A comparative analysis of the prevalence and predictors of chronic pain in older adults with and without intellectual disability in Australia

ABSTRACT

Background

There is little research comparatively assessing prevalence of pain between older people either with or without intellectual disability. This paper explores health and social factors associated with chronic pain in these two groups.

Method

A cross-sectional survey was undertaken in New South Wales and Queensland, Australia. Inclusion criteria were adults either with or without intellectual disability, aged 60 years and older, and currently living in community-settings. Univariate and multivariable analyses were undertaken on a sample of 391 adults with intellectual disability and 920 adults without intellectual disability.

Results

Key findings included higher prevalence of pain in the intellectual disability group, along with higher rates of osteoarthritis, falls, oral health problems, and mood disorders.

Conclusions

Mitigating risk factors for conditions that cause chronic pain in older adults is crucial. As longevity increases, the healthcare sector needs to prioritise chronic pain management for people with intellectual disabilities through appropriate treatment strategies.

KEYWORDS:

• Ageing
• Australia
• chronic pain
• intellectual disability
• well-being

Chronic pain in older people

Chronic pain at any point across the lifespan is of significant concern for individuals and the wider health system, but in older adults it significantly affects their perceived quality of life and general wellbeing (Ferretti et al., 2018). In recognition of this, the World Health Organization (WHO) Rehabilitation Strategy 2030 placed considerable emphasis on better understanding and recognising chronic pain (WHO, 2017). Whilst there is still debate regarding a definition, the International Association for Study of Pain (IASP) consensus statement describes chronic pain as including symptoms of pain that are either ongoing or recurring for a period of longer than three months (IASP, 2022).

Prevalence figures for chronic pain in older people vary considerably across different studies, depending on the health conditions, age ranges, and sex. Data from European large-scale studies have reported prevalence of chronic pain at 28% in people aged 60 years and older (Breivik et al., 2006) and 38% in adults aged 65 years and older (Larson, 2017). Similarly, the Australian Institute of Health & Welfare (AIHW) reported approximately that 20% of people aged 55–84 years had chronic pain, with higher proportions of women than men experiencing chronic pain (AIHW, 2020). Findings from the Australian nationwide General Practice study, Bettering the Evaluation and Care of Health, which ran for 18 years from 1998 to 2016, reported similar outcomes with the prevalence of chronic pain being 30% in adults aged 65–74 years and 36% in adults aged 75 years or older (Henderson et al., 2013).

The impact of chronic pain on individuals’ lives can occur at multiple levels: they include specific functional limitations that impact on activities of daily living (Breivik et al., 2006), disturbed sleep patterns (Finan et al., 2013) and impaired quality of life (Hadi et al., 2019). These issues can then affect the capacity of older people to remain living independently in community settings (Eggermont et al., 2014). The main predictors of chronic pain involve physical illness, including musculoskeletal conditions such as osteoarthritis and osteoporosis (Stubbs et al., 2014). It has also been found that adults with chronic pain experience high levels psychological distress, which can then in turn worse their pain (Lerman et al., 2015).

Chronic pain in older individuals with lifelong intellectual disability

There is dearth of literature regarding prevalence of chronic pain amongst older adults with lifelong intellectual disability (Barney et al., 2020). The increasing life expectancy and ageing of people with intellectual disability (e.g., Coppus, 2013) highlights the need to examine potential predictors and correlates of chronic pain in this highly vulnerable population sub-group in order to better support their healthcare needs (Hussain et al., 2019). There have been disagreements regarding whether people with intellectual disability have a higher pain threshold than people without intellectual disability, or if they even experience pain in the same way (Beacroft & Dodd, 2010). Further, there is variation in reported and perceived levels of pain intensity amongst people with intellectual disability depending on the degree of cognitive limitation, communication impairments associated with their intellectual disability, or whether data were provided by the affected individual or by a proxy respondent (de Knegt et al., 2013; Defrin et al., 2006; McGuire & Kennedy, 2013; Wark et al., 2013, 2015).

