Characterizing Chronic Pain Phenotypes in Multiple Sclerosis

A Nationwide Survey Study

Anna L. Kratz; Daniel Whibley; Kevin N. Alschuler; Dawn M. Ehde; David A. Williams; Daniel J. Clauw; Tiffany J. Braley


Pain. 2021;162(5):1426-1433. 

In This Article


Preliminary Results

A total of 1220 individuals representing 49 US states (except Wyoming) and the District of Columbia accessed the survey, indicated an MS diagnosis, and were invited to continue the survey. Analyses were completed on the 842 (69%) respondents who endorsed chronic pain (lasting at least 3 months) and who had scores (nonmissing data) on the painDETECT and FM Survey Criteria. Those whose data were included in the analyses were statistically significantly older (Mincluded = 51.83 ± 11.98 [SD follows all ± symbols hereafter], Mexcluded = 49.77 ± 12.84; F(1,1216) = 7.3, P = 0.007) but were not significantly different in terms of sex distribution (P = 0.67). Descriptive statistics for the study sample are in Table 1. Distribution characteristics for all variables subjected to ANOVA tests (eg, PROMIS pain intensity and pain relief scores) met normality criteria for conducting parametric statistical tests (all skew values <|0.76|, all kurtosis values <|1.2|).

Distribution of Pain Subtypes

On the ACR FM Survey Criteria (measure of centralized pain), the sample mean was 12.19 ± 5.65 and the median was 12 (IQR = 8,16; Figure 1.). Using the ACR FM Survey Criteria cut-point of ≥13, 346 (41.1%) of the sample scored positive for FM.

Figure 1.

Distribution of FM Survey Criteria scores (N = 842).

For the measure of neuropathic pain, the painDETECT, the sample mean was 15.73 ± 8.18 and the median = 16 (IQR = 10, 21; Figure 2.). Most of the sample did not show strong evidence of neuropathic pain, with 303 (36.0%) scoring in the range indicating an unlikely neuropathic component, 234 (27.8%) scoring in the unclear/ambiguous range, and 305 (36.2%) scoring in the likely neuropathic range on the scale.

Figure 2.

Distribution of painDETECT scores (N = 842).

Using the median score on the ACR FM Survey Criteria[12] and the positive cut-point on the painDETECT (≥19) to identify probable pain phenotypes in this sample (Figure 3), the largest subgroup showed low scores on measures of both neuropathic and nociplastic pain (n = 341, 40.5%) and was labeled "nociceptive type." The next largest subgroup scored high on measures of both neuropathic and nociplastic pain (n = 226, 26.8%) and was labeled "mixed type." The group that reported pain that did not show neuropathic characteristics but scored high on the FM Survey (n = 196, 23.3%) was labeled "nociplastic type." The smallest subgroup, which consisted of people with MS who showed evidence on the painDETECT of probable neuropathic pain but no evidence of nociplastic pain (n = 79, 9.4%), was labeled "neuropathic type."

Figure 3.

Distribution of pain types based on surveys of neuropathic and nociplastic pain (N = 842).

Pain Intensity and Chronic Overlapping Pain Conditions

The pain subtypes differed significantly in terms of PROMIS pain intensity T-scores (F(3, 841) = 123.11, P <0.001). The nociceptive type had the lowest (mean = 44.98 ± 6.89) average pain intensity, and the mixed neuropathic/nociplastic type had the highest (mean = 55.06 ± 5.77) average pain intensity. The neuropathic (mean = 50.82 ± 6.02) and nociplastic types (mean = 50.71 ± 5.59) had nearly identical average pain scores. Post hoc multiple comparison tests revealed that all pain subtypes were significantly different from each other in terms of pain intensity scores (all P < 0.001), with the exception of no significant difference between neuropathic and nociplastic subtypes (P = 0.99; Cohen's d = 0.02). Large differences in mean pain scores were observed between the nociceptive and neuropathic (Cohen's d = 0.90), nociplastic (Cohen's d = 0.91), and mixed pain subtypes (Cohen's d = 1.58). Medium effects for mean pain differences were observed between mixed pain and both nociplastic (Cohen's d = 0.77) and neuropathic pain subtypes (Cohen's d = 0.72).

Migraine was the most common chronic overlapping pain condition, followed by chronic pelvic pain; TMD was relatively rare. Prevalence of chronic pain conditions significantly differed by pain subtype (Table 2); in all cases, the mixed neuropathic/nociplastic pain type reported the highest frequency of chronic pain syndromes. Nociceptive type showed the lowest frequency of chronic overlapping pain conditions, with the exception of chronic pelvic pain, for which neuropathic pain type had the lowest frequency.

Pain Treatments

Nonsteroidal anti-inflammatory drugs were the most commonly used (66.5%) and steroids the least commonly used (6.5%) medications for analgesia across all pain types (Table 3). For all pain medications, frequency of use significantly differed across the pain subtypes. Participants with high centralized forms of pain—nociplastic or mixed nociplastic/neuropathic pain—most frequently reported use of cannabinoids, opioids, serotonin-norepinephrine reuptake inhibitors, SSRIs, antispasmodics, and benzodiazepines. By contrast, participants categorized with the nociplastic pain type used NSAIDs 10% more frequently than the other groups. For anticonvulsants, those with the nociceptive pain type reported use frequencies~15% lower than the other 3 groups. Steroid use, although uncommon in general, was highest for those with any type of neuropathic pain—either alone or with nociplastic pain (mixed type).

Narcotic pain medications received the highest average pain relief ratings (mean = 6.99 ± 1.78, median = 7.00, and IQR = 6,8), followed closely by cannabinoids (mean = 6.29 ± 2.17, median = 7.00, and IQR = 5,8). Selective serotonin reuptake inhibitors were associated with least pain relief (mean = 2.72 ± 3.12; median = 1, IQR = 0,5) across all pain types. Analgesic ratings significantly differed by pain subtype for only 2 classes of medications: Relief ratings for cannabinoids were significantly higher for those with mixed pain compared with neuropathic pain (Tukey HSD P = 0.02; Cohen's d = 0.65), and relief ratings for NSAIDs were significantly higher for those with nociceptive pain compared with nociplastic (P = 0.001; Cohen's d = 0.41) and mixed pain subtypes (P < 0.001; Cohen's d = 0.70) and for those with nociplastic pain compared with mixed pain (P = 0.04; Cohen's d = 0.30). Comparisons of pain relief ratings for benzodiazepines were underpowered because of low frequency of use and did not reveal statistically significant differences, despite substantial mean differences, indicating that the neuropathic type rated this class of drugs 2.32 points higher in terms of pain relief relative to nociceptive type (Cohen's d = 0.82), 2.84 points higher relative to nociplastic type (Cohen's d = 1.02), and 1.96 points higher relative to mixed pain type (Cohen's d= 0.78) on the pain relief scale.