This study was conducted at the Fragile X Research and Treatment Center of the MIND Institute, University of California, Davis Medical Center, with participants in the Genotype-Phenotype Relationships in Fragile-X families data set. Participants with FXTAS were identified through families of known members affected by fragile X syndrome, referral from physicians, and via self-referral. All participants gave written informed consent from the UCD IRB and were categorized by the stages of FXTAS (32): stage (0) no tremor/ataxia; (1) questionable tremor and/or ataxia; (2) mild tremor and/or balance problems with minimal interference in activities of daily living (ADLs); (3) moderate tremor and/or ataxia with significant interference in ADLs; (4) severe tremor and/or ataxia requiring a cane or walker; (5) requiring daily use of wheelchair; (6) bedridden. We have eight participants who do not meet the diagnostic criteria for FXTAS (2), although all have premutation and subtle neurological symptoms apart from one patient who has the MCP sign without tremor or ataxia, a phenomenon previously reported in five males (33). These patients likely represent the most subtle end of the spectrum involving neurological deficits in the premutation and are included for comparison purposes. Participants who reported medical conditions such as a history of stroke, head injury, or primary language other than English were excluded from the study.

Participants underwent medical and neurological evaluation to assess FXTAS diagnosis, stage, and FMR1 molecular studies for CGG repeat size and mRNA level. A medical history was taken at the time of visit to document FXTAS phenotype and symptoms including tremor and ataxia in addition to age of onset and severity, current and past medication use, self-reported anxiety and depression, musculoskeletal and autoimmune problems such as osteopenia, osteoporosis, osteoarthritis, rheumatoid arthritis, fibromyalgia, lupus, ANA positivity, and other immunological diseases. We collected data on restless legs syndrome, severe cramps, disc/spine problems, and muscle pain. Migraine with and without aura, age of onset, and frequency was collected along with symptoms of neuropathy, allodynia, back pain, and chronic pain in general. Activation ratio was measured using ratios of signal intensity with Southern blot as previously described (34). FMR1 CGG repeat allele length was quantified using a combination of both PCR and Southern blot analysis as previously described (35).

Pain medication categories were separated into non-overlapping categories of opiate analgesics, non-opiate analgesics, cannabinoids, anesthetics, antimigraine, nerve pain, and muscle pain/relaxation. Non-opiate analgesics included non-steroidal anti-inflammatory drugs (NSAIDs) and acetaminophen. Anesthetics included lidocaine. The category of antimigraine medications did not overlap with other medications categories and contained triptans, topiramate, and botulinum injections. Nerve pain medications were based on first line treatments and include tricyclics, serotonin and norepinephrine reuptake inhibitors (SNRIs), and anti-epileptic drugs (AEDs) including gabapentin, pregabalin, and carbamazepine. Muscle pain/relaxation medications included antispasticity and antispasmodic agents such as baclofen, tizanidine, and cyclobenzaprine.

Statistical analyses of data were performed with an open-source R software. Results were expressed as mean ± standard deviation (or error) of mean for continuous variables and proportion (%) for categorical variables. For quantitative variables, group differences in means or medians were determined by t-test, analysis of variance (ANOVA), or Kruskal-Wallis test as appropriate. For categorical variables, proportions were compared between groups using t-test for proportions, chi-squared test, or Fisher's exact test as appropriate. Linear regression analysis was performed to assess associations between two variables with the main effects of the predictor and gender and their interaction term included in the linear models. Missing values were excluded from the statistical inference tests. Two-tailed p < 0.05 were considered statistically significant as appropriate.

The participants are 104 individuals (41 female and 63 male) with FXTAS and the premutation confirmed via molecular studies. No significant gender difference was found for race and ethnicity (p = 0.513), education level (p = 0.199), age (p = 0.743), CGG repeat size (p = 0.266), or FXTAS stage (p = 0.404). The mean (SD) age of females was 68.1 (8.8) years and age of males was 67.5 (8.0) years. Race and ethnicity are primarily white (100% female vs. 95.7% male). Of the 41 females in the study, 34 (85%) are FXTAS stage >2. Of the 63 males, 57 (90.5%) are FXTAS stage >2. The mean CGG repeat size for females is 87 (SD = 18), and the mean for males is 90 (SD = 16). Refer to Table 1. There is a significant difference in FXTAS diagnostic categories (p = 0.001) with more males diagnosed with “Definite FXTAS” than females (48.3% male vs. 9.8% female). Females are more represented in the Probable category (30.0% male vs. 58.5% female).

