Pain catastrophizing in the elderly: An experimental pain study

2.1 Subjects

Sixty-six healthy young and elderly volunteers participated in the study, 34 young: 20–35 (24 ± 3.2 years, 17 F) and 32 elderly: 65–87 (74 ± 6.2 years, 18 F). Sample size was determined using an a priori calculation estimating type 1 error at 5% (alpha = 0.05) and type 2 error at 15% (85% power), with a conservative effect size moderate-to-large (d = 0.5–0.8).

Participants were recruited via posted advertisements and signed informed consent. They were screened with an interview prior to participation to exclude conditions that could affect their ability to follow instructions, answer the questionnaire, and report pain perception. In the interview, the participants were also assessed with the Mini Mental State Examination (MMSE) to ensure that their cognitive capabilities were intact. Individuals that scored >28 on MMSE were included in the study. Additionally, the following exclusion criteria were determined: if the participant reported the presence of severe ongoing pain, diabetes, signs of rheumatic or arthritic disease, especially on the neck and shoulders, and mental conditions such as anxiety and depression. All participants were pain-free, and none had taken any analgesic or sedative for at least 48 h prior to the experiment.

The study protocol was approved by the local ethical committee (N-20130071; N-20210065), and it was in accordance with all relevant national regulations, institutional policies, and followed the Helsinki Declaration and IASP’s guidelines for pain research in humans. The study was part of a larger study on assessing pain in dementia as a part of the EU-COST action TD1005. The subjects in this study were recruited as healthy controls in the primary study.

2.2 Apparatus

2.3 Procedure

The experiment took place in a quiet and climate-controlled room, where the subjects were seated in a comfortable chair. The subjects filled in the PCS and were familiarized with the experimental procedure. Before the test started, subjects were trained until they understood the procedure and could follow the instructions.

Six pressures were applied on the trapezius muscle starting from the lowest intensity on the dominant and non-dominant side of the body in an alternated order: 50 kPa (dominant) – 50 kPa (non-dominant) – 200 kPa (dominant) – 200 kPa (non-dominant) – 400 kPa (dominant) – 400 kPa (non-dominant). Randomization for starting with the dominant/non-dominant side of the body was applied across subjects and individually assessed.

After each stimulation, subjects were instructed to rate the pain intensity and unpleasantness of the pressure stimulation.

2.4 Data analysis

Data were checked for normality using Shapiro–Wilk tests and calculating data frequency in histograms and Q–Q plots. Data were non-normally distributed; consequently, non-parametric tests were used in the following analyses.

Mann–Whitney U-tests were used to compare pain catastrophizing and subscales of rumination, magnification, helplessness, and NRS pain intensity and unpleasantness between young and elderly. Effect size estimations were calculated using the following formula: r = z √ N , with values of 0.1 small, 0.3 medium, and 0.5 large effect size [38].

Spearman’s r correlations were used to evaluate the relationship between NRS pain intensities, unpleasantness responses, and pain catastrophizing measures in the elderly and young groups. The significance of these correlations was assessed using the bias-corrected and accelerated method, which is effective in controlling type 1 errors associated with multiple comparisons. The 95% confidence interval (CI) using bootstrap resampling (1,000 sample, bias-corrected confidence limits) was computed for each correlation. Significant coefficients with an associated CI that did not include zero were considered statistically significant.

Box plots were used for descriptive statistics, where results were presented as the median, range, and interquartile range (IQR).

Data were statistically analyzed using SPSS version 28.0. A significant level of 0.05 was used. One elderly subject was excluded from the analyses since the PCS was incomplete.

