Background
One of the most common sleep disturbances among older adults is insomnia [
1‐
5]. Insomnia is defined as a “complaint of insufficient and non-restorative sleep described by the inability to initiate and/or maintain sleep” [
6]. In older adults, the overall prevalence of insomnia ranges from 30 to 48% [
7,
8] with an annual incidence of near 5% [
9]. Furthermore, the prevalence of clinical insomnia in this population is estimated to be over 20% [
10], and the incidence of insomnia symptoms is expected to be even higher among men 85 years and older [
9].
The impact of insomnia is particularly high in the elderly [
2,
11,
14‐
26]. Insomnia incurs substantial adverse consequences for the individual and for society [
8]. Hence, insomnia in older adults is associated with decreased quality of life [
11,
12], impaired concentration and memory [
13], cognitive decline [
14‐
16], increased incidence of medical and psychiatric disorders [
8,
17‐
19], and increased risk of falls, fractures, and mortality [
20‐
22]. Moreover, large societal costs and increased health care resource use have been noted [
8,
23]. For example, an overview study reported that the total direct, indirect, and related annual costs of insomnia in United States were estimated to be between 30 and 35 billion US dollars [
23].
Traditionally, it has been argued that certain alterations in the circadian rhythm are connected to insomnia as people age [
19]. The most significant demographic predictors of insomnia are older age and female sex [
7,
9,
17,
24,
25]. However, insomnia in older adults is probably the result of morbidity rather than aging per se [
26‐
28]. Thus, in the majority of older adults, insomnia is believed to be due to medical disorders (especially chronic pain and respiratory and neurological disorders), psychiatric disorders (especially depression and/or anxiety), and medications (especially anticholinergics and antidepressants) [
8,
17,
19,
26,
27,
29,
30]. Older adults with chronic pain seem to be a group with a high risk of being afflicted by insomnia [
27,
31,
32] with a prevalence of various sleep problems including insomnia ranging from 13 to 62% [
32]. In addition, impaired sleep quality is associated with higher pain intensity [
33,
34]. In this vein, a recent article has suggested that insomnia in older adults should be considered a “multifactorial geriatric syndrome” [
35]. Because sleep complaints are influenced by a combination of medical, physical, cognitive, psychological, and social issues, diagnostic assessments and treatment plans should take into account all these issues rather than considering insomnia as an inevitable consequence of aging [
35].
Considering all of this evidence, it is imperative that investigations of insomnia in older adults should account for level of pain. Although extensive research has been conducted on the relationship between insomnia and pain in older adults, to the best of our knowledge, no study has examined insomnia severity with respect to different pain groups or duration and course of pain. In addition, these studies have produced contradicting results as the definition of insomnia varies between studies [
7,
36,
37]. Some studies have also emphasized that emotional symptoms like depression and anxiety have a stronger association with insomnia in pain populations than the psychical symptoms like the pain itself [
34,
38‐
40]. However, little is known about the relationship between depression, anxiety, and pain symptoms that contribute to insomnia in the elderly people. Hence, this study evaluates the severity of insomnia and its relationship with, age, sex, pain intensity, pain spreading, anxiety, and depression in a large-scale population-based study in older adults with pain (i.e., chronic and subacute pain) and without pain.
Discussion
The main findings of this study were as follows:
-
The overall prevalence of clinical insomnia in the total sample was 20.0%. Older adults with CP had the highest prevalence of clinical insomnia (24.6%).
-
Age and sex were not associated with either total ISI score or ISI categories, regardless of pain group.
-
Higher pain intensity, pain spreading, anxiety, and depression were more common in severe insomnia, moderate clinical insomnia, and sub-threshold insomnia compared to no clinically significant insomnia.
-
The multivariate stratified analyses revealed the following associations: pain intensity, pain spreading, anxiety, and depression were independently related to insomnia in the CP group. Anxiety and depression were independently related to insomnia in the SP group, but only anxiety was significantly associated with insomnia in the NP group.
Overall, 20% of the participants reported clinical insomnia (ISI ≥ 15) [
44], a rate lower compared to that reported in most other studies of elderly populations [
1,
4,
11,
50]. Foley et al., for example, in a large sample of 9000 participants aged 65 years and older found that the estimated insomnia symptoms ranged between 23 and 34% [
4]. Our study found a higher prevalence of insomnia than the findings of an investigation of 47,700 individuals in Norway (i.e., a similar geo-cultural environment to Sweden) where the prevalence rate of insomnia symptoms was 13.5% [
17]. That study, however, did not report results specifically for the elderly. Consistent with previous reports [
24‐
27,
29,
30], we also found that the respondents in CP reported higher prevalence (24.6%) of clinical insomnia than respondents in the SP and NP groups. Indeed, there is evidence that patients with CP tended to demonstrate more sleep fragmentation, longer sleep latency, lower sleep quality, and shorter sleep duration [
54]. Explanations for the observed differences of prevalence estimates may include the great variability of measurements and definitions of insomnia [
7,
36,
51]. Studies using DSM-5 [
52] criteria for insomnia reveal lower prevalence of insomnia in the elderly (4–12%). This discrepancy may also be the result of the age range of the studied population, pain definitions, and socioeconomic and cultural differences of the participants [
7,
17,
24,
35].