Mean age of onset of migraine is 28.7 years (SD = 17.6), chronic pain is 51.4 years (SD = 17.3), peripheral neuropathy is 57.4 years (SD = 11.4), tremor is 58.6 years (SD = 9.2), and ataxia is 60.7 (9.4), and no gender differences were found (see Table 1).

The age of onset of peripheral neuropathy symptoms was positively associated with the age of onset of ataxia (β = 0.56 ± 0.17, p = 0.004) and tremor (β = 0.56 ± 0.17, p = 0.004). There were no significant associations between the onset of chronic pain and migraine with the onset of ataxia and tremor (see Table 2 and Figure 1). In addition, there were no gender differences for any of those associations.

There were significant gender differences in the prevalence of several symptoms. Back pain was significantly higher in females than males (48.5 vs. 23.4%, p = 0.008), as was migraine (43.9 vs. 14.5%, p = 0.0008), fibromyalgia (26.8 vs. 0%, p = 0.0071), thyroid problems (34.2 vs. 14.5%, p = 0.0182), osteoarthritis (65.9 vs. 43.5%, p = 0.027), allodynia (20 vs. 2.0%, p = 0.008), peripheral neuropathy pain (43.9 vs. 25.4%, p = 0.0488), and anxiety (82.9 vs. 39.7%, p < 0.001). No significant gender differences were found in autoimmune problems, musculoskeletal pain, or depression (see Table 3).

Among participants who experienced pain, there were no gender differences in CGG repeat length. Among participants who had experienced autoimmune problems, there was a significant gender difference in CGG repeat length. Females with autoimmune had significantly lower CGG repeats than that of males (76.5 ± 12.9 female vs. 105.5 ± 26.5 male, p = 0.032). Refer to Supplementary Table 4.

Overall, the prevalence of patients with musculoskeletal pain (severe cramps + muscle pain + arthritis + back pain + disc or spine problems) was significantly higher than the prevalence of patients with peripheral neuropathic pain (PNP) (88.4 vs. 33.7%, p < 0.0001). The difference between the proportion with neuropathic pain and the proportion with musculoskeletal pain was significant in both males and females, p = <0.0001 (85.5 vs. 25.5%) and p = <0.0001 (92.5 vs. 45%), respectively (see Supplementary Table 5).

The data does not provide evidence that activation ratio is significantly associated with either neuropathic pain or musculoskeletal pain, p = 0.241 and p = 0.770, respectively. The mean activation ratio for those with neuropathic pain is 0.52, while the mean activation ratio is 0.60 for those without neuropathic pain. The mean activation ratio for those with musculoskeletal pain is 0.57 while the mean activation ratio is 0.61 for those without musculoskeletal pain (see Supplementary Table 6).

Females with FXTAS are significantly more likely to be taking any pain medication (58.5% females vs. 36.5% males, p = 0.0271) as well as nerve pain medication (31.7% females vs. 14.3% males, p = 0.331). No significant gender differences were found for other medication types. Refer to Supplementary Table 7.

Both men and women with FXTAS experience chronic neuropathic and musculoskeletal pain. However, women with FXTAS experience significantly more chronic pain than men and take more pain medications. It's currently unclear why women with FXTAS experience more chronic pain than men, particularly central sensitivity syndromes, neuropathic pain, and back pain. Possible mechanisms include differences in endocrinology, autoimmune conditions, connective tissue, inflammation, and neuropsychiatric differences.