3.1 Distribution of pain catastrophizing between young and elderly

The distribution of pain catastrophizing total score and subscale scores resulted in the following for (1) the young group: PCS tot (Med = 10; IQR = 8.8; 95% CI = 8.8–13.7), rumination (Med = 4; IQR = 6; 95% CI = 3.5–5.6), magnification (Med = 2; IQR = 3; 95% CI = 1.2–2.6), helplessness (Med = 4; IQR = 5.3; 95% CI = 3.8–6.0), and (2) the elderly group: PCS tot (Med = 5; IQR = 14; 95% CI = 4.8–11.0), rumination (Med = 2; IQR = 4; 95% CI = 1.5–5.6), magnification (Med = 2; IQR = 2; 95% CI = 0.9–2.0), and helplessness (Med = 2; IQR = 7; 95% CI = 2.1–5.3).

In addition, 19 subjects (24% of the sample) reported PCSs above the 75th percentile, corresponding to PCSs of ≥15 (Med = 20). A larger percentage of these individuals was young, 57.8%, whereas a smaller percentage, 42%, was elderly, with a larger percentage of men, 68.3%, as compared with women, 31.5%. Notably, in the elderly group, men were significantly older than women (EM, Med = 76.5; EW, Med = 71; U = 190; N = 31; P < 0.005).

Mann–Whitney U-tests showed that elderly individuals reported significantly lower pain catastrophizing total scores than young individuals (U = 360.5, N1 = 34, N2 = 31, Z = −2.193, r = −0.27, p = 0.028, two-tailed asymptotic). Additionally, comparisons within the subscale scores showed that elderly individuals had significantly lower ruminations (U = 347.5, N1 = 34, N2 = 31, Z = 2.390, r = −0.29, p = 0.017, two-tailed asymptotic) and helplessness scores (U = 378.0, N1 = 34, N2 = 31, Z = −1.970, r = −0.24, p = 0.049, two-tailed asymptotic) than young individuals. No significant difference was observed in the magnification dimension (U = 469.5, N1 = 34, N2 = 31, Z = −0.784, r = −0.09, p = 0.433) (Figure 1).

Figure 1

Significant lower pain catastrophizing total scores (p = 0.028*), rumination (p = 0.017*), and helplessness (p = 0.049*) scores were observed in elderly individuals when compared with young ones. Outliers are presented as ° in the figure.

3.2 Pressure stimuli 50, 200, and 400 kPa: NRS intensities and unpleasantness

The different pressure stimulations resulted in the following: 400 kPa moderately pain intensity (dominant Med = 6; IQR = 4.13; no-dominant Med = 6; IQR = 7.5) and unpleasant pain (dominant Med = 5; IQR = 4.13; no-dominant Med = 5; IQR = 2.5) perceptions; 200 kPa slightly pain intensity (dominant Med = 4; IQR = 5; no-dominant Med = 5; IQR = 3.7) and unpleasant pain (dominant Med = 3; IQR = 4.5; no-dominant Med = 4; IQR = 4.5) perceptions; and 50 kPa no-pain intensity (dominant Med = 0.5; IQR = 1; no-dominant Med = 0.5; IQR = 1) and unpleasant pain (dominant Med = 0.0; IQR = 0.5; no-dominant Med = 0.0; IQR = 0.5) perceptions. No statistical difference was observed between the dominant and non-dominant sites. Since pressures of 400 kPa could elicit moderate pain stimulations, these were further analyzed.

Mann–Whitney U-tests showed that elderly individuals (Med = 5) reported significantly lower NRS intensity pain scores (U = 366, N1 = 34, N2 = 31, Z = −2.122, r = −0.26, p = 0.034, two-tailed asymptotic), and lower NRS unpleasantness pain scores (Med = 4.5) (U = 334, N1 = 34, N2 = 31, Z = −2.546, r = −0.31, p = 0.011, two-tailed asymptotic), than young individuals (Med = 6.3, NRS intensity; Med = 5.5, NRS unpleasantness), respectively (Figure 2).

Figure 2

Significant lower pain intensity (p = 0.034*) and unpleasantness (p = 0.011*) NRS scores were observed in elderly individuals when compared with young ones.