In contradiction to some previous research findings [
4,
9,
17,
24,
25], age and sex were not associated with insomnia symptoms or clinical insomnia. Although we found that in the CP group severe clinical insomnia compared to no clinically significant insomnia was more common in the older ages (Table
1, part b), this difference was not confirmed in the regression analysis. Nevertheless, our results are similar to some prospective studies that found no age effect related to insomnia [
53,
54]. Moreover, our results (Tables
2,
3) suggest that insomnia cannot be explained by the chronologically age-related changes in the elderly population, but rather is explained by aging-related changes due to various mental and physical morbidities including pain [
10,
11,
21,
27,
28,
30,
31,
48]. Therefore, as recently suggested, insomnia should be considered a “multifactorial geriatric syndrome” [
32]. Similarly, we found that the total scores of insomnia were higher in women compared to men in the whole sample and in the CP group, but these sex differences were not confirmed in the regression analyses (Tables
2,
3). This finding is consistent with the suggestion that the relationship between female sex and insomnia (especially in the elderly) may be explained by an array of other factors such as depression, chronic diseases, living alone, marital status, occupational status, and social support deficits [
35,
55].
The stratified multivariate analysis of the three pain groups revealed that in the CP group higher levels of pain intensity, pain spreading, anxiety, and depression had direct links to higher severity of insomnia. The most important regressors for both total ISI score and clinical insomnia were pain intensity and anxiety. This result is somewhat consistent with previous reports [
24,
33,
34,
38,
56]. Our results, however, are not in line with some studies that report that insomnia in CP is more strongly related to depression and low mood than to pain intensity [
38,
39], but in line with studies reporting that anxiety is more strongly related to insomnia than to pain intensity [
34,
40]. Generally, the results of the regressions are consistent with previous reports that found strong relationships between both pain and psychological strain and the variability of sleep duration and fragmentation in various populations, including older adults [
17,
24,
29,
34,
57].
In the SP group, both anxiety and depression had clear positive associations with total ISI score and clinically significant insomnia, whereas in the NP group only anxiety was significantly associated with insomnia. Hence, in these two groups, anxiety also had a stronger relationship with insomnia than depression, a finding that is not substantiated in previous studies [
4,
26,
38,
39,
58]. One possible explanation for the stronger impact of anxiety in the present population could be the fact that in elderly depression symptoms are conflated with somatic complaints, and probably those symptoms are being underestimated and unreported by the respondents [
59]. On the other hand, anxiety may more likely be the result of stressful events associated with aging, and perhaps this anxiety increases the likelihood of insomnia via increased levels of arousal [
60]. It is also possible that anxiety and depression are interrelated with insomnia through different paths. According to previous prospective research, anxiety may more likely act as a pathway to insomnia while depression may more likely act as a consequence of insomnia [
61,
62]. In any case, the relationship between insomnia and pain as well as other morbidities is difficult to interpret [
27,
31,
32] and more studies, especially longitudinal studies, concerning the nature and the direction of these associations are needed.
To our best knowledge, this is the first study to evaluate insomnia severity in a large random sample of older adults with and without pain. This study’s limitations include the issues associated with collecting data via a self-reported instrument and the inherent limitations of a cross-sectional study design. In addition, this study evaluated a limited set of variables. We did not examine the role of other factors such as medications, especially hypnotics or other physical comorbidities; life-style factors such as nicotine, caffeine, and alcohol use; or stressful life traumatic events that might be independently related to insomnia [
25].
Conclusions
In conclusion, this study suggests that the relationship between insomnia severity and chronic pain in older adults is very complex, ensuing from pain intensity, pain spreading, anxiety, and depression. In subacute pain, pain symptoms seem to have no effect while a predominance of a psychological strain may account for higher levels of insomnia and clinical insomnia. Conversely, in older individuals without pain, anxiety had a clear positive relationship to both insomnia and clinical insomnia. Taken together, these results indicate that ongoing specific attention from health care providers on this topic is required to ensure that the best possible insomnia treatment modalities are made available to the elderly. For example, approaches including both psychological and pain management components would be beneficial for individuals with CP, whereas psychological approaches targeting depression and/or anxiety might be suitable for individuals who experience short durations of pain or no pain. This study also suggests that a comprehensive assessment including both pain and psychological aspects are essential when older people are seeking primary health care for insomnia complaints. Further research is also needed that defines these multifactorial relations so that elderly patients suffering from insomnia receive the most effective insomnia treatments regardless of their pain level.
Authors’ contributions
L-ÅL, LB, BL, and BG were involved in study conception and study design. ED and BG performed the data analyses and drafted the manuscript. All authors discussed the results and commented on the manuscript. All authors read and approved the final manuscript.