Association Between Isolated Regional Pain and Subsequent Fibromyalgia Documentation: A Retrospective Text-Mined Cohort Study

Fibromyalgia is a common chronic pain condition characterized by widespread pain, fatigue, sleep disturbance, cognitive symptoms, and multiple somatic symptoms. Its case definition has changed substantially over the last three decades: the 1990 American College of Rheumatology (ACR) criteria emphasized widespread pain and tender points, whereas the 2010 and 2016 criteria shifted toward symptom severity and generalized pain distribution. 1–3 Contemporary clinical reviews describe fibromyalgia as a polysymptomatic syndrome that overlaps with other chronic pain, sleep, mood, and functional disorders. 4 , 5 Population estimates vary by method and setting, but large surveys consistently show that fibromyalgia is not rare and disproportionately affects women. 6 , 7 A contemporary systematic review and meta-analysis of more than 30 million individuals estimated a pooled general-population prevalence of 1.40% (95% confidence interval 0.49–2.47), with substantial heterogeneity and very low certainty; prevalence was markedly higher in selected at-risk groups, including chronic back pain (19.2%), systemic lupus erythematosus (12.6%), ankylosing spondylitis (13.6%), and several autoimmune or inflammatory conditions. 8 The biological framing of fibromyalgia has also evolved. Central sensitivity syndromes, central sensitization, and more recently nociplastic pain describe processes through which persistent peripheral nociceptive or neuropathic input may increase spinal and supraspinal excitability, facilitate temporal summation, impair endogenous inhibitory control, and contribute to spatially generalized pain. 9–13 Sleep disturbance, psychological distress, and other modulatory factors may further influence pain amplification. Prospective community studies have shown that chronic widespread pain can emerge after regional pain and in association with psychosocial distress, sleep disturbance, trauma exposure, altered pain sensitivity, and neuropathic-like symptoms. 14–20 These mechanisms are biologically plausible but cannot be identified from documentation patterns alone. The clinical pathway from an isolated pain-clinic presentation to later fibromyalgia documentation therefore remains incompletely described. Patients often enter specialty pain care with a single dominant complaint, such as low back pain, neck pain, headache, pelvic pain, or a regional musculoskeletal syndrome. Which presentations are most often followed by later fibromyalgia documentation in real-world practice is a practical question for clinicians, researchers, and trial designers. Most previous longitudinal studies of chronic widespread pain have been population-based or questionnaire-based. Those designs are essential for estimating risk in the community, but they do not fully capture the diagnostic language and referral patterns of tertiary pain practice. A pain clinic is enriched for persistent, treatment-refractory, overlapping, and diagnostically ambiguous syndromes. It is also the setting in which clinicians must decide whether a patient with apparently regional pain should be monitored for broader fibromyalgia features. Describing this transition in specialty-care notes can therefore complement epidemiologic studies by identifying documentation phenotypes that are visible to clinicians before fibromyalgia is explicitly recorded. Electronic health records (EHRs) are a natural substrate for this question because they preserve longitudinal clinical language across many encounters. Free-text phenotyping can recover information not captured by billing codes, especially in pain medicine, where the note narrative often contains anatomy, mechanism, medication history, and diagnostic uncertainty. At the same time, EHR phenotypes require transparent definitions, patient-level denominators, adequate follow-up, and cautious interpretation because documentation is not the same as biological onset. 21–24 We therefore designed a retrospective text-mined cohort study to identify patients whose first observed pain-clinic presentation contained one isolated, localized, non-fibromyalgia regional pain category and who had at least five years of observed follow-up. We aimed to quantify later fibromyalgia documentation, compare subsequent-documentation proportions across index pain categories, and describe age and sex patterns among patients with and without later fibromyalgia documentation.