The personal impact of chronic pain needs to be understood using a life-course perspective. The reasons for chronic pain are multifactorial, and, for older people with intellectual disability, their chronic pain may result from reduced opportunities arising earlier in life. Examples could include limited healthy food choices and restricted options for physical activity leading to a sedentary lifestyle and increasing levels of obesity at much younger age, as well as the earlier occurrence of physical and mental health conditions with higher levels of comorbidity (Hussain et al., 2020; McCarron et al., 2013). People with intellectual disability are also at higher risk of falls due to gait and balance issues as well as accidental injury, more so as they age (Enkelaar et al., 2012), and which in turn can result in both acute and potentially chronic pain.

People with intellectual disability in some settings were traditionally not proactively treated for pain – in spite of diagnosed health morbidities that would indicate a high likelihood of associated pain (e.g., Wark & Kingstone, 2019) – with this failure possibly due to the concerning perception that they simply didn’t experience pain (Beecroft & Dodd, 2010). However, it is not clear whether older people with intellectual disability experience pain at similar levels, and from similar causes, to the wider ageing population. To the best knowledge of the authors, there are few contemporary research studies that have undertaken a concurrent comparative assessment of older Australians with and without intellectual disability in relation to prevalence and correlates of pain. This paper aims to provide knowledge in this limited research area by examining the physical and mental health and social factors associated with chronic pain in two groups, older community-residing individuals either with or without lifelong intellectual disability, with the arising knowledge helping to inform future healthcare delivery for people with intellectual disability.

Methods

Overview

The project utilised a cross-sectional survey with a sample composed of people either with or without lifelong intellectual disability. The regions of interest included all metropolitan and rural regions of New South Wales (NSW) and Queensland. A specific focus was on ensuring rural participants, as it is known that intellectual disability research often fails to include this group (Wark, 2018). The inclusion criteria for the two groups were the same; aged 60 years and older and currently living in community-settings. Small group homes of up to five people with intellectual disability were included, but large congregate-care services, such as residential aged-care facilities, were excluded for both cohorts as their health support is structured differently than for people living in the community and therefore pain management regimes may be different. For both groups, the ability to give informed consent for participation was considered essential to overcome problems regarding the use of proxy-respondents. It was recognised in advance that this choice may result in the potential exclusion of respondents with moderate-to-severe cognitive impairments, but no decision was made to include or exclude based solely on a previously assessed level of intellectual disability (i.e., mild; moderate or severe/profound) or dementia.

A steering committee comprising eminent disability and aged-care experts, including people with intellectual disability, oversaw the project and implementation. All survey protocols, including the pilot instruments, the final survey pro-forma and the recruitment strategies, were formally approved by the University of New England’s Human Research Ethics Committee (Approval number: HE14/019), as well as being reviewed by the project’s industry partners.

Recruitment

People with intellectual disability: A list of all disability agencies in the Australian Disability Enterprise directory who operated in metropolitan or rural regions of either/both NSW and Queensland (n = 767) was established. Consultation with industry partners identified 568 organisations as being potentially relevant, and these were individually approached to assist with recruitment of respondents. As the target population in these services were over 60, there was a strong trend for individuals to have been supported by the same organisation continuously for many years, and assessments of intellectual disability level (e.g., mild; moderate; severe/profound) were either not available or were decades old and potentially out-dated. Of the 568 organisations, 209 reported having eligible participants that met the inclusion criteria.

To assist their participation, individuals with intellectual disability were facilitated to complete the survey form through a face-to-face interview with trained project interviewers. The person with intellectual disability could opt to have a support person of their choosing present in the interview to assist with answering questions, particularly in relation to clarifying exact timeframes and/or medical diagnoses. Ultimately, 399 participants completed the survey; another 53 initially planned surveys did not take place either due to participant illness on the day or if it became apparent that the individual did not meet eligibility criteria (e.g., was not able to provide informed consent or was below 60 years of age).

People without intellectual disability: Since potential participants resided independently in community-settings, recruitment was open-ended. Information was sent to Independent Living, Home Care, and ageing service coordinators, was posted on local print media, social media and community noticeboards, and promotional interviews were held on local radio stations across the two states. A toll-free phone number, website and email were provided for potential participants to seek more information. Participants were given the choice to complete the survey in either the hard or electronic format; survey forms were sent with a reply-paid envelope and it was also made available online, with 1020 respondents ultimately completing the survey.

