Причины использования медицинских услуг до появления множественных склерозов: Общенациональное совпадающее исследование кохарта в Швеции.

Introduction

Over the past decade, evidence from epidemiological studies suggests that changes in health and behavior precede the clinical presentation of multiple sclerosis (MS). For example, studies reported an increase in healthcare use, including higher rates of hospitalizations, physician visits, and prescription drug dispensations before the classical MS onset [ 1 , 2 , 3 , 4 ]. Individuals who later developed MS experienced non‐MS‐specific symptoms or health issues more frequently compared to the matched general population. These included fatigue, pain, migraine, gastric and urinary disturbances, mood disturbances such as depression and anxiety, and cognitive impairment [ 1 , 2 , 5 , 6 , 7 , 8 , 9 , 10 , 11 , 12 , 13 ]. The rise in serum neurofilament light chain levels 6–9 years before the first clinical manifestations of MS has also provided biological evidence of neuronal injury [ 14 , 15 ]. Collectively, these findings support the existence of an MS prodromal phase.

Identifying and understanding this early phase of disease carries significant implications, including opportunities to advance the understanding of disease mechanisms, support risk assessment, enable more timely diagnosis and intervention, and consequently improve long‐term outcomes. However, well‐defined markers for identifying the MS prodrome remain lacking, and determining the length of the prodrome window continues to be challenging. Further, as MS is a highly heterogeneous disease, a uniform prodromal phase across diverse populations may not exist. Therefore, discussions about the potential utility of recognizing the MS prodrome for advancing research and improving patient care are ongoing [ 16 , 17 ], underscoring the need for further investigation in this area.

In this study, we aim to better characterize the prodromal phase of MS by examining health services utilization patterns in the Swedish MS population. Specifically, we investigated reasons for seeking specialist and hospital care, as well as patterns of prescription medication use prior to clinically determined MS symptom onset or the first demyelinating event. In addition, we explore potential heterogeneities related to sex and age to understand how these factors may influence the prodromal phase.

Methods

Data Sources

We utilized comprehensive, prospectively collected nationwide health administrative data from Sweden. The Swedish Multiple Sclerosis Registry (SMSreg), a national quality register covering approximately 80% of Sweden's MS population, provided neurologist‐recorded MS symptom onset dates [ 18 , 19 ]. We also accessed health administrative data encompassing hospital admissions, outpatient physician visits (i.e., specialist visits), and prescriptions dispensed in community pharmacies throughout Sweden. Specifically, we sourced hospitalization and outpatient visit data from the National Patient Register, and dispensation information from the National Prescribed Drug Register [ 20 ]. Sex, birth and death dates, immigration and emigration records, and county of residence were obtained from the Total Population Register [ 21 ]. Annual family income data were extracted from the Longitudinal Integrated Database for Health Insurance and Labour Market Studies (LISA) [ 22 ]. Data from all sources spanned from January 1, 2001, to December 31, 2019, except for prescription data which were available from July 1, 2005. The data linking is facilitated by a unique Personal Identity Number maintained by the Swedish Tax Agency.

Study Design and Cohorts

In this matched cohort study, we established two study cohorts: a clinical cohort and an administrative cohort. The clinical cohort comprised individuals registered in the SMSreg diagnosed with MS by a neurologist, with the index date defined as the earliest occurrence of either the neurologist‐recorded MS symptom onset date or the date of the first demyelinating event indicative of MS, as identified through outpatient or hospital diagnoses coded according to the Swedish version of the International Classification of Diseases (ICD) systems (ICD‐9‐SE codes 323X, 340, 341A/W/X, 377D (before 1997) or ICD‐10‐SE codes G35, H46‐H48, G37.3, G36) [ 4 ]. The administrative cohort was constructed using a validated algorithm applied to hospital and outpatient care data, requiring three or more ICD‐9/10‐SE codes for MS (340/G35) [ 23 ]. The index date for this cohort was the date of the first MS/demyelinating disease‐related ICD code (see above). Individuals could be included in both cohorts if they met each criterion. Eligibility criteria for both cohorts mandated that all individuals had resided in Sweden for a minimum of five consecutive years immediately before their index date through until their MS diagnosis date (clinical cohort) or third MS ICD code (administrative cohort). Each person with MS (PwMS) was matched with up to five randomly selected individuals from the general population of the same sex, birth year, county of residence at the index year, and duration of residency in Sweden. Matched individuals were assigned the same index date as the matched PwMS and had to fulfill the same residency requirement and have no MS/demyelinating disease‐related ICD codes.