We conducted a retrospective longitudinal text-mined cohort study using local pain-clinic workbook extracts containing 1,260,439 document sections from visits dated June 23, 2009 through June 1, 2026. Each row contained structured visit metadata and a free-text clinical section. The unit of longitudinal follow-up was the patient, using the ID field as the patient key after aggregate metadata review showed that ID represented 58,650 longitudinal entities, whereas Medical_Record behaved like a finer-grained document or encounter field. Patient identifiers and raw clinical text were used only in memory during processing and were not exported. For each patient, the index date was the earliest dated pain-clinic patient-day in the extract. All free-text sections from that same patient-day were aggregated conceptually for phenotype assignment. To ensure a minimum five-year observation window, eligible index dates were required to occur on or before June 1, 2021, with data available through June 1, 2026. The primary cohort required exactly one detected localized non-fibromyalgia pain category on the index date, no fibromyalgia signal on that date, and at least five years of observed follow-up, defined as a last observed pain-clinic date at least five years after index. Patients with multiple detected index pain categories were excluded to create a clearly defined inception cohort of apparently isolated presentations and to reduce inclusion of patients whose multisite pain at baseline could already represent unrecognized fibromyalgia. This criterion does not prove that fibromyalgia or widespread symptoms were absent before the captured index visit. We used prespecified bilingual English/Hebrew regular-expression dictionaries to identify fibromyalgia and candidate localized regional index pain categories. Classification was performed at the patient-day level after considering all sections recorded on the same date. A fibromyalgia signal required at least one prespecified fibromyalgia expression that was not captured within a local negation context. Expressions associated with terms such as “no”, “without”, “denies”, or “rule out”, and their Hebrew equivalents, were excluded. The first non-negated signal on a date after the index date defined subsequent fibromyalgia documentation. Index categories included axial low back/lumbar pain, neck/cervical pain, radiculopathy/sciatica, neuropathic pain, diabetic neuropathy, headache/migraine, trigeminal neuralgia, occipital neuralgia, postherpetic neuralgia, complex regional pain syndrome (CRPS)/reflex sympathetic dystrophy (RSD), pelvic pain/vulvodynia, endometriosis-associated pain, abdominal pain, cancer-related pain, joint pain/osteoarthritis, shoulder pain, knee pain, hip pain, foot/ankle pain, hand/wrist pain, myofascial/muscle pain, phantom limb pain, central/post-stroke pain, failed back surgery/post-laminectomy pain, postoperative pain, and rheumatologic/inflammatory pain. Hierarchy rules reduced double-counting: for example, diabetic neuropathy overrode generic neuropathic pain, failed back surgery overrode generic low back pain and postoperative pain, and specific neuralgias overrode generic neuropathic pain. Dictionaries were built as clinically interpretable families rather than narrow billing-code surrogates. This choice was intentional because pain notes frequently describe anatomy, syndrome names, procedures, and working diagnoses in the same section. The isolated-index rule was applied only after hierarchy suppression, so a patient with both generic neuropathic language and a more specific neuralgia could remain classified under the specific syndrome. The workflow separately tracked the number of patients with any index category, the number without fibromyalgia on the index date, patients excluded for multiple baseline categories, patients lacking later follow-up, and primary-cohort patients with later fibromyalgia documentation. These checks prevented row-level counts from being mistaken for patient-level denominators. No independent manual adjudication set or criterion-based diagnostic gold standard was available; accordingly, dictionary matches are reported as documentation signals rather than validated incident diagnoses. The outcome was the first patient-day after the index date containing a non-negated fibromyalgia signal. This was interpreted as later documentation rather than confirmed incident disease onset or fulfillment of ACR diagnostic criteria. Age and sex were inferred from explicit age/sex phrases in index-date text when present; otherwise, values were retained as missing. Follow-up time was measured from the index date to first later fibromyalgia documentation or the last observed pain-clinic date. Time to fibromyalgia documentation was reported among patients with the outcome. Analyses used patient-level denominators. We summarized continuous variables using medians and interquartile ranges (IQRs), and categorical variables using counts and percentages. Later-documentation proportions were reported with Wilson 95% confidence intervals (CIs). Category-specific results in the main table were limited to groups with at least 30 eligible patients; point estimates from smaller groups were described as imprecise and hypothesis-generating. No pairwise hypothesis tests were used to rank pain categories. Prespecified sensitivity analyses measured fibromyalgia documentation more than 90 days after index, documentation on at least two later dates, the less restrictive any-follow-up cohort, and a lower-bound denominator including isolated-index patients with five-year potential observation as patients without later documentation. Analyses were performed with reproducible local scripts that wrote only aggregate outputs. This study was approved by the Institutional Helsinki Committee of Chaim Sheba Medical Center, Tel HaShomer, Israel, from which the patients were recruited. The approval was granted under protocol number SMC-8148-21; the Ministry of Health application number was 202121003. The study was conducted in accordance with the principles of the Declaration of Helsinki and applicable Israeli Ministry of Health regulations. All data were handled in accordance with institutional privacy and data-governance requirements, and the analytic workflow was designed to avoid export of patient identifiers, row-level records, or raw clinical text.