Instrument & measures

A customised survey questionnaire was purposively-developed to facilitate direct comparisons between the two groups. This development was an iterative process that included several stages of review and pilot-testing. Chronic pain was pre-defined as the presence of “pain” of three months or more in the 12-months preceding the survey. This definition allowed distinction from acute pain for the respondents and aligns with the 2022 IASP definition for chronic pain.

The main domains were demographic and socio-economic information; health status included list of health condition adapted from a well-documented tool for health screening for people with intellectual disability (Esralew et al., 2017), and the Medical Outcome Study Short Form (MOS-SF) 12, henceforth referred to as SF-12 (Ware et al., 1996). The SF-12 scale has two validated sub-scales, the PCS (physical component score) and MCS (mental component score), which are computed based on internationally validated algorithm. The PCS assesses energy, mobility, activity, and pain, while the MCS assesses mood-associated issues. Scores range from 0–100 with higher scores indicative of better self-reported physical and mental health.

Wellbeing was assessed using the validated Personal Wellbeing Index scale developed by Cummins & Lau for Adults, the PWI-A (International Wellbeing Group, 2013), and its customised version for people with intellectual disability, the PWI-ID, which includes a pre-test for comprehension (Cummins & Lau, 2005). There are 11 domains in the PWI scales using a 10-point Likert-type scale of which eight are used in the construction of overall PWI score: overall satisfaction, satisfaction with standard of living, health, life achievements, personal relationships, sense of safety, feeling part of community, future security. The three additional domains of the PWI scale focus on religion and spirituality. Cummins and Lau (2005) noted the exclusion of these domains does not yield psychometrically different data.

Social support was measured using the Duke Social Support Index (DSSI) (Koenig et al., 1993). The DSSI scale scores are analysed by segmenting into two validated subscales: social interaction (size and frequency of social network) and satisfaction (degree of satisfaction with social networks and interactions). The 13-item Adverse Life Events scale was used to measure a range of adverse events in the preceding three years (Lee et al., 2005), and has been used previously in Australian settings (Powers et al., 2004; Sharpley et al., 2015). It includes a range of adverse events from serious and major personal and family illness, death, and loss of a loved one and being physically assaulted.

Data collection, editing & analysis

The surveys by the participant group without intellectual disability were completed in either a print or online format. The online version was through SurveyMonkey (www.surveymonkey.com), while data from the paper-based forms were subsequently entered into SurveyMonkey by the project team. Data from the face-to-face surveys with people with intellectual disability were manually checked by interviewers at the conclusion of the interview, and then entered into SurveyMonkey.

After data collection was completed, the data from both surveys were exported to SPSS (IBM Corporation, 2020). Preliminary audits were conducted in SPSS to check veracity of data and undertake cross-validation. Identified errors were then checked against the paper-based copies to minimise data entry errors. Survey data where information was missing on age or postcodes were removed as it was not possible to ascertain whether they met the inclusion criteria. Overall, 10% (n = 100) of the survey forms for people without intellectual disability had missing or inaccurate data on age and/or postcode compared to 4.0% (n = 8) of surveys from people with intellectual disability. This resulted in a final analysis sample of 391 (399 initial) people with intellectual disability and 920 (1020 initial) people without intellectual disability.

The de-identified data for SF-12 and DSSI scale were sent under formal agreement to a national statistical consulting agency for computation of sub-scales using established international algorithms for data imputation and re-categorisation. For this paper, univariate and bivariate analysis was undertaken, using statistical methods such as chi-square test, t-tests, and ANOVA. Before undertaking multivariable logistic regression, correlational analysis for variables of interest was undertaken to check for multicollinearity. Only variables with a correlation coefficient of less than 0.5 were included in the regression models. All statistical analyses were undertaken using SPSS.

Results

Demographic characteristics

The main demographic characteristics of the two study samples are presented in Table 1 below. Respondents with intellectual disability were younger (mean age 65.2 years, SD, 4.4) than those without intellectual disability (mean age 71.9 years, SD, 7.2). The sex ratio was considerably different with more males (62.7%) in the intellectual disability group and more females (67.9%) in group without intellectual disability. This was perceived to be potentially due to enrolment factors as disability agencies had more men registered with their various services; the steering committee confirmed this pattern. There was no remarkable difference observed for geographic location across the two samples, with the proportions were reflective of the national metropolitan versus regional/rural population distribution.