Reasons for Healthcare Use

Healthcare use was measured as (1) physician outpatient and hospital visits, grouped according to the ICD‐10 chapter (using the primary code in the outpatient register or discharge code in the inpatient register, Table S1 ) and (2) prescription dispensed, grouped according to the World Health Organization's Anatomical Therapeutic Chemical (ATC) classification system main anatomical group (1st level, Table S2 ), and pharmacological or therapeutic subgroups (2nd level) were also investigated when feasible.

Statistical Analysis

The clinical and administrative cohorts were analyzed separately. For each ICD chapter or ATC group (1st and 2nd level), the MS and non‐MS comparators were first compared by the total number of events in the 5 years pre‐index date, enabling comparisons with prior work. Guided by the 5‐year analysis, the ICD chapters or ATC groups were then interrogated annually as far back pre‐index as was feasible. Overdispersed‐Poisson regressions were used for comparison with the logarithm of person‐time included as an offset. The overdispersion was accounted for using scaled Pearson's chi‐squared statistic. The primary comparisons were based on the overall cohort. Secondary comparisons were carried out by sex and age groups at the index date (18–29, 30–49, and ≥ 50 years). Pediatric onset‐related (< 18 years old at index) findings are reported elsewhere [ 24 ]. The analysis of pregnancy and childbirth‐related visits was restricted to females aged ≤ 45 at the index date. Analyses were conducted using R (v.4.3.1, R Foundation for Statistical Computing, Vienna, Austria).

Ethics

Institutional ethical approval was gained (Regional Ethical Review Board, Karolinska Institutet/Swedish Ethical Review Authority/University of British Columbia's Clinical Research Ethics Board [Dnr: 2017/1378‐31/H20‐03232‐A007]).

Results

The clinical cohort included 7604 PwMS and 37,974 matched population comparators, with a median age of 35.5 (interquartile range: 27.3–44.7) at the index date, Table 1 . The administrative cohort comprised 10,269 PwMS and 51,297 comparators. Their median age was 5 years older than the clinical cohort. Approximately two‐thirds of each cohort are female. Supplementary Results in Appendix S1 provide additional details. The mean pre‐index observation period was 11 years for both cohorts.

Outpatient Visits and Hospitalizations by ICD Chapters

Clinical Cohort

We focus on chapters with significant overall 5‐year findings for outpatient visits (Figure 1 ). For these chapters, we also summarize corresponding results for hospitalizations, notable sex and age differences, and yearly trends. The chapters are organized by overall 5‐year outpatient visit RR, starting from the highest.

During the 5 years preceding MS symptom onset, visits for diseases of the sense organs were considerably higher in the MS cohort (RR = 1.73 for outpatient; 1.71 for hospital), followed by ill‐defined symptoms/signs (RR = 1.47 both outpatient and hospital), nervous system (RR = 1.33 for outpatient; 1.92 for hospital), and musculoskeletal system (RR = 1.17 for outpatient; 1.28 for hospital). Outpatient visits were modestly elevated for mental disorders, digestive system, genitourinary system, and other health system contact (RRs range = 1.11 to 1.12). Pregnancy/childbirth‐related visits among females were lower in the MS cohort (RR = 0.92 for outpatient; 0.90 for hospital).

