Complex regional pain syndrome and use of psychotropic drugs as a proxy for psychological health
The consumption of psychotropic drugs in the general Swedish population without a diagnosis of CRPS or a nerve injury/disorder was around 15%, which corresponds well with the prevalence of use of those psychotropic drugs in Sweden according to the National Board of Health and Welfare (/ These figures mirror the prevalence of impaired psychological health according to our definition24. However, psychotropic drug-use was higher. i.e., 26%. in those with a nerve injury in the upper or lower limb, and even higher, i.e., 36%, in individuals with a diagnosis of CRPS in the upper or lower limb, particularly in individuals with CRPS type 2, i.e., those with an associated nerve injury, suggesting that psychological health is linked, without causation22 to CRPS1,14,15,16,17,18,19,20,20.
Furthermore, compared to the general population, the unadjusted relative risk for consumption was about two times higher in patients with nerve injuries and CRPS type 1 and almost three times higher in patients with CRPS type 2. As expected, those relative risks became considerably reduced (PRs around 1.10 and 1.4 for nerve injuries/CRPS type 1 and CRPS type 2, respectively; thus, same risk relation), when adjusting for the demographic and socioeconomic factors age, sex, occupational qualification level, and country of birth. However, those factors are not clearly confounders as CRPS diagnoses may be mediators in the causal pathway leading to impaired psychological health or distress in individuals with low socioeconomic positions.
CRPS can affect the upper or lower limb, and the ICD codes are the same regardless of the limb affected. Our study is innovative as it not only investigates the risk of the use of psychotropic drugs related to CRPS diagnoses in the whole population but also how this risk is modified in different demographic and socioeconomic strata. Such information is often hidden behind average population risks. For instance, women have a higher risk of using psychotropic drugs than men, but the influence of CRPS or nerve injuries on the use of psychotropic drugs was similar within both strata. CRPS is more frequently observed in the upper limb and with an overrepresentation of women4,32,33 which is in accordance with our earlier study consisting of around 70% of women with CRPS type19,20 as well as that distal radius fractures are common among women1. This finding may be related to the types of included nerve injuries and nerve entrapment disorders, where women dominate. The relation between CRPS type 1 and CRPS type 2 was 3.5:1, which is in accordance with other studies4,32; i.e., the present database is valid.
It is known that immigrants have a lower risk of using psychotropic drugs than natives, which may be mainly attributed to insufficient access to health care34,35,36,37. However, the association between nerve injuries, CRPS type 1 and CRPS type 2, and use of psychotropics was certainly higher in the immigrant stratum than in the native stratum (higher ARDs for CRPS type 2 for both natives, 8%-units, and immigrants, 14%-units). That is, suffering from such diagnoses might affect more immigrants than natives, which should be further investigated. Previous psychotropic drug consumption was highly associated [16.09 (16.00-16.18)] with continuous consumption of the drugs, which is in line with opioid consumption in surgically treated neuroma patients, although it is reported to be reduced in such surgically treated patients38.
Individuals with CRPS generally have a poor quality of life, which differs from individuals with other chronic pain conditions. Based on in-depth interviews, individuals with CRPS have described it as a war-like experience (i.e., “dealing with the unknown enemy”, “building an armory against a moving target”, “battles within the war”, “developing battle plans with allies” and “warrior or prisoner of war”39). Factors, such as psychological behaviour, depression, preoperative psychological distress, or an increased pain level, are not predictive for the development of CRPS2,10,21 but anxiety and depression, on the other hand, are linked to CRPS13,14. In addition, individuals with CRPS type 1 are not psychologically different from the general population40,41. One may argue that psychological factors and psychiatric conditions should not be interpreted as causing the condition but rather are a result of the condition leading to psychological stress1,16,17,18,18. Our recent data suggest that both aspects may be relevant19,20. In this context, one may stress that individuals with CRPS may suffer from pain, sleeping problems, limitations in daily activities, anxiety, and depression, which may have an impact on their life20. Such patients may also have a low sense of coherence and high pain catastrophizing scores associated with a worse outcome, impacting the development of CRPS20,42,43. One previous large study showed an association between depression and CRPS, but the cross-sectional design makes it impossible to draw any conclusions on causality44.
Women were at greater risk of consuming psychotropic drugs than men, indicating that women with CRPS may suffer more psychological distress4 than men with CRPS. However, the ARDs were similar between men and women in the three groups of individuals with the diagnoses although being lower for nerve injuries and CRPS type 1 (ARDs < 2.41%-units) and higher for CRPS type 2 (ARDs 9.2–9.5%-units). One other study found contradictory results, with men reporting more depression and more kinesiophobia than women with CRPS45. This might be explained by that men reported worse coping strategies and were less likely to seek support. Hence, women may be more prone to seek help for their depressive symptoms and, therefore, more likely to be prescribed medication or more likely to receive such help when seeking care, which may mirror the findings concerning occupational and income levels (see below). Central sensitisation, i.e., amplification of neural signaling in nociceptive spinal and trigeminal neurons, can induce facilitation of pain hypersensitivity; a feature of CRPS together with a dysregulated sensory processing4. Data from functional MRI studies also suggest several different alterations in the brain related to such findings4.