Table 1. Demographic characteristics of study participants in older people with and without intellectual disability.

Characteristics	People with intellectual disability (n  =  391)	People without intellectual disability (n  =  920)
	% (n)	% (n)
Age in years – Mean & (SD)	65.2 (4.4)	71.9 (7.2)
Sex
Female	37.3 (146)	67.9 (625)
Male	62.7 (245)	32.1 (295)
Employment
FT/PT	53.7 (211)	16.9 (155)
No/Retired	46.3 (181)	83.0 (764)
Location
Metro	60.4 (236)	52.2 (516)
Rural	39.6 (155)	40.7 (403)
Employment
FT/PT	53.7 (211)	16.9 (155)
No/Retired	46.3 (181)	83.0 (764)
Financial Vulnerability
≥ −−2	30.4 (119)	20.5 (188)
−1	39.8 (155)	66.0 (607)
0	21.9 (85)	10.4 (96)
≥1	6.7 (30)	3.2 (29)

Approximately half of the sample of people with intellectual disability remained in some form of employment, whereas most of the group without intellectual disability had retired. This difference likely relates to lack of post-employment retirement programs for people with intellectual disability in Australia, with individuals remaining in the workplace past retirement age due to the scarcity of alternative support options (Putnam & Bigby, 2021). However, the income obtained from such employment was often minimal and most of respondents in the intellectual disability group received disability pensions and had no meaningful assets, such as home-ownership. While most individuals without intellectual disability did receive age pensions, proportionately more of this group lived in their own homes [which are not means-tested for age pension eligibility]. The financial vulnerability composite variable showed that nearly 70% of people with intellectual disability and over 86% of the group without intellectual disability experienced financial hardship.

Prevalence and correlates of chronic pain

The overall prevalence of reported pain across the two groups varied slightly. The proportion of adults with intellectual disability experiencing chronic pain was approximately 27%, compared to 24% in the group without intellectual disability (see Table 2). For each group, the association of chronic pain with known health conditions, which can either be the source of chronic pain or can exacerbate pain, was examined. These conditions included: osteoarthritis; oral health; musculoskeletal pain associated with previous fall/s; diabetes; cancer; and chronic insomnia, as well as anxiety and depression.

Table 2. Prevalence and univariate predictors of chronic pain by various health conditions in older people with and without intellectual disability.

Characteristics	People with intellectual disability (n = 391)	People without intellectual disability (N = 920)
	% (n)	% (n)
Chronic pain	26.9 (105)	24.1 (222)
Osteoarthritis	29.2 (114)^a	24.1 (222) ^a
Recent history of recent falls	29.2 (114)^a	9.5 (87)^a
Gum problems	9.0 (35)^a	4.9 (45)^b
Diabetes	25.6 (100)^a	12.7 (117)^n/s
Cancer	10.5 (41)^n/s	8.3 (76)^n/s
Anxiety	22.0 (86)^a	16.0 (147)^a
Depression	22.0 (86)^a	15.3 (141)^a
Sleep	22.3 (87)^a	27.2 (250)^a
SF-12_PCS (Mean, SD)^§	45.8 (9.9)^a	43.6 (11.9)^a
SF12_MCS (Mean, SD)^§	50.4 (9.9)^a	51.5 (10.0)^a

Notes: Statistically significant difference within each cohort for difference by prevalence of chronic pain: a = p < 0.001, b = <0.05, n/s = non-significant. For continuous variables denoted by ^§, statistical difference was assessed using ANOVA.