Distinct sex and age differences emerged (Figures S2 and S3 ): for sense organs, the elevation in outpatient visits was more substantial in females (RR = 1.90 [95% confidence interval: 1.74–2.08] vs. males RR = 1.35 [1.16–1.58]). Although the outpatient RRs were similar by sex for ill‐defined symptoms/signs and the genitourinary system, the annual rate among non‐MS was higher in females than in males (Table S3 ), hence the absolute increases differed by sex. For mental disorders, an elevation in hospitalizations was suggested in males (RR = 1.30 [0.98–1.73]) but not in females (RR = 0.70 [0.54–0.92]). Nervous system‐related outpatient visits were elevated in those aged 30–49 and ≥ 50 (RR = 1.48 and 1.72), but not in those aged 18–29 (RR = 0.82), whereas hospitalizations for the genitourinary system were increased in those aged 18–29 (RR = 1.58 [1.12–2.24]), but not in the older groups.

Annual comparisons were conducted up to 19 years pre‐index when feasible. As no consistent patterns were observed during years 11 to 19, only results for the 10 years leading to the index date were displayed (Figure 2 ; Figures S4 and S5 ). For outpatient visits, a higher rate for sense organs persisted up to 6 years pre‐onset; RRs ranged from 1.34 to 1.56 from year −2 to −6 and rose to 2.84 in year −1. Yearly RRs for the sense organs were significant up to 6 years pre‐onset for females (ranging from 1.53 to 3.01) and only 1 year pre‐onset for males (RR = 2.45). The rate was also consistently elevated for ill‐defined symptoms/signs, ranging from 1.17 to 1.32 from year −5 to year −2 and rising to 2.28 in the year before the index. The RRs for the musculoskeletal system ranged from 1.07 to 1.26 over the 5 years pre‐index, with years −1 and −5 being significant. Although not significant, the yearly RRs for mental and behavioral disorders were consistently above one, ranging between 1.07 to 1.17 over the 5 years pre‐index. The RRs for the digestive system in the first 3 years were above one (RR = 1.17–1.24), with year −3 being significant. The RRs for the genitourinary system in years −1 to −4 ranged from 1.06 to 1.20, with years −1 and −3 being significant. The elevations for the nervous system and other health system contact were driven by a sharp rise in the year pre‐index (RR range = 1.26 to 2.43).

The increases in hospitalizations mainly occurred in the year pre‐index with the most pronounced rise for visits related to the nervous system and sense organs—nearly seven times higher than the non‐MS comparators (RR = 6.96 [4.81–10.06] and 6.91 [4.17–11.47]) (Figure S5 ). Elevations were also observed for ill‐defined symptoms/signs (3.24 [2.59–4.04]), other health system contact (1.97 [1.14–3.42]), musculoskeletal system (1.54 [1.06–2.23]), and digestive system (1.42 [1.04–1.96]).

Administrative Cohort

Generally, similar patterns were observed in the administrative cohort with higher magnitudes and longer durations (Figures S1–S3 , S6 , S7 , and Supplementary Results in Appendix S1 ). For example, for outpatient visits, the highest elevation over the 5 years was related to the nervous system (RR = 3.70, Figure S1 ), and a higher rate relative to the non‐MS cohort was detected 9–10 years pre‐index for sense organs and ill‐defined symptoms/signs (Figure S6 ).

Prescription Dispensations by ATC Level

Due to a later data start date, 5055/25,235 individuals with/without MS from the clinical cohort and 6501/32,467 individuals from the administrative cohort contributed to the analyses.

Clinical Cohort

We focus on main anatomical groups (1st level) with significant overall 5‐year findings (Figure 3 ). For these groups, we also summarize sex and age differences, yearly trends, and significant findings at the therapeutic subgroup level (2nd level) (Figures S9–S13 , and Tables S4 and S5 ). The groups are organized by overall 5‐year RR, starting from the highest.

By the main anatomical groups, there were significantly more dispensations in the 5‐year pre‐index for drugs for the nervous system (RR = 1.23), musculoskeletal system (RR = 1.22), blood and blood‐forming organs (RR = 1.18), alimentary tract and metabolism (RR = 1.14), sensory organs (RR = 1.10), respiratory system (RR = 1.06), and dermatological agents (RR = 1.06). While no noticeable sex differences in terms of RRs were observed (Figure S9 ), the dispensation rates in the non‐MS cohort were generally higher in females than in males (Table S4 ). Distinct age‐specific patterns emerged: increases were more marked in the 18–29 and 30–49 age groups for blood‐related, metabolism, and respiratory system agents than in the ≥ 50 age group (Figure S10 ).