3.3 Correlations between NRS pain responses and pain catastrophizing levels

Correlations were evaluated between NRS pain intensities and unpleasantness responses and total score of pain catastrophizing, rumination, magnification, and helplessness scores in elderly and young groups. No significant correlations were observed in the young group. In the elderly group, significant, positive correlations were observed between NRS pain unpleasantness scores and total PCSs (r _s = 0.416, p = 0.020, two-tailed), rumination (r _s = 0.420, p = 0.019, two-tailed), and helplessness (r _s = 0.434, p = 0.015, two-tailed). No significant correlations were observed with magnification and between all sub-scales of PCSs and NRS pain intensities.

4.1 Different emotional regulatory strategies in elderly individuals

The study supported the view that older adults are less bothered or emotionally affected by pain and might use pain-related coping strategies associated with psychosocial well-being more effectively. The results were also in line with empirical evidence from large-scale studies showing that elderly, non-demented, individuals are in a better mood than their younger counterparts [10,39] and have fewer negative emotional experiences and greater emotional control [40,41]. Different types of explanations exist, for example, an improved resilience due to more experience with life-related stressors, enhanced use of coping strategies, or simply the acceptance of pain as part of the aging process [3,42,43]. Greater pain acceptance or reluctance to express pain has been often related to stoic attitudes toward pain and have been reported in older adults [44]. However, a recent study that compared stoic attitudes between young and elderly during an acute experimental pain induction have not found any significant difference among the groups [45].

The suggestion that elderly make a more effective use of pain-related coping strategies has been corroborated by studies showing that older people appear to use active [46,47] and adaptive coping strategies regardless of the overall pain intensity [42] and diverse use of emotion regulation strategies [48]. Today, it is acknowledged that pain catastrophizing is a complex phenomenon resulting in several interrelated processes, and it can be accounted as an emotional regulatory process [11].

Studies focusing on emotional regulatory strategies have found that reappraisal strategies (reframing the way one thinks about pain to reduce its emotional impact) partially mediate the relationship between the affective dimensions of pain and catastrophizing. Older women used reappraisal strategies more than younger women, whereas older men reported a lower use of reappraisal [48] and a larger use of suppression strategies (intentionally avoiding thoughts and feelings related to pain) than women. This is consistent with the idea that men tend to avoid emotional reactivity [49]. Additionally, distress, anxiety, and negative lifestyles mediated the relationship between catastrophizing and pain-related outcomes in older adults [20].

The present study also examined the contribution of the individual dimensions of pain catastrophizing between young and elderly. Rumination and helplessness were both reduced in the elderly as compared with the young ones. Previous psychological studies have shown that helplessness leads to a deterioration in motivation, cognition, and affect [50,51], and in elderly, the symptoms of helplessness have been associated with the risk of developing addictive behaviors [52]. Research on patients with rheumatological diseases revealed that helplessness is associated with a number of negative outcomes, such as poorer quality of life, functional disability, more severe symptoms, pain, fatigue, and higher distress [29,53,54]. Additionally, helplessness influenced patient’s attitudes toward illness and interfered negatively with treatment compliance [55,56]. Conversely, a reduced sense of helplessness enabled engagement in meaningful activities for copying with pain [57]. In elderly, the self-perception to control over health-related issues is fundamental for maintaining physical and psychological well-being [58]. Similarly, studies investigating the influence of rumination over health showed that individuals with high levels of rumination, and who continued to think about stressful events, had a slower recovery of heart rate and higher levels of cortisol [59]. Less rumination is observed in older individuals above the age of 63 when compared with other age groups [30], whereas excessive rumination and worrying are observed in young college students, influencing the relationship between pain and anxiety/depressive symptoms [60].

The results support the view that generally, healthy older adults are more resilient than younger ones [32]. Psychological resilience has shown to contrast the negative effects of illness [61], especially in older adults [62,63], and contribute to a successful healthy aging. Conversely, the elevated ruminative scores observed in the young adults point to the growing concern about the steadily increasing poor mental health in young people [64], as well as the early onset of mental disorders such as depression and anxiety [65]. More studies are reporting associations between mental disorders and pain symptoms and higher rates of chronic pain in young adulthood [66,67]. Consequently, the different degree of catastrophizing responses observed in the present sample, between young and elderly, could lead to risk or protective factors depending on the age group considered.