The prevalence of osteoarthritis in people with intellectual disability was higher (29.2%) compared to the group without intellectual disability (24.1%), with statistically significant univariate association within each group between osteoarthritis and chronic pain (p < 0.001). History of recent falls was comparatively higher for people with intellectual disability (29.2%) to people without intellectual disability (9.5%), which could be possibly associated with issues of poor balance, gait, and/or side effects of medications in this group. However, despite the difference in prevalence of falls, the univariate association with chronic pain was statistically significant in both groups. Similarly, gum problems were more common in people with intellectual disability (9.0%) compared to the group without intellectual disability (4.9%), with statistically significant univariate association with chronic pain in both groups. A quarter of people with intellectual disability had diabetes (25.6%), whereas the proportion of the group without intellectual disability with diabetes was nearly half at 12.7%. The univariate association with chronic pain was statistically significant only for the non-intellectual disability group (Table 2). Cancer prevalence in both groups was low, which could be attributed to a selection bias amongst the respondents as individuals with significant health issues potentially likely to not opt to participate, and showed no statistically significant association with chronic pain.

In addition to physical health concerns, a separate analysis for anxiety and depression was undertaken. The proportion of people with intellectual disability experiencing anxiety was 22.0%, and those having depression was also 22.0% [the similarity in figures is coincidental as on multiple checks no data errors were found]. For the group without intellectual disability, the prevalence of anxiety was slightly higher (16.0%) than depression (15.3%); and univariate analysis showed statistically significant association between anxiety and chronic pain; and depression and chronic pain (p values <0.001). Highly statistically significant univariate associations between anxiety and chronic pain; and depression and chronic pain (both p < 0.001) were seen for the group without intellectual disability. The final health condition included in the current analysis was insomnia (poor sleep or unusual sleep patterns). A higher proportion of people without intellectual disability had insomnia (27.2%) compared to people with intellectual disability (22.3%). Univariate associations between insomnia and chronic pain were found to be statistically significant for both groups.

The association between the composite sub-scales of the MS-12 health measure (PCS and MCS) were assessed. For PCS, people with intellectual disability had slightly higher values (Mean = 45.8, SD ± 9.9) compared to the group without intellectual disability (Mean = 43.6, SD ± 11.9). The corresponding values for MCS were Mean = 50.4, SD ± 9.9 and Mean = 51.5, SD ± 10.0, respectively (see Table 2). Univariate analysis showed statistically significant association between PCS values and chronic pain, and for MCS values and chronic pain, for both groups.

The findings for social factors associated with pain – the social interaction and satisfaction sub-scales of DSSI, the cumulative score for PWI-ID and PWI-A wellbeing scales, and adverse life events – are presented in Table 3 below. People with intellectual disability had lower scores for social interaction (Mean 7.3, SD ± 1.6) than did the group without intellectual disability (Mean 8.2, SD ± 1.9). At the univariate level, the association of social interaction for the group without intellectual disability was statistically significant, but this was not the case for people with intellectual disability (Table 3). Although satisfaction with social support was higher for people without intellectual disability (Mean 17.9, SD ± 3.2), the difference in scores between the groups was small, with corresponding values for the group with intellectual disability marginally lower (Mean 17.7, SD ± 3.0). Like the results for the DSSI interaction scale, the univariate association between social satisfaction and chronic pain was only statistically significant for people without intellectual disability.

Table 3. Prevalence and univariate predictors of chronic pain by social indicators in older people with and without intellectual disability.

Characteristics	People with intellectual disability (n = 391)	People without intellectual disability (N = 920)
Adverse Life Events (Mean, SD)^§	2.2 (1.5)^a	1.5 (1.3)^a
DSSI – Interaction subscale (Mean, SD)^§	7.3 (1.6)^n/s	8.2 (1.9)^a
DSSI – Satisfaction subscale (Mean, SD)^§	17.7 (3.0)^n/s	17.9 (3.2)^a
Personal Wellbeing scale (Mean, SD)	55.5 (9.1)	50.8 (12.7)^a
Adverse Life Events (Mean, SD)^§	2.2 (1.5)^a	1.5 (1.3)^a

Notes: Statistically significant difference within each cohort for difference by prevalence of chronic pain: a = p < 0.001, b = <0.05, n/s = non-significant. For continuous variables denoted by ^§, statistical difference was assessed using ANOVA.

The results for personal wellbeing scores for the two groups are also presented in Table 3. Unlike the results described above for DSSI sub-scales, people with intellectual disability had higher personal wellbeing score (Mean 55.5, SD ± 9.1) compared to the group without intellectual disability (Mean 50.8, SD ± 12.7). Univariate results for association with chronic pain was not statistically significant for people with intellectual disability, whereas for the group without intellectual disability it was highly significant (p < 0.001). People with intellectual disability experienced more adverse life events (Mean 2.2, SD ± 1.5) compared to the group without intellectual disability (Mean 1.5, SD ± 1.3). Univariate analysis showed that for both groups the association of adverse events with chronic pain was highly significant (p < 0.001).