When assessed by year, rates for the nervous system drugs were moderately increased up to year −7 (RR = 1.17–1.27; Figure S11 ); the duration appeared to be shorter in males, extending only up to year −5. Significant or near significant increases were seen in all the 2nd level subgroups (Table S5 ), except for other nervous system drugs (N07). In particular, elevated analgesic dispensation s (N02) were significant up to 5 years pre‐index (RR = 1.18–1.39), psycholeptics (N05) up to 7 years (RR = 1.24–1.36), and psychoanaleptics (N06) up to 3 years as well as in years −6 and −7 (RR = 1.12–1.19), Figure S13a .

Dispensations for musculoskeletal drugs were increased up to year −6 (RR ranged from 1.14 to 1.38). This was driven by antiinflammatory/antirheumatic , joint and muscular pain ‐related drugs, and muscle relaxants (M01–M03; 5‐year RR = 1.22–1.34). A ntiinflammatory/antirheumatic‐related drugs (M01) were elevated up to 9 years pre‐index (RR = 1.11–1.37), significant in all years except for year −7 (Figure S13b ). Meanwhile, dispensations for antigout agents (M04) were noticeably lower (5‐year RR = 0.35 [0.20–0.62]). The RRs of drugs for the alimentary tract and metabolism ranged between 1.05 and 1.25 during years −1 to −6, with years −1, −4, and −5 reaching statistical significance. Drugs for gastrointestinal disturbances (A01‐03, 06), antiobesity (A08), and vitamins (A11) were more frequently dispensed in the MS versus non‐MS cohort. Blood and blood‐forming organ‐related drug dispensations were only increased in year −1 (RR = 1.36 [1.20–1.54]), which included elevations for antianemic preparations (B03) and antihemorrhagics (B02). Dispensations for sensory organ‐related drugs increased in year −1 (RR = 1.22 [1.07–1.39]), particularly for ophthalmological agents (S01; RR = 1.29 [1.11–1.51]). Drugs for the respiratory system increased in year −1 (RR = 1.09 [1.01–1.18]), particularly for cough and cold preparations (R05; RR = 1.32 [1.15–1.52]).

Within the genitourinary system and sex hormones group, there was a 60% increase in urological agents (G04) in the 5 years pre‐index (5‐year RR = 1.60 [1.39–1.85]; yearly RRs for years −5 to −1 ranged from 1.31–1.83; Figure S14a ), and were similar by sex and age. Among females, there was a mild but significant increase in prescriptions dispensed for gynecological antiinfective agents and antiseptics and other gynecological agents (G01‐G02; 5‐year RR = 1.26 and 1.09, respectively) with the yearly RR for G01 peaking around year −4 (RR = 1.54 [1.20–1.99]; Figure S14b ), followed by year −6 (RR = 1.39 [1.05–1.83]).

At the ATC 2nd level, there was a significantly higher dispensation rate for calcium channel blockers (C08) (particularly in the 30–49 age group, RR = 1.55 [1.25–1.92]), antifungals for dermatological use (D01), antiacne agents (D10), and antimycotics for systemic use (J02) in the 5 years pre‐index (RR range = 1.12 to 1.21).

Administrative Cohort

Generally, higher and more prolonged increases were observed in the administrative cohort (Table S5 , Figures S8–S10 and S12–S14 ). Significant elevations in the 5 years pre‐index were observed in all anatomical groups except for cardiovascular‐related drugs and antiparasitics. The elevation in drugs for the nervous and musculoskeletal systems extended to 10 years pre‐index.

Discussion

In this population‐based matched cohort study, we observed increased specialist and hospital care for several ICD chapters, 1 to 6 years before clinical symptom onset, most noticeably for visits related to the sense organs , nervous system , and ill‐defined symptoms/signs . We also found higher rates of prescription drugs dispensed, including urological agents , muscle relaxants, analgesics , and psycholeptics , 5 to 7 years before clinical onset. These findings were supported by the comparisons in the years before the first demyelinating event in the larger administrative cohort.