Females with FXTAS were more likely to have conditions that fit under the umbrella of central sensitivity syndromes, characterized by altered pain regulation such as fibromyalgia, allodynia, chronic pain, and chronic headaches (6). We also found that the defining symptom of central sensitization, allodynia, is significantly higher in women with FXTAS (p = 0.008). Sensitization occurs following intense or chronic peripheral noxious stimuli, tissue injury, or nerve damage that keeps the peripheral nervous system in a constant state of activity. Chronic pain conditions commonly known to lead to hypersensitivity such as back pain, arthritis, and fibromyalgia are significantly prevalent among women with FXTAS. Central sensitivity symptoms are heavily influenced by psychological stressors (21), which are significantly worse for females with FXTAS in our cohort. Despite the known role of FMRP in altering sensitization of pain and neuropathy in animal models, we did not find that these pain phenotypes correlated with CGG repeat length.

We found a quarter of participants (26.2%) reported experiencing migraines, with rates significantly higher among females than males (43.9% females vs. 14.5% males, p = 0.0008). Age of migraine onset had no bearing over age of ataxia and tremor onset. This confirms a previous study on FMR1 premutation carriers and the prevalence of migraines (13). Author Au hypothesized that migraines in FXTAS are related to mitochondrial dysfunction leading to increased oxidative stress, which worsens with age. Several chronic pain syndromes prevalent in female FMR1 premutation carriers are comorbid with migraine including fibromyalgia (36), allodynia (37), and chronic fatigue syndrome (36), suggesting diffuse alterations to the nociceptive nervous system. Individuals with migraines are also at an increased risk for anxiety and depression (38).

We found that women with FXTAS self-reported significantly more anxiety symptoms (82.9% females vs. 39.7% males, p = 0.001); depressive symptoms were present in a quarter of participants (24.4% female vs. 25.8% male). This finding is consistent with previous studies that found women with FXTAS have more depression and emotional symptoms (4, 7). The relationship between pain and psychopathology is multi-directional; chronic pain is a strong predictor of both onset and persistence of comorbid psychiatric disorders in general, and vice versa, psychiatric disorders are also a powerful predictor of chronic pain persistence (39–41). Emotional states can also amplify the pain felt from central sensitivity syndromes (21). Women with FXTAS may report more chronic pain partly because that pain is viewed through a lens intensified by symptoms of anxiety, which play a role in pain perception (42). How significant a role mental health plays in FXTAS across gender may also be influenced by reporting bias, as men are less likely to disclose mental health problems than women in general (43).

Consistent with previous research on bone density in premutation carriers (44, 45), females were found to have more skeletal related problems such as osteopenia, osteoporosis, and osteoarthritis. With respect to pain, women with FXTAS reported significantly more back pain than men (48.5% of women vs. 23.4% of men, p = 0.008). Refer to Table 3. Post-menopausal women in the general population do tend to have higher incidence of back pain than men (46, 47), but percentages are similar between men and women in the 65–74 age group within the US (48). The cause of these gender differences is likely secondary to hormonal differences in post-menopausal women. The role estrogen plays in bone density is additionally important for women with the premutation because the mean age of natural menopause is reduced by about 5 years, from the typical age of about 51 to 46. Separate from this overall earlier menopause, 20% of carriers (49) can develop fragile X-associated primary ovarian insufficiency (FXPOI) with typically onset at 33 years (50). This directly decreases the lifetime exposure of estrogen and decreases bone mineral density, as early menopause is associated with osteoporosis (51). The significantly increased back pain among women is likely secondary to decreased bone density-related sequelae, which can have a devastating impact on women with FXTAS.

More women report feeling peripheral neuropathic pain than men (43.9 vs. 25.4%, p = 0.0488). Women were also more likely to be taking medications that treat nerve pain (31.7 vs. 14.3%, p = 0.0331), however some of these medications overlap with treatment for anxiety, which was also significant among women with FXTAS. We did not find CGG repeat number correlated with nerve pain in men or women. Nerve conduction studies have shown men and women have similar sensory nerve fiber abnormalities (52), but men have additional motor nerve abnormalities in conduction velocity and latency (53). Pathogenesis of neuropathy is related to RNA toxicity causing the creation of intracellular inclusion bodies in neurons and glial cells (54) and peripheral tissues (55, 56) that causes axonal dysfunction.