The source extract contained 58,650 patients with valid key and date information. A detected index pain category was present in 42,819 patients, and 55,504 patients had no fibromyalgia documentation on the index date. After requiring exactly one localized non-fibromyalgia index pain category, 23,103 patients met the isolated-presentation definition. Of these, 13,391 had an index date allowing a five-year potential observation window, and 2,783 had at least five years of observed follow-up. Later fibromyalgia documentation occurred in 299 primary-cohort patients, yielding an overall later-documentation proportion of 10.74% (95% CI 9.65–11.95). Cohort characteristics stratified by later fibromyalgia documentation are summarized in Table 1 . Table 1 Cohort Characteristics by Later Fibromyalgia Documentation Characteristic All Eligible Later Fibromyalgia No Later Fibromyalgia Patients, n 2783 299 2484 Later fibromyalgia, n (%) 299 (10.74%) 299 (100.0%) 0 (0.0%) Female among known sex, % 89.44 91.76 89.17 Male among known sex, % 10.56 8.24 10.83 Sex unknown, n 159 20 139 Age known, n 1364 165 1199 Age, median (IQR), years 59.0 (45.0–69.0) 51.0 (40.0–61.0) 60.0 (46.0–70.0) Follow-up, median (IQR), years 8.42 (6.41–11.15) 9.47 (6.99–12.35) 8.32 (6.37–10.97) Time to fibromyalgia, median (IQR), years 4.49 (1.76–7.61) 4.49 (1.76–7.61) NA Notes : Age and sex were inferred from explicit index-date free-text expressions when available. Abbreviations : IQR, interquartile range; NA, not applicable. Cohort Characteristics by Later Fibromyalgia Documentation Notes : Age and sex were inferred from explicit index-date free-text expressions when available. Abbreviations : IQR, interquartile range; NA, not applicable. In the primary cohort, sex was identified in 2,624 patients and age in 1,364 patients. Women represented 89.4% of patients with known sex. Median age was 59 years (IQR 45–69), and median follow-up was 8.42 years (IQR 6.41–11.15). Patients with later fibromyalgia documentation were younger than those without it, with a median age of 51 years (IQR 40–61) versus 60 years (IQR 46–70), and had slightly longer observed follow-up, 9.47 years (IQR 6.99–12.35) versus 8.32 years (IQR 6.37–10.97). Among patients with later documentation and known sex, 91.8% were female and 8.2% were male; these percentages largely reflect the female-predominant source cohort. Median time from index presentation to fibromyalgia documentation was 4.49 years (IQR 1.76–7.61). Later-documentation proportions varied across isolated index categories ( Table 2 and Figure 1 ). The highest point estimates included myofascial/muscle pain (12 of 33; 36.36%, 95% CI 22.19–53.38), occipital neuralgia (5 of 14; 35.71%, 95% CI 16.34–61.24), hand/wrist pain (12 of 50; 24.00%, 95% CI 14.30–37.41), headache/migraine (17 of 102; 16.67%, 95% CI 10.67–25.08), shoulder pain (23 of 140; 16.43%, 95% CI 11.20–23.45), pelvic pain/vulvodynia (5 of 31; 16.13%, 95% CI 7.09–32.63), CRPS/RSD (8 of 51; 15.69%, 95% CI 8.17–28.01), and neck/cervical pain (13 of 88; 14.77%, 95% CI 8.84–23.65). Low back/lumbar pain was the largest index group, with 114 patients with later documentation among 1,303 patients (8.75%, 95% CI 7.33–10.41). Radiculopathy/sciatica accounted for 53 patients with later documentation among 