Previous studies of these Swedish cohorts found a rise in annual general healthcare use and sickness absence as MS onset approached [ 4 , 25 ]. In particular, the overall outpatient visits were consistently higher than their non‐MS comparators up to 5 years pre‐index in the clinical cohort and 8 years in the administrative cohort. The current study offers a more granular view of reasons for healthcare use before the index date, capturing subtle temporal changes specific to different anatomic systems. Our findings corroborate that many symptoms and issues commonly associated with MS might manifest well before the clinical symptom onset. For example, sense organ‐related issues led to increased outpatient visits up to 6 years before symptom onset and 9 years before the first demyelinating event. Urological disturbance, reflected in higher use of urological agents, was observed up to 5 years pre‐index date in the clinical cohort and 11 years pre‐index in the administrative cohort.

Musculoskeletal system‐related issues were evident in increased drug use up to 6 to 9 years pre‐index dates. Similarly, psychiatric issues, marked by elevated use of psycholeptics and psychoanaleptics , can be traced 7 years pre‐index date in the clinical cohort and 10 years pre‐index in the administrative cohort. The use of analgesics was increased 5 years pre‐index in the clinical cohort and 12 years pre‐index in the administrative cohort, suggesting early management of pain‐related issues, which are common after MS diagnosis. A higher use of antianemic agents may suggest anemia. These findings concur with, while greatly extending results from previous studies [ 1 , 2 , 7 , 8 , 9 , 11 , 13 , 26 ], most of which relied on administrative case definitions to identify PwMS. One study [ 2 ] reported increases in sense organ‐related visits, as well as urological and musculoskeletal drugs in a clinical cohort. These studies often lacked an observation period beyond 5 years or did not examine the patterns yearly.

The increase in the digestive system‐related visits and drug use also aligns with previous findings [ 1 , 2 , 12 ]. We did not see an increase in circulatory system‐related visits or the overall use of cardiovascular drugs. However, calcium channel blockers were more frequently dispensed among PwMS. This may suggest more prevalent hypertension among those who later developed MS, possibly signaling autonomic dysfunction or cardiovascular risk. However, hypertension‐related findings are inconsistent in the previous studies [ 1 , 27 ]. Others have reported dermatological issues [ 2 ]; while we did not see an increase in skin‐related visits, specific dermatological agents were more frequently used in PwMS, possibly a manifestation of the underlying inflammatory activity or hormonal changes.

There were clear elevations in visits recorded as ill‐defined signs and symptoms and abnormal laboratory results, detectable up to 5 years pre‐index in the clinical cohort and 10 years in the administrative cohort. These represent conditions with unknown etiology or transient nature, for example, fatigue which has been reported pre‐index [ 1 , 9 , 28 ]. A more detailed analysis of conditions within this chapter could improve our understanding. A Canadian study [ 26 ] reported a higher occurrence of symptoms involving the nervous and musculoskeletal systems, head and neck, and general symptoms in PwMS.

We found a lower use of anti‐gout agents. A previous study reported a negative association between gout and neurodegenerative diseases, including MS, suggesting that uric acid plays a role in MS development [ 29 ]. Similar to previous studies [ 2 , 30 ], we found fewer pregnancy and childbirth‐related visits in PwMS, which may reflect changes in the health‐related behavior of females who later developed MS.

Our study reveals heterogeneities in the prodrome associated with sex and age, which had been investigated in few studies. For example, encounters related to sense organs rose earlier and more sharply in females, while the elevation in mental/behavioral‐related encounters tended to be higher in males. Similar to a Canadian study [ 31 ], we observed a higher RR among males for genitourinary prescriptions. However, analysis of the clinical cohort by therapeutic subgroups revealed compatible increases in urological agents between sexes and milder increases in gynecological drugs, accounting for the disparity at the anatomical group level. This prior study also reported that cardiovascular drug dispensations were more elevated among those under 50, a trend we also observed. Additionally, we found increases were more pronounced among the younger individuals for blood‐related, metabolism, and respiratory system dispensations, which, to our knowledge, have not been reported before.