The results of the multivariable logistic regression for both groups are presented separately. The outcome variable was chronic pain coded as a dichotomous variable. As mentioned earlier, correlation analyses were conducted to check for multicollinearity between variables presented in Tables 1–3. None of these variables had a correlation coefficient higher than 0.4. Therefore, regression analyses for each group proceeded, except financial vulnerability due to small number of cases across various categories. As there was variation in age and sex profile of the respondents in the two groups, both these variables were included in the regression models as adjustment / control variables. Rurality was included as a proxy for gaps in health service provision. All the variables listed in Tables 2 and 3, apart from cancer, were included. Results for statistically significant variables are presented as adjusted odds ratio (AOR) with corresponding 95% confidence intervals (CI) and p values.

At the multivariable level, none of the demographic variables such as age, sex, employment, and rurality were found to be statistically significant (see Table 4). Among the health conditions, there was a strong association between osteoarthritis and chronic pain after adjusting for other health conditions for people with intellectual disability (AOR: 3.8, 95% CI: 2.1–6.9, p < 0.001). This was similar to the group without intellectual disability; although the value of AOR was comparatively lower, the association between osteoarthritis and chronic pain was statistically significant (AOR: 1.95, 95% CI: 1.29–2.96, p = 0.002). The other musculoskeletal condition, a recent history of falls, was not found to have a statistically significant association with chronic pain for respondents in either group.

Table 4. Multivariable regression for predictors of chronic pain in older people with and without intellectual disability.

	People with intellectual disability	People without intellectual disability
Variable	Adjusted OR (95% CI)	AOR (95% CI)
Age (in years)	1.04 (0.94–1.07)^n/s	0.98 (0.95–1.01)^n/s
Sex
Male	Ref	Ref
Female	0.84 (0.47–1.52)^n/s	1.3 (0.85–1.99)^n/s
Location
Metro	Ref	Ref
Rural	1.05 (0.59–1.88)^n/s	0.93 (0.64–1.37)^n/s
Employment
None/Retired	1.13 (0.62–2.06)^n/s	0.88 (0.48–1.60)^n/s
Employed	Ref	Ref
Osteoarthritis
No	Ref	Ref
Yes	3.85 (2.13–6.65)^c	1.96 (1.29–2.96)^b
Recent history of falls
No	Ref	Ref
Yes	1.07 (0.56–2.40)^n/s	0.84 (0.47–1.59)^n/s
Sore gums
No	Ref	Ref
Yes	2.87 (1.15–7.19)^a	0.79 (0.36–1.74)^n/s
Diabetes
No	Ref	Ref
Yes	0.86 (0.46–1.62)^n/s	1.65 (0.97–2.80)^n/s
Anxiety
No	Ref	Ref
Yes	1.68 (0.87–3.25)^n/s	1.72 (1.11–2.97)^a
Depression
No	Ref	Ref
Yes	1.16 (0.56–2.40)^n/s	2.00 (1.13–3.58)^a
Sleep
No	Ref	Ref
Yes	1.51 (0.78–2.95)^n/s	0.79 (0.51–1.23)^n/s
SF-12_PCS	0.90 (0.87–0.94)^c	0.92 (0.89–0.93)^c
SF12_MCS	0.95 (0.92–0.98)^b	0.96 (0.93–0.98)^b
DSSI – Interaction subscale	1.09 (0.92–1.30)^n/s	1.01 (0.90–1.13)^n/s
DSSI – Satisfaction subscale	1.03 (0.93–1.15)^n/s	0.92 (0.86–0.99)^a
Cummins Wellbeing scale	1.02 (0.93–1.15)^n/s	1.02 (1.01–1.04)^a
Adverse Life Events	1.10 (0.90–1.34)^n/s	1.14 (0.97–1.33)^n/s

Notes: Ref = Reference category; Statistical Significance: a = p < 0.05 b = p < 0.01 c = p < 0.001.