Consistent with previous research that neuropathy is a presenting feature of FXTAS (8), we found the age onset of peripheral neuropathy was positively associated with the age onset of ataxia (β = 0.63 ± 0.25, p = 0.019) and tremor (β = 0.56 ± 0.17, p = 0.004), with no gender difference. This further supports the recommendation to consider a FXTAS diagnosis in a patient presenting with neuropathy and a family history of intellectual disability, premature ovarian failure, autism, or movement disorder (8).

We found that musculoskeletal pain is significantly more prevalent than peripheral neuropathy pain (88.4 vs. 33.7%, p < 0.0001) in FXTAS. There was no significant gender difference in musculoskeletal pain (92.5 vs. 85.5%, p = 0.3077). Of interest, every patient that reported neuropathic pain also reported musculoskeletal pain, meaning all who experienced neuropathic pain also reported arthritis, back pain, severe cramping, or muscle pains. However, 54.7% of those with musculoskeletal pain did not have neuropathic pain (see Supplementary Table 6). Clinically, premutation carriers can manifest connective tissue-related features such as hyperextensible finger joints, large ears, and connective tissue dysplasia that are more subtle than those seen in FXS (57). Case report evidence including five females with premutation and Ehlers Danlos syndrome phenotype suggests there may be related commonalities in pathogenesis (58). Possible pathophysiologic mechanisms hypothesized include FMRP deficiency, mRNA toxicity, and sex effects (58). FMRP is known to regulate multiple connective tissue pathways including elastin, actin, and matrix metalloproteinase (MMP9), a class of enzymes involved in bone development, wound healing, and pathology such as arthritis and intracerebral hemorrhage (59, 60). Levels of connective tissue involvement correlate with level of FMRP depletion in FXS and premutation CGG repeats > 120 are more likely to be associated with connective tissue problems (61). Because the FXTAS population has minimal alterations to FMRP which result in altered regulation of these pathways, this mechanism could contribute to generating chronic musculoskeletal pain in FXTAS.

We did not find a significant gender difference (p = 0.2301) in rheumatoid arthritis, systemic lupus erythematosus, or antinuclear antibodies positivity, but the prevalence for these conditions was quite low. Confirming previous research, we found FXTAS women are more likely to have fibromyalgia (26.8 vs. 0%, p = 0.0071) and thyroid problems (34.2 vs. 14.5%, p = 0.0182). Fibromyalgia is generally thought to be immune mediated (62), but its etiology remains unclear with some evidence linking it to small nerve fiber neuropathy (63) and central sensitization (64). Thyroid problems consisted almost entirely of autoimmune causes, including hypo and hyperthyroid states. Females with autoimmune conditions had significantly lower CGG repeats than that of males, likely because women are more likely to have autoimmune conditions at baseline. Autoimmune conditions and inflammation have strong links to pain and the primary pathophysiology of FXTAS involves RNA toxicity and inflammation. Inflammatory mediators from damaged cells induce pain and sensitization. Local tissue inflammation can result in hypersensitivity to pain via secondary hyperalgesia, where inflammatory mediators diffuse to uninjured nearby tissues. Females produce a larger proinflammatory immune response to tissue damage than males (65). Females with FXTAS may have increased chronic pain because they develop more inflammation contributing to sensitization and maintenance of pain.

Non-opiate pharmacologic therapies can be selected based on type of pain (nociceptive, neuropathic, central sensitization, and combination) or by targeting comorbid conditions.

Considering both psychiatric and pain symptoms of a patient is crucial as psychiatric disorders can exacerbate pain conditions and impede treatment adherence (66). Treatment of depression and anxiety in fragile X-associated neuropsychiatric disorders (FXAND) typically involves SSRIs or SNRIs (15, 67). In addition to targeting symptoms of depression and anxiety, antidepressants such as tricyclics (TCAs), SSRIs, SNRIs, and norepinephrine dopamine reuptake inhibitors (NDRIs) have evidence for use in chronic pain and are commonly used in the treatment of patients with FXTAS. Of the antidepressants, evidence suggests SNRIs and TCAs are the most effective for treating chronic neuropathic pain (68) as well as centralized pain (21). There is also strong evidence for the use of anticonvulsants pregabalin and gabapentin, and with moderate evidence for use of SSRIs (21).