The strengths of this study include the large, population‐based clinically identified MS cohort, up to 19 years of longitudinal secondary care data, 14 years of drug dispensation data, and a matched study design. Neurologist‐recorded symptom onset reduced the risk of bias from diagnosis delays or changes in the diagnostic criteria. The consistent findings from the administrative cohort, which may include underrepresented individuals in the clinical cohort, enhance the study's reliability and facilitate comparison with previous studies.

However, we could not examine signs or symptoms untreated or without requiring secondary care or a prescription medication. The absence of primary care data could lead to an underestimation of the scope and duration of the MS prodrome. We also lacked the information needed to contextualize the reasons for drug use. Despite the matched design, potential confounders, for example, socioeconomic status and lifestyle factors, may exist.

In summary, our study not only validates previous findings, but substantially extends them by exploring much longer periods before two different, but complementary, index dates in two cohorts (clinical and administrative), and by incorporating sex‐ and age‐specific analyses. Key early indicators included issues related to sensory organs, urological and musculoskeletal conditions, mental health, and ill‐defined signs and symptoms, emerging 6 to 9 years before clinical symptom onset. However, their manifestation varied by age and sex. Translating this knowledge into clinical practice will be a key focus of future research. First, profiling individuals at risk for MS could guide who should undergo neurological evaluation and testing—such as MRI or serum biomarkers (e.g., neurofilament light chain and glial fibrillary acidic protein)—and when. To ensure both sensitivity and specificity, additional research is needed to understand the mechanisms underlying these healthcare use patterns and their potential role in MS development. Second, examining whether the frequency, type, and timing of healthcare use predict disease course and subsequent disability progression will aid the development of patient‐specific interventions and management strategies.

Author Contributions

All authors contributed to the conception and design of the study. A.M., Y.Z., and F.Z. contributed to the acquisition and analysis of data. A.M., Y.Z., and H.T. contributed to drafting a significant portion of the manuscript or figures. All authors contributed to interpreting the results and reviewing the manuscript.

Funding

The study was funded in part by: the National MS Society and MS Canada (formerly Multiple Sclerosis Society of Canada) (RG5063A4/1/RFA‐2103‐37392; EGID: P002/903124/3743; PI: Tremlett). Funders had no role in the methods, analyses, manuscript preparation, or decision to submit for publication.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Appendix S1: ene70464‐sup‐0001‐AppendixS1.pdf.

Acknowledgments

We would like to acknowledge the members of the team who facilitated gaining funding for the wider program of work related to the MS prodrome (Drs Jan Hillert, M. Ehsan Karim, José Wijnands, and Fardowsa L.A. Yusuf). JW also contributed to designing the technical plan whilst a research associate at the University of British Columbia, Canada.

Manouchehrinia A., Zhao Y., Zhu F., et al., “Reasons for Healthcare Use Before Onset of Multiple Sclerosis: A Nationwide Matched Cohort Study in Sweden,” European Journal of Neurology 32, no. 12 (2025): e70464, 10.1111/ene.70464.

Ali Manouchehrinia and Yinshan Zhao contributed equally.

Data Availability Statement

The data that support the findings of this study are provided by the Swedish MS Registry, National Board of Health, and Statistics Sweden. Data‐sharing agreements prohibit public availability, although the relevant data steward may grant access.

References

1. Disanto G., Zecca C., MacLachlan S., et al., “Prodromal Symptoms of Multiple Sclerosis in Primary Care,” Annals of Neurology 83, no. 6 (2018): 1162–1173.

2. Wijnands J. M., Zhu F., Kingwell E., et al., “Five Years Before Multiple Sclerosis Onset: Phenotyping the Prodrome,” Multiple Sclerosis 25, no. 8 (2019): 1092–1101.

3. Wijnands J. M. A., Kingwell E., Zhu F., et al., “Health‐Care Use Before a First Demyelinating Event Suggestive of a Multiple Sclerosis Prodrome: A Matched Cohort Study,” Lancet Neurology 16, no. 6 (2017): 445–451.

4. Tremlett H., Zhu F., Everett K., et al., “Healthcare Use Is Elevated Two Decades Before a First Demyelinating Event and Differs by Age and Sex,” Annals of Clinical Translational Neurology 12, no. 2 (2025): 415–432.