565 patients (9.38%, 95% CI 7.24–12.07). These descriptive estimates were not used to infer pairwise differences between categories. Table 2 Later Fibromyalgia by Isolated Index Pain Category Index Pain Category Patients Later FM Later Documentation % (95% CI) Follow-Up y, Median Time to FM y, Median Myofascial/muscle pain 33 12 36.36% (22.19–53.38) 7.66 2.13 Hand/wrist pain 50 12 24.0% (14.3–37.41) 8.7 5.23 Headache/migraine 102 17 16.67% (10.67–25.08) 7.65 2.63 Shoulder pain 140 23 16.43% (11.2–23.45) 7.45 5.42 Pelvic pain/vulvodynia 31 5 16.13% (7.09–32.63) 8.89 6.65 CRPS/RSD 51 8 15.69% (8.17–28.01) 8.85 4.69 Neck/cervical pain 88 13 14.77% (8.84–23.65) 8.61 5.39 Trigeminal neuralgia 33 4 12.12% (4.82–27.33) 8.36 3.61 Neuropathic pain 147 14 9.52% (5.76–15.35) 8.19 2.66 Radiculopathy/sciatica 565 53 9.38% (7.24–12.07) 7.69 4.12 Low back/lumbar pain 1303 114 8.75% (7.33–10.41) 9.14 5.45 Knee pain 46 4 8.7% (3.43–20.32) 7.58 5.43 Abdominal pain 49 3 6.12% (2.1–16.52) 7.94 2.98 Hip pain 36 2 5.56% (1.54–18.14) 8.04 1.85 Notes : The table displays categories with at least 30 eligible patients in the five-year observed-follow-up cohort. Smaller categories are described in the Results and should be treated as hypothesis-generating because of limited precision. Abbreviations : CI, confidence interval; CRPS, complex regional pain syndrome; FM, fibromyalgia; RSD, reflex sympathetic dystrophy. Figure 1 Later fibromyalgia documentation after isolated localized regional pain. Bars show the percentage of patients with subsequent fibromyalgia documentation by isolated index pain category; labels report the number with later documentation and the eligible denominator. Categories displayed had at least 30 eligible patients in the five-year observed-follow-up cohort. A horizontal bar graph showing later fibromyalgia documentation by isolated index pain category. X-axis: Later fibromyalgia documentation (%), range 0-40. Y-axis: Categories. Bar labels: % and counts. Myofascial/muscle pain: 36.36% (12/33). Hand/wrist pain: 24.00% (12/50). Headache/migraine: 16.67% (17/102). Shoulder pain: 16.43% (23/140). Pelvic pain/vulvodynia: 16.13% (5/31). CRPS/RSD: 15.69% (8/51). Neck/cervical pain: 14.77% (13/88). Trigeminal neuralgia: 12.12% (4/33). Neuropathic pain: 9.52% (14/147). Radiculopathy/sciatica: 9.38% (53/565). Low back/lumbar pain: 8.75% (114/1303). Knee pain: 8.70% (4/46). Abdominal pain: 6.12% (3/49). Hip pain: 5.56% (2/36). Later Fibromyalgia by Isolated Index Pain Category Notes : The table displays categories with at least 30 eligible patients in the five-year observed-follow-up cohort. Smaller categories are described in the Results and should be treated as hypothesis-generating because of limited precision. Abbreviations : CI, confidence interval; CRPS, complex regional pain syndrome; FM, fibromyalgia; RSD, reflex sympathetic dystrophy. Later fibromyalgia documentation after isolated localized regional pain. Bars show the percentage of patients with subsequent fibromyalgia documentation by isolated index pain category; labels report the number with later documentation and the eligible denominator. Categories displayed had at least 30 eligible patients in the five-year observed-follow-up cohort. Several high point estimates were based on small