, anxiety, and fatigue, may hamper efforts to take up or maintain healthy behaviors. Holistic strategies to overcome these barriers and break the feedback loop need to be considered.

Pain is common in PwMS, with substantial pain present in over a quarter of our study population. Fatigue, depression, reduced mental health QOL, and anxiety were strongly associated with the presence of substantial pain. For the first time, we showed that tobacco smoking, obesity, lack of exercise, and poor diet were also associated with pain using multivariable analyses. These modifiable factors have previously been suggested to be targets for management of pain in other populations and in the secondary prevention of MS. Further research is needed to disentangle the complex interplay between these lifestyle factors, symptoms, and comorbidities, which may contribute to pain, or may be caused or maintained by the presence of pain.

The Health Sciences Human Ethics Sub-Committee at the University of Melbourne provided ethical approval for the study (Ethics ID: 1545102). Individuals were provided with participant information and could not enter the survey until they confirmed they were 18 years or older and consented to participate.

CM, TW, and GJ were involved in conceiving the study. CM was involved in data collection, and CM, CB, and AL contributed to cleaning and analyzing the data. CM drafted the manuscript with the help of PJ and FK. All authors edited and contributed to the final version and approved the manuscript.

GJ receives royalties for his books Overcoming Multiple Sclerosis and Recovering from Multiple Sclerosis. GJ, SN, and KT have received remuneration for conducting lifestyle educational interventions for PwMS. All other authors declare no conflict of interest.