Musculoskeletal pain and co-morbid insomnia in adults; a population study of the prevalence and impact on restricted social participation

Musculoskeletal pain is common in adults and is a frequent reason for consultation to primary health care [1]. One quarter to one third of the general population report low back, hip or shoulder pain and one in five experience chronic pain (i.e. pain that lasts for three months or more) [2, 3]. Musculoskeletal pain impacts on physical and mental health, and mortality risk [4, 5]. Thirty percent of adults in the general population report significant sleep disturbance and 6% to 10% meet diagnostic criteria for insomnia [6], defined as difficulty initiating or maintaining sleep, or nonrestorative sleep, for at least 1 month [7]. Left untreated, insomnia is associated with an increased incidence of depression, anxiety [8] and worse physical health [9]. Insomnia frequently occurs in patients with chronic pain with the prevalence of co-occurrence ranges between 50% and 88% [10, 11]. However, the prevalence and impact of co-existing musculoskeletal pain and insomnia in adults living in the community who may present to primary care is unclear.

Musculoskeletal pain and insomnia have a reciprocal relationship, with each condition increasing the risk of the other which may augment the burden on health and well-being [12]. The mechanisms underlying the association appear to be complex and multi-factoral. Sleep disruption may be attributed to musculoskeletal pain arising from painful stimuli during sleep, which can induce microarousal and increase wakefulness [11]. There is also evidence to suggest that recurrent sleep deprivation and disruption (especially disruption of slow wave sleep) for three consecutive nights or more can decrease an individual’s pain threshold, amplify negative mood and produce somatic symptoms [10, 11].

Social participation involves taking part in social activities and hobbies as well as fulfilling social roles such as being a worker, carer or community member [13]. Maintaining social participation is a clinically important outcome, potentially modifiable by intervention, and a target for intervention in primary care [14]. Restricted social activity is associated with higher rates of morbidity and mortality and lower life satisfaction and health-related quality of life [15, 16]. The aim of this prospective cohort study was to report the prevalence of co-morbid pain and insomnia in community-dwelling adults registered with primary care and its association with restricted social activities.

Of 4757 participants who were eligible to take part, 2662 (56.0%) returned a completed questionnaire at baseline. Compared to participants, non-participants were younger (mean age: 47 years cf. 51 years; p < 0.001) and more likely to be male (52 cf. 42%; p < 0.001). After excluding those who refused further contact (n = 247), decedents (n = 84), those who did not respond at follow-up (n = 340), or had incomplete data (n = 810), 1181 persons were available for analysis at follow up (Fig. 1). Comparison to UK Census data [30] showed the analytical sample to be similar in terms of gender (55.7 female cf. 51.9%) with fewer young adults (19.1% aged 18–34 cf. 29.4%) and more middle aged adults (61.4% aged 35–64 cf. 50.0%).

The study reports the extent of co-occurrence of pain and insomnia in a primary care population, and the extent of its association with problems getting out and about, and engaging in social activities. The findings highlight the common occurrence of musculoskeletal pain and insomnia in adults in primary care patients; almost one third patients reported co-existing pain and insomnia. More than four in every five respondents with insomnia reported pain. Baseline pain and each sleep problem were associated with RSA at 12 month follow-up. The association between RSA at follow-up and pain, difficulty maintaining sleep, early wakening and non-restorative sleep attenuated with adjustment for depression, anxiety, physical health related quality of life and baseline RSA. However, delayed sleep onset was significantly associated with RSA at 12 month follow-up following adjustment for all confounders.

The prevalence of insomnia symptoms in this study is comparable with the 10-40% reported in other population surveys [31, 32]. The prevalence of pain (72%) was high but is comparable to that reported in previous studies of community-dwelling adults [33‐35]. The levels of co-occurring insomnia and pain were comparable with other studies [36, 37]. The association between musculoskeletal pain and RSA and insomnia and RSA have been reported in older adults but not in the general adult population. A systematic review of available literature (details available from the corresponding author) found no comparable population studies that have reported on the impact of estimates of co-occurring pain and insomnia and their impact on restricted social activity.

This study includes a number of strengths and limitations. The 1181 participants were recruited from general practice registers, which provide a sampling frame representative of the general population [17]. Furthermore, the response rates were comparable with other population based prospective studies that have used postal questionnaires. As in all longitudinal studies there was some attrition and missing data. Responders to the baseline survey were more likely to be younger and female. Those included in the analysis, compared to those who responded at baseline but not to the follow-up questionnaire, were less likely to have pain at baseline (p = 0.002) and more likely to be younger (p < 0.001), but they were no more likely to have insomnia at baseline (p = 0.67) or be female (p = 0.89). Non-participation bias may occur but it is unknown if the relationship between pain, insomnia and RSA in those who did not respond is different to those included in the analysis. The generalizability of the data may be limited by the characteristics of the study sample; the area covered by the study is more deprived in terms of health, education, and employment, but has fewer barriers to housing and services, than England as a whole.