denominators and were interpreted cautiously. Occipital neuralgia and endometriosis-associated pain had high observed proportions but fewer than 20 eligible patients each, resulting in wide CIs and limited precision. By contrast, headache/migraine, shoulder pain, radiculopathy/sciatica, neuropathic pain, and low back/lumbar pain had more stable denominators. Categories with no observed later documentation, such as diabetic neuropathy and central/post-stroke pain in this five-year cohort, had very small denominators and should not be interpreted as protective. The descriptive later-documentation proportions were 10.91% among women and 8.30% among men. Because the cohort was predominantly female and no adjusted sex-effect analysis was performed, these data should not be interpreted as demonstrating female predominance in subsequent documentation. By age group, the later-documentation proportion was highest in patients aged 30–44 years (22.58%, 95% CI 17.52–28.59) and 45–59 years (17.08%, 95% CI 13.56–21.29), lower in patients younger than 30 years (7.83%, 95% CI 4.17–14.21), and lowest in patients aged 75 years or older (0.60%, 95% CI 0.11–3.29). These descriptive estimates are summarized in Table 3 . Table 3 Later Fibromyalgia by Sex and Age Strata Stratifier Level Patients Later FM Later Documentation % 95% CI Sex Female 2347 256 10.91 9.71–12.23 Sex Male 277 23 8.3 5.6–12.15 Sex Unknown 159 20 12.58 8.29–18.63 Age group <30 115 9 7.83 4.17–14.21 Age group 30–44 217 49 22.58 17.52–28.59 Age group 45–59 363 62 17.08 13.56–21.29 Age group 60–74 501 44 8.78 6.61–11.58 Age group 75+ 168 1 0.6 0.11–3.29 Age group Unknown 1419 134 9.44 8.03–11.08 Notes : Percentages are descriptive. Abbreviations : CI, confidence interval; FM, fibromyalgia. Later Fibromyalgia by Sex and Age Strata Notes : Percentages are descriptive. Abbreviations : CI, confidence interval; FM, fibromyalgia. As summarized in Table 4 , requiring fibromyalgia documentation more than 90 days after the index date produced a similar later-documentation proportion of 10.67% (297 of 2,783; 95% CI 9.58–11.87). Requiring fibromyalgia documentation on at least two later dates yielded a more specific estimate of 8.19% (228 of 2,783; 95% CI 7.23–9.27). The less restrictive any-follow-up cohort reproduced the earlier estimate of 4.01% (644 of 16,076; 95% CI 3.71–4.32). Including all 13,391 isolated-index patients with five-year potential observation as patients without later documentation gave a conservative lower-bound estimate of 3.61% (484 of 13,391; 95% CI 3.31–3.94). Table 4 Sensitivity Analyses Analysis Patients Later FM Later FM, % 95% CI Primary: any later fibromyalgia documentation; requires ≥5 years observed follow-up 2783 299 10.74 9.65–11.95 Outcome lag: fibromyalgia first documented >90 days after index 2783 297 10.67 9.58–11.87 Specificity: fibromyalgia documented on at least two later dates 2783 228 8.19 7.23–9.27 Comparator: isolated-index patients with any later follow-up 16076 644 4.01 3.71–4.32 Lower-bound denominator: includes isolated-index patients with five-year potential observation with no later documentation 13391 484 3.61 3.31–3.94 Abbreviations : CI, confidence interval; FM, fibromyalgia. Sensitivity Analyses Abbreviations : CI, confidence interval; FM, fibromyalgia.