Finally, whether pain catastrophizing changes as function of gender is not completely determined [68]; although evidence points to higher levels of pain catastrophizing in women [13,69,70], a recent meta-analysis showed that men with chronic pain syndrome who had psychosocial symptoms had elevated pain catastrophizing [71]. Although the present study did not address gender difference due to small sample size, it was observed that high levels of pain catastrophizing were scored by elderly men, which were significantly older (>70) than the women (<70). Although limited, the studies that considered oldest-old individuals showed an increasing demand of care and support for pain management in this population [72,73]. Consequently, pain catastrophizing and gender difference in oldest groups are warranted for investigations.

4.2 Differences in age-related changes in pain perception

Experimental studies have shown a greater variability in pain perception among healthy old and young individuals. This variability has been attributed to the type of noxious stimulus used and the modality of the pain induction. For example, decreased sensitivity has been observed when phasic, short duration stimuli were applied on the skin surface, whereas increased sensitivity has been observed when tonic, deep, and diffuse stimuli were used [4,74,75]. However, age-related changes in nociceptive pathways have also been implicated in these differences. Experimental studies have highlighted a reduced pain tolerance and a decrement in the endogenous pain modulatory system in healthy elderly when compared with the young ones [76,77,78].

In the present study, the pain intensity and pain unpleasantness ratings to pain pressure stimuli were significantly reduced among elderly individuals. The results are in line with some of the aforementioned findings, and these decrements could be due to an actual reduction of pain sensitivity due to the noxious experimental modality utilized. Alternatively, it could also be speculated that the decrements in pain ratings could had been the product of supraspinal processing that top-down regulates the descending modulatory pain system. Less catastrophizing was driven by less rumination and less helplessness dimensions which are cognitive processes governed by prefrontal cortical structures [79]. Additionally, elderly individuals exhibit positive associations with catastrophizing and unpleasant ratings, indicating that these two variables were somehow related and influenced by each other, mainly in elderly but not in the young ones. It could be hypothesized that compensatory cognitive and emotional mechanisms take place in healthy elderly individuals when dealing with physiological pain-related changes. Future studies should investigate this hypothesis and underline the interactive mechanisms of resilience observed in elderly individuals.

4.3 Methodological considerations and limitations

The present study has some limitations. First, the sample utilized here was selected as control sample for a primary study on pain in dementia. Consequently, the generalizability of the findings would be difficult, especially for the elderly group who might not be representative of the population since they were selected to be very fit. Young and older individuals with persistent and chronic pain might struggle with low mood, poorer physical function, disability, and social isolation and are at risk of depression. Depressed mood is the strongest predictor of health decline in the elderly [80], and catastrophizing in individuals with mild cognitive impairment could be a limiting factor for proper cognitive management and preventive treatments for dementia for this group [81]. However, studies investigating catastrophizing in individual with cognitive impairment are lacking at the present. In the elderly, dementia is likewise a major hurdle for optimized pain management [82] and for using different pain scales/questionnaires [83,84].

A better stratification of aging groups: youngest-old (ages 65–74), middle-old (ages 75–84), and oldest-old (>85), could have been more informative about the development of pain catastrophizing over the lifespan, especially now that the global aging population tends to be larger and the old classification of above 65 is becoming too reductive.

Second, this study investigated pain catastrophizing with the sole use of the PCS. Pain catastrophizing is a complex phenomenon, and it could be a limitation to measure it with only one scale [85]. Additionally, although age-related difference has been suggested, pain catastrophizing could also depend by personality traits and individual pain experiences gathered during life. Future studies should attempt to replicate the present findings adding a larger range of psychological variables considering the multidimensionality of pain catastrophizing, personality variables, social context of the sample, as well as assessing the pain descending modulatory system.