The available data covered potential confounders of the relationship between pain and insomnia and RSA, however other potential confounders of the relationship have not been included (e.g. educational attainment). The questionnaires used to measure insomnia, anxiety, depression and RSA have been validated for use in general population samples and in postal surveys. RSA was operationalised as a binary measure, produced by dichotomising the six-level of responses to a single question. A sensitivity analysis, performed to check whether the choice of cut-off affected the relationship between sleep and RSA, found the association was not dependent on the cut-off used to define RSA. Pain was assessed using blank body manikins, a standard data-capture method used in postal surveys. This method has been shown to be a valid and reliable assessment of pain in mid-life adults, although the validity of manikin-derived pain in older people is less clear. High levels of inter-rater reliability for pain scoring (Kappa >0.60) has been demonstrated using this data-capture method [20].

In the pilot work for this study, there were some responders who reported pain interfering with work and but did not indicate pain on the manikin. This may have occurred because pain that interferes with work may not last for one day or longer and would not be reported on the manikin. The high prevalence of pain suggests that recall bias may be present, however the estimates are comparable with other studies that have measured the same phenotype (any pain that have lasted for a day or longer in the last month). It is also possible that pain captured includes that which is not musculoskeletal in origin which would inflate the prevalence estimates, however the majority of pain lasting one day or longer that is reported by adults in primary care is likely to be musculoskeletal in origin [35]. It is also possible that pain intensity may influence the association of musculoskeletal pain upon subsequent social activity restriction, as musculoskeletal pain intensity has been shown to predict greater negative health impact [38]. This study considers pain arising in any part of the body, future research could examine whether the specific location of pain, or pain phenotype (e.g. widespread, regional or none) influence the association with RSA. The co-occurrence of sleep problems was high, with early awakening and delayed sleep onset rarely occurring in isolation (3.7% and 12.3% of individuals reported isolated symptoms respectively). While early awakening was most likely to co-occur with waking several times (93.1% of those with early awakening), delayed sleep onset often co-occurred with each of the other insomnia symptoms (60.7%, 73.0% and 77.0% respectively) suggesting it may be a marker of more severe symptoms. This may explain why delayed sleep onset was found to be most strongly associated with subsequent restricted participation.

This study has implications for primary care clinicians. Firstly, results support the idea that all adults consulting with musculoskeletal pain should be asked about concurrent insomnia, and patients who present with sleep problems should be asked about pain. The study indicates the importance of offering interventions to people where pain or sleep problems are identified. Current management options are limited and tend to target pain and insomnia separately: General practitioners can prescribe analgesics which target pain or, drugs which influence sleep (although these should not be prescribed long-term, due to the risk of addiction) and there is evidence to support such approaches [39]. However, there are known adverse effects to some medications that may limit their usefulness [39]. Psychological therapies and exercise are also known to be beneficial but are not commonly available [40]. A multidisciplinary holistic approach to management is likely to be the way forward [41]; however access to pain clinics which offer such an approach may be limited.

Approaches that reduce pain and insomnia is necessary to offer acceptable treatments for the large number of adults in the general population who experience both symptoms. Primary care clinicians provide support to patients to manage their symptoms and act as gate-keepers to onward referral. They should systematically explore the range of symptoms when a patient presents a single problem, such as pain or sleep disturbance, including symptoms of anxiety and depression. In addition, the clinician should explore social circumstances of the patient, including range of social contacts. The clinician can then give specific advice about the constellation of symptoms and problems that the patient may have, and which are likely to be interacting with each other. Such advice should include education about the possible interaction between sleep disturbance and pain and facilitation of increasing social activity, and engagement with the third sector to promote patient and self-management. Patients receiving interventions that target mood and physical ability have potential to improve social participation which acts as a buffer against morbidities such as cardiovascular disease [42] and maintains healthy ageing [16, 43].

This study highlights co-occurring pain and insomnia as a prevalent problem in the general population, and suggests that those reporting both conditions experience greater impact upon social activity participation than those with only one condition. The study findings suggest co-occurring pain and insomnia may be an important target for clinical trials, rather than considering treatments for isolated symptoms. RSA risk factors which are potentially modifiable and treatable, such as anxiety and depression, present important targets for interventions and should be considered by those commissioning services. Further studies are warranted to determine how best to promote social activity participation in older people with comorbid pain and insomnia.