In this five-year longitudinal pain-clinic cohort, approximately one in nine patients with an isolated, localized, non-fibromyalgia regional pain presentation and at least five years of observed follow-up had subsequent documentation consistent with fibromyalgia. This finding represents an association between patterns of clinical documentation, not evidence that the index pain caused fibromyalgia or that biological disease onset occurred during follow-up. The overall proportion was higher than in the less restrictive any-follow-up cohort, consistent with reduced early censoring and greater opportunity for fibromyalgia to be recognized. Myofascial/muscle pain and several regional presentations, including hand/wrist pain, headache/migraine, shoulder pain, pelvic pain/vulvodynia, CRPS/RSD, and neck/cervical pain, showed higher-than-average point estimates, although estimates from small categories were imprecise. These results fit the broader conceptual shift from viewing fibromyalgia as a tender-point syndrome to viewing it as a polysymptomatic disorder considered within central sensitization and nociplastic pain frameworks. 1–13 Persistent peripheral input can plausibly promote central neural plasticity, temporal summation, and impaired descending inhibition, while sleep and affective disturbances may amplify symptoms. However, the present EHR design did not measure these mechanisms. A localized complaint may therefore represent a genuinely regional condition followed by broader symptoms, an early focal manifestation of an evolving polysymptomatic disorder, or delayed recognition and documentation of pre-existing fibromyalgia. The study cannot distinguish among these explanations and maps documentation trajectories rather than biological progression. The high later-documentation proportion after myofascial/muscle pain is clinically plausible but not mechanistically established. Myofascial pain shares clinical language with tenderness, widespread muscle pain, trigger points, fatigue, and diffuse pain descriptions and may reflect either an early regional phenotype or partial documentation of already evolving centralized pain. Similarly, headache/migraine, pelvic pain, and selected neuralgias have been discussed within overlapping pain and central sensitivity frameworks. 9–13 The hand/wrist category was necessarily heterogeneous and could include regional musculoskeletal, joint, soft-tissue, neuropathic, inflammatory, or early multisite symptoms documented through a focal complaint. Its point estimate (12/50; 24.00%, 95% CI 14.30–37.41) should not be interpreted as higher than that for CRPS/RSD (8/51; 15.69%, 95% CI 8.17–28.01): the CIs overlap substantially, the counts are small, and no pairwise comparison was prespecified. These categories should be considered sentinel documentation phenotypes rather than causal exposures. The distinction between observed proportion and absolute burden is also important. Low back/lumbar pain and radiculopathy/sciatica had lower later-documentation proportions than several smaller categories, but because they were common, they contributed many later fibromyalgia cases in absolute terms. From a health-system perspective, a moderate observed proportion in a common localized presentation may still identify many patients who could benefit from systematic symptom assessment. Mechanistic interpretations are not supported by these data; higher point estimates may instead identify phenotypes in which nociplastic features, overlapping symptoms, or diagnostic uncertainty are more readily documented. Women constituted most patients with later fibromyalgia documentation, but women also constituted 89.4% of the cohort with known sex. The later-documentation proportions were 10.91% in women and 8.30% in men; this descriptive difference was not adjusted for age, referral patterns, follow-up, or other covariates and should not be presented as evidence of an independent female effect. Later documentation was highest in middle adulthood rather than older age. Age and sex were extracted from free text rather than structured demographic fields, and age was missing in a substantial proportion. The low proportion in patients aged 75 years or older may reflect competing diagnoses, shorter follow-up, lower likelihood of applying a fibromyalgia label in older adults, or different documentation practices. The five-year requirement changes the interpretation of the estimate. It does not create a population incidence rate, because patients still needed observed pain-clinic follow-up, but it does reduce the problem of labeling short-follow-up patients as nonconverters. The any-follow-up comparator produced a much lower proportion, illustrating how early censoring can dilute observed fibromyalgia documentation. The lower-bound analysis, which treated all patients with five-year potential observation as