Oral health was significantly associated with chronic pain for people with intellectual disability, with those adults reporting sore gums being twice as likely to have chronic pain (AOR 2.8, 95%CI: 1.14–7.1, p < 0.05). The small number of cases is probably the most likely explanation for the wide range of the 95% confidence interval. In comparison, there was no statistically significant association between sore gums and chronic pain for group without intellectual disability. Diabetes was not statistically significant at the multivariable level after adjusting for other health conditions, though the p values were closer to being statistically significant for the group without intellectual disability (Table 4).

For mental health conditions, although the value of AOR was elevated for association between anxiety and chronic pain for people with intellectual disability, the results were not statistically significant. However, for the group without intellectual disability, anxiety was significantly associated with chronic pain (AOR: 1.71, 95%CI: 1.01–2.91, p < 0.05). A similar pattern of non-statistically significant results was observed for the association between depression and chronic pain for people with intellectual disability, whereas for the group without intellectual disability the risk ratio was elevated and statistically significant (AOR: 2.00, 95%CI 1.28–3.57, p < 0.05). In both groups, there was no statistically significant association between insomnia and chronic pain.

The association between short-term measures of physical functioning, i.e., SF-12-PCS, was statistically significant for chronic pain in both groups: people with intellectual disability (AOR: 0.90, 95% CI 0.87–0.93, p < 0.000) and without intellectual disability (AOR: 0.91, 95%CI: 0.89–0.93, p < 0.000). A similar pattern was observed for SF-12 MCS scores in both groups: people with intellectual disability (AOR: 0.95, 95% CI: 0.92–0.93, p < 0.01) and without intellectual disability (AOR: 0.95, 95%CI: 0.93–0.98, p < 0.01). In summary, as PCS and MCS score values increase, the association with chronic pain decreases, and this pattern holds for people with and without intellectual disability.

The two DSSI subscales measuring the degree of social interaction and satisfaction with social interaction were not observed to have any statistically significant association with chronic pain for people with intellectual disability. However, for the group without intellectual disability, while the DSSI interaction scores were not significantly associated with chronic pain, the DSSI satisfaction subscale scores did show a statistically significant association (AOR 0.92, 95%CI: 0.86–0.99, p < 0.05). Similarly, in the multivariable association between personal wellbeing scores and chronic pain, there was variation across both groups. For people with intellectual disability, the association was not statistically significant, whereas for the group without intellectual disability there was a marginally significant association with chronic pain (AOR: 1.02, 95% CI: 1.00–1.04, p < 0.05). The last variable in the regression model was the adverse life events score, with no statistically significant association evident with chronic pain in either group.

Discussion

This study has shown that reported chronic pain is a significant health problem for older people, with 27% of those with intellectual disability and 24% without intellectual disability reporting chronic pain. These figures for people without intellectual disability are broadly in line with previous national and international data. For example, the 2017 Healthcare Survey conducted by ABS using a sample of the Australian general population reported a prevalence of persistent [chronic] pain of 19–20% for people aged 55–64 and 64–74 years respectively and 24% for the oldest age groups (85 years and older) (ABS, 2017). Similarly, international estimates show pain prevalence to be around 20% in adult population (Breivik et al., 2006; Goldberg & McGee, 2011). While chronic pain is a problem amongst older people in general, the current project reveals that people with intellectual disability comparatively experience chronic pain at an even higher rate than for people without intellectual disability, and this finding underlines the need to proactively support this cohort in managing this vital aspect of healthcare. The following discussion considers key aspects of the findings from this study in relation to pain management.

Health conditions associated with chronic pain

There are several health conditions associated with chronic pain. Data from the 2016 round of the Global Burden of Disease (GBD) study found a considerable increase in chronic pain associated with musculoskeletal disorders, which include a range of conditions, and most notably pain associated with osteoarthritis in older age groups (Hay et al., 2017). Similarly, analysis of the more recent 2019 GBD data indicates that 20% of the chronic pain burden was associated with musculoskeletal conditions (Cieza et al., 2020). Another possible correlate of chronic pain associated with musculoskeletal conditions in elderly people are due to foot problems. These issues can aggravate existing pain, such as that associated with osteoarthritis due to gait disturbance, and can be a risk factors for falls (Menz, 2016). In the current study, chronic pain associated with osteoarthritis, a degenerative musculoskeletal disorder, was 29.2% in people with intellectual disability and 24.1% in the group without intellectual disability. The epidemiology of pain due to musculoskeletal problems and implications for policy and gerontological/geriatric practice is provided in considerable detail by Blythe and Noguchi (2017).