We recommend avoidance of opiates to treat chronic pain associated with FXTAS. Opiates are commonly used to treat pain in the general population and FXTAS population, but anecdotal evidence suggests those on opioids can have a faster progression of their FXTAS symptoms (69). Opiates are reported to trigger white matter changes in chronic users (70, 71). In vitro evidence suggests premutation neurons are more vulnerable to environmental toxins than normal neurons (67). Drug and alcohol use has been reported to be increased in premutation carriers compared to controls (72, 73), which has resulted in opiate overdose (74). Long-term opiate use leads to tolerance, dependence, withdrawal symptoms, and can worsen chronic pain through development of opiate-induced hyperalgesia (75).

Topical treatments can be recommended for chronic pain such as diclofenac and ketoprofen for musculoskeletal issues (76). For osteoarthritis, topical NSAIDs, low-concentration capsaicin, or topical rubefacients. For neuropathic pain, topical lidocaine or high-concentration capsaicin can be used.

CBD is a compound typically derived from hemp that targets CB1, CB2 receptors and is an allosteric modulator of μ- and δ-opioid receptors. Evidence in FMR1 knockout mice suggests CB2 receptor is necessary for protection against neuropathic pain, raising interest in targeting this receptor for treatment (77). A controlled trial examined pain in patients with multiple sclerosis, spinal cord, and other neurological conditions found pain control improved with CBD (78) in addition to two other studies that found pain relief (79, 80). CBD is also being explored as an intervention in FXS (81) in an open-label trial (82) because of its benefits on sleep quality, anxiety (83), and cognitive impairment (84).

Curcumin is a polyphenol with anti-inflammatory properties extracted from turmeric that has shown promising preclinical results in treating FXTAS (85) but has not been studied clinically. Curcumin has a notable anti-inflammatory effect and pain relief for the treatment of general pain syndrome and osteoarthritis (86). However, more comprehensive long-term studies are needed, as many of the controlled trials are of low quality with industry funding. Sulforaphane is another dietary supplement with antioxidant/anti-inflammatory properties that was found to improve markers of oxidative stress and mitochondrial function in fibroblasts from FXTAS patients (87), but has not been studied for pain symptoms in FXTAS.

Allopregnanolone, a natural neurosteroid and positive allosteric modulator at GABA-A, has been studied as a safe therapy for neuropathic pain (88, 89), treatment for post-partum depression (90), and has been studied in an open-label trial for treatment of FXTAS in six patients with limited benefits on executive function and neuropathy symptoms (91). Allopregnanolone is a promising candidate for FXTAS because it improves mitochondrial dysfunction and acts to prevent reactive oxidative species (ROS) overproduction, a primary pathological mechanism driving FXTAS phenotype. ROS produced in mitochondria have been shown to contribute to central sensitization associated with pain (92). Allopregnanolone was able to protect neuronal cells against oxidative stress through improved mitochondrial antioxidant activity (93).

For a non-pharmacologic approach, counseling and therapy for chronic pain and psychiatric conditions has been helpful. The cognitive behavioral therapy approach to pain teaches patients skills to anticipate pain and divert attention to other thoughts, helping the patient better cope with pain (94). We also recommend daily exercise to stimulate neurogenesis, improve mitochondrial function, and decrease chronic pain (67). Focusing on maintaining sleep quality can have an impact on pain (95). We recommend other techniques for reducing overall stress such as biofeedback for relaxation, which can lead to improvements in various pain-related outcomes (96). Mindfulness stress reduction has evidence for reducing chronic pain and improving quality of life (97). Acupuncture has been studied in hundreds of controlled trials and appears to be effective for chronic pain, particularly back pain (98) and headaches (99).

The limitations of this study include a sample bias, as the cohort consisted of participants with higher socioeconomic status and majority with white ethnicity. Data on anxiety and depression are self-reported and the symptomatology of pain is recorded through medical interviews. Future studies characterizing pain would benefit from measures of intensity. Furthermore, participants were enrolled through referral from physicians or family members and thus this cohort might not be a true representative of individuals with FXTAS, but the more affected ones. In addition, men are less likely to disclose health issues, and at more advanced FXTAS stages may not be able to assess themselves accurately.