The present data showed a higher consumption of psychotropic drugs among individuals with CRPS type 1 compared to the general population, even when adjusted for the demographical and socioeconomic factors. Additionally, there was a higher risk for high consumption of psychotropic drugs in CRPS type 2, i.e., in those with an associated nerve injury. The risk of psychotropic consumption among the individuals with injury(ies) to peripheral nerve(s) was similar to CRPS type 1 but lower than observed for the individuals with CRPS type 2. This suggests that the nerve injury per se is relevant for psychological health or psychological distress but with an additive effect of the CRPS diagnosis. Thus, CRPS with an associated nerve injury presented the highest risk for psychotropic drug use, exceeding that of individuals with treated nerve entrapment disorders24indicating that the type of nerve injury may also be relevant in this context. Both the peripheral and the central sensitization of the nervous system in CRPS might also affect the risk of psychotropic drug consumption46.
The occupational qualification levels were also associated with a risk for the use of psychotropic drugs and nerve injury/CRPS with an especially higher risk among those with high and middle high levels (ARDs 25–35%-units) for CRPS type 2. The ARDs were much lower for the other occupational levels and nerve injuries or CRPS type 1 (ARDs < 2.31%-units). In accordance, regarding income levels, the highest association was observed in the high-income stratum (ARD = 17%-units) between having CRPS type 2 and the use of psychotropics, while ARDs were much lower in the low- and middle-low-income strata (7 and 8%-units). This observation that CRPS type 2 affects high-income individuals more (AR increase 13–30%) warrants further investigation because the ARDs were much lower for nerve injuries and CRPS type 1 (< 3.22%-units). Higher ARDs (34 and 32%-units) were particularly observed in individuals with CRPS type 2 at age 25–34 and 55–64 years. In contrast to the findings of CRPS type 2, low occupational or income levels have an impact on use of psychotropic and psychoactive drugs in other nerve entrapment disorders in the upper limb24,25. However, still, our data propose that socioeconomic factors, as income level, occupational status, country of birth, as well as sex and age of the individuals influence the consumption in accordance with observations of high risk of consumption of opioids and gabapentoid drugs as well as psychotropic drugs in patients with nerve entrapment disorders24,25. Disability and pain severity are more intensely associated with psychological factors in CRPS than in other diagnoses with pain, i.e., low back pain47,48. A history of previous CRPS was also associated with a higher risk, which is plausible since there, for different reasons, including genetic heredity, is a propensity in some individuals to develop CRPS16,49,50,51. The reason why a previously surgically treated nerve injury in individuals (see Supplemental Tables 1 and 2) is not associated with the use of psychotropic drugs is more obscure but may hypothetically be related to less risk of having a permanent nerve injury or prevention of formation of a neuroma38.
This study has several limitations that should be considered when interpreting the findings.
Classification of CRPS: The primary limitation is the inability to distinguish between CRPS affecting the upper versus lower limbs. This limitation arises from the structure of ICD-10 coding, which does not provide site-specific classification for CRPS. The updated ICD-11 system allows for such differentiation and may enable more detailed analyses in future studies.
Incomplete Exclusion of Rare Causes: Although we excluded CRPS cases potentially caused by stroke, it was not possible to comprehensively rule out all less frequent etiologies. This may introduce heterogeneity into the study population.
Proxy Measures of Psychological Health: Psychological health was approximated using registry-based data on dispensed psychotropic medications. This approach has inherent limitations as detailed clinical data on CRPS severity, concomitant injuries (including nerve damage), or treatment outcomes were not available.
Also, misclassification may have occurred, as dispensation records do not confirm medication adherence or actual use.
Residual Confounding and Confounding by Indication: Despite adjustments, residual confounding may persist due to unmeasured factors, such as comorbidities, unrecorded psychological conditions, and variability in healthcare access. Confounding by indication is particularly relevant in this context: psychotropic drugs may be prescribed not only for psychological conditions (e.g., depression, anxiety) but also for chronic pain, insomnia, or other off-label indications. This overlap may inflate the observed association between CRPS and psychotropic drug use. To mitigate this, we excluded medications primarily indicated for neuropathic pain and adjusted for prior psychotropic drug use. However, complete disentanglement of psychological distress from pharmacological pain management is inherently challenging in register-based studies.
Sensitivity Analyses: A sensitivity analysis that included previously excluded medications used for neuropathic pain (ATC codes N06AA09, N06AX21, and N06AX16, as per national guidelines) yielded consistent results: PR for nerve injuries: 1.09 (95% CI, 1.07–1.12), PR for CRPS type 1: 1.27 (95% CI, 1.15–1.40), PR for CRPS type 2: 1.41 (95% CI, 1.18–1.69). These findings support the robustness of the primary results.
Definition of Exposure: Psychotropic drug exposure was defined as at least one dispensation during the one-year follow-up period. We did not conduct sensitivity analyses to distinguish acute from chronic use, nor did we apply a washout period as used in some pharmacoepidemiologic research. These decisions reflect our aim to capture broad, population-level patterns of psychotropic drug use following CRPS diagnosis, the limited clinical detail in the registry data regarding treatment duration and indication, and consistency with previous studies24,25,25. Nonetheless, potential misclassification of exposure timing remains a consideration and should be addressed in future research with more detailed data.
The strength is that we included data from a substantial part of the Swedish population, which allowed us to distinguish the individuals with nerve injuries in the upper or lower limbs from the ones with CRPS type 1 and CRPS type 2. Ideally, the association between CRPS diagnoses and the use of psychotropic drugs should be investigated in individuals neither with a previous CRPS diagnosis nor with previous psychotropic use. However, we adjusted for those variables in the model as this restriction was impossible to implement because the number of individuals with specific types of CRPS was relatively small. In any case, our results implied a higher risk of psychotropic drug use in patients with CRPS, particularly in those with an associated nerve injury (CRPS type 2), and in those with nerve injuries, and this information seems clinically relevant.
link