Physical and/or mental functioning and chronic pain

This study revealed a strong statistically significant association between physical functioning [as measured by PCS scores of the SF-12 scale] and chronic pain. Since PCS explores functional limitations in the 4-week period prior to the survey and chronic pain is of a longer duration, some temporal association can be cautiously inferred. However, physical functioning can be limited by both subjective and objective perception of pain, as well as presence of other health conditions that limit mobility. A Dutch study that recruited older participants from general practice clinics found that negative thoughts and perceptions due to functional limitations associated with chronic pain led to a decline in physical functioning. Although the overall change in physical functioning over time was small, there was considerable variation in physical functioning across the respondents (Ilves et al., 2019).

Some studies highlight the bi-directional link between mental/psychological health and chronic pain (Chou, 2007; Roy-Byrne et al., 2008). For older people with intellectual disability in the current study, neither anxiety or depression had any statistically significant association with chronic pain after controlling for other health conditions. However, MCS values remained statistically significant as a predictor of chronic pain. In contrast, for older people in the non-intellectual disability group, all three variables had statistically significant association with chronic pain. Mossey and Gallagher (2004) in their longitudinal study of residents in continuing residential care centres in the US found that 13% of the older population suffered from both depression and chronic pain. Moreover, social isolation and poor quality of life outcomes have also been linked to chronic pain (Domenichiello & Ramsden, 2019). It is also known that medical practitioners and carers for people with intellectual disability may misdiagnose mental health conditions and overlook ongoing pain due to communication limitations (e.g., Wark et al., 2014), which results in either a delayed or complete absence of pain relief and effective pain management. The problems associated with chronic pain and social isolation are also potentially further magnified for individuals with intellectual disability who may display behaviours of concern in response to undiagnosed and untreated chronic pain (Wark et al., 2016). It is considered vital that healthcare practitioners across all disciplines remain cognisant of the higher likelihood of chronic pain in this cohort, and proactively pursue assessment and management of potentially untreated pain.

Limitations

There are a number of key limitations of the present study. The cross-sectional survey design precludes making any inferences about temporal association of predictors of pain beyond correlational observations in both groups. Pain intensity was not assessed for either group as the literature indicates several methodological issues associated with administering valid pain measurement scales for people with mild-to-moderate intellectual disability (Chibnall & Tait, 2001). Since this study was a comparative assessment of correlates of chronic pain across both groups, the issues for intellectual disability group also led to excluding questions on pain intensity for the group without intellectual disability. In absence of clinical records and physical examination, pain severity can be graded into pain-related distress and/or functional impairment. It would be useful in the future to assess both of these aspects, rather than just functional impairment. This would work best with a mixed-methods study design where pain-related distress and functional impairment could be assessed both quantitatively as well as seeking participant’ perceptions through qualitative measures. Further, it was not possible to obtain accurate current assessments of the level of intellectual disability (e.g., mild; moderate; severe/profound) in order to see whether there were any in-group differences. It is recommended that this could be a foci within any follow-up studies specific to people with intellectual disability.

Conclusions

Chronic pain is a multi-dimensional issue. Although it is relatively well researched in the mainstream gerontology literature, there is paucity of literature that actively considers the impact on older adults with intellectual disability, and there are few comparative analyses using similar tools to assess prevalence and correlates of chronic pain across the two groups. As the longevity of this cohort increases, it is important that greater attention at all levels of healthcare is focused on chronic pain management with widespread implementation of adequate and appropriate proactive preventive and treatment strategies.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

The project was funded by a linkage grant from the Australian Research Council [grant number LP130100168] and by contributions from the following industry partners: Ability Options, Endeavour Foundation, The Ascent Group and Uniting. We are deeply indebted to our industry partners, the project Steering Committee and the study participants who generously provided time and support.