nonconverters, provides a conservative floor rather than a preferred estimate. The study has several strengths. It used a large specialty pain dataset spanning more than 16 years, patient-level longitudinal ordering, same-day aggregation at the index visit, a minimum observed five-year follow-up requirement, bilingual dictionaries, hierarchy rules to reduce double-counting, negation handling for fibromyalgia, and sensitivity analyses that tested temporal lag, outcome specificity, any follow-up, and a conservative lower-bound denominator. The design directly addresses a clinically actionable question: when a patient presents with isolated, localized regional pain, which documented phenotypes are most often followed by later fibromyalgia documentation? Limitations are important. First, this was a retrospective text-mined cohort, not a manually adjudicated incidence study. The outcome reflects documentation consistent with fibromyalgia, not necessarily first biological onset, fulfillment of ACR criteria, or a newly developed disorder. Delayed recognition of pre-existing widespread symptoms and early focal manifestations of fibromyalgia cannot be distinguished from true progression. Second, left truncation is likely: some patients may have had fibromyalgia symptoms or diagnoses before their first captured pain-clinic visit. Third, patients with longer follow-up had more opportunity for outcome detection, and patients with later documentation had longer observed follow-up. Fourth, dictionary-based phenotyping can misclassify spelling variants, copied-forward text, differential diagnoses, family history, or historical diagnoses despite negation filters; no independent manual gold standard was available. Alternative or overlapping conditions, including hypermobile Ehlers-Danlos syndrome and small fiber neuropathy, may also have contributed to symptoms or diagnostic misclassification. Fifth, the localized categories were phenotype-based and etiologically heterogeneous. The available text-derived variables did not permit uniform classification of trauma, inflammation, infection, visceral disease, or other causes of regional pain, and the hand/wrist category did not reliably distinguish joint, soft-tissue, or nerve-related pain. Sixth, potential confounders such as autoimmune or inflammatory disease history, sleep and mood disorders, baseline symptom burden, medication exposure, and healthcare utilization were not systematically assessed and may be associated with both the index presentation and later fibromyalgia documentation. Seventh, age and sex were inferred from free text and missing in a meaningful fraction of patients. Eighth, the cohort reflects a specialty pain-clinic population and should not be generalized to population incidence without external validation. These findings suggest practical next steps. Clinically, patients presenting with isolated myofascial/muscle pain, headache/migraine, pelvic pain, selected neuralgias, and regional upper-limb pain may benefit from structured screening for widespread pain, sleep disturbance, fatigue, cognitive symptoms, and symptom severity at follow-up. For research, the dictionary-derived sentinel categories can be used to design prospective validation studies with manual chart review, structured ACR 2016 criteria, patient-reported outcomes, and time-to-event modeling. For informatics, the study demonstrates how bilingual clinical text can be converted into patient-level longitudinal phenotypes while preserving privacy and avoiding row-level export. A prospective follow-up study should not simply repeat the present text-mining approach. The next step would enroll patients at the time of an isolated regional pain presentation, measure pain distribution and symptom severity directly, collect sleep and mood measures, and follow participants at prespecified intervals. That design would distinguish true incident fibromyalgia from delayed documentation, estimate incidence rates with person-time, and test whether the sentinel categories identified here remain predictive after adjustment for baseline symptom burden, treatment intensity, and follow-up frequency. In conclusion, later fibromyalgia documentation occurred in 10.74% of pain-clinic patients whose first observed presentation was an isolated, localized non-fibromyalgia regional pain problem and who had at least five years of observed follow-up. The observed proportions varied across index presentations, with substantial uncertainty in small subgroups. Myofascial/muscle pain and several regional pain documentation phenotypes may help identify patients for earlier assessment of widespread pain and associated symptoms. These associations do not establish incident disease, biological progression, or causality and require prospective validation with structured diagnostic criteria and measurement of relevant confounders.