5. Sinay V., Perez Akly M., Zanga G., Ciardi C., and Racosta J. M., “School Performance as a Marker of Cognitive Decline Prior to Diagnosis of Multiple Sclerosis,” Multiple Sclerosis 21, no. 7 (2015): 945–952.

6. Cortese M., Riise T., Bjornevik K., et al., “Preclinical Disease Activity in Multiple Sclerosis: A Prospective Study of Cognitive Performance Prior to First Symptom,” Annals of Neurology 80, no. 4 (2016): 616–624.

7. Hoang H., Laursen B., Stenager E. N., and Stenager E., “Psychiatric Co‐Morbidity in Multiple Sclerosis: The Risk of Depression and Anxiety Before and After MS Diagnosis,” Multiple Sclerosis 22, no. 3 (2016): 347–353.

8. Marrie R. A., Walld R., Bolton J. M., et al., “Rising Incidence of Psychiatric Disorders Before Diagnosis of Immune‐Mediated Inflammatory Disease,” Epidemiology and Psychiatric Sciences 28, no. 3 (2019): 333–342.

9. Yusuf F., Wijnands J. M., Kingwell E., et al., “Fatigue, Sleep Disorders, Anaemia and Pain in the Multiple Sclerosis Prodrome,” Multiple Sclerosis 27, no. 2 (2021): 290–302.

10. Chertcoff A. S., Yusuf F. L. A., Zhu F., et al., “Psychiatric Comorbidity During the Prodromal Period in Patients With Multiple Sclerosis,” Neurology 101, no. 20 (2023): e2026–e2034.

11. Eid K., Torkildsen O., Aarseth J., et al., “Migraine in the Multiple Sclerosis Prodrome: A Prospective Nationwide Cohort Study in Pregnant Women,” Journal of Headache and Pain 25, no. 1 (2024): 225.

12. Yusuf F. L. A., Zhu F., Evans C., et al., “Gastrointestinal Conditions in the Multiple Sclerosis Prodrome,” Annals of Clinical Translational Neurology 11, no. 1 (2024): 185–193.

13. Bergeron‐Vitez H., Yusuf F., Zhu F., et al., “Anticonvulsant and Gabapentinoids Pharmacotherapy in the Multiple Sclerosis Prodrome: A Population‐Based Matched Cohort Study,” Multiple Sclerosis and Related Disorders 93 (2025): 106167.

14. Bjornevik K., Munger K. L., Cortese M., et al., “Serum Neurofilament Light Chain Levels in Patients With Presymptomatic Multiple Sclerosis,” JAMA Neurology 77, no. 1 (2020): 58–64.

15. Britze J., Larsen M. H., Pedersen A. G., et al., “Temporal Dynamics of Plasma Neurofilament Light in Blood Donors With Preclinical Multiple Sclerosis,” Neurology Neuroimmunology & Neuroinflammation 12, no. 1 (2025): e200335.

16. Harding K. E., Kreft K. L., Ben‐Shlomo Y., and Robertson N. P., “Prodromal Multiple Sclerosis: Considerations and Future Utility,” Journal of Neurology 271, no. 4 (2024): 2129–2140.

17. Marrie R. A., Allegretta M., Barcellos L. F., et al., “From the Prodromal Stage of Multiple Sclerosis to Disease Prevention,” Nature Reviews. Neurology 18, no. 9 (2022): 559–572.

18. Hillert J. and Stawiarz L., “The Swedish MS Registry—Clinical Support Tool and Scientific Resource,” Acta Neurologica Scandinavica 132, no. 199 (2015): 11–19.

19. Alping P., Piehl F., Langer‐Gould A., Frisell T., and Group C‐MS , “Validation of the Swedish Multiple Sclerosis Register: Further Improving a Resource for Pharmacoepidemiologic Evaluations,” Epidemiology 30, no. 2 (2019): 230–233.

20. Ludvigsson J. F., Andersson E., Ekbom A., et al., “External Review and Validation of the Swedish National Inpatient Register,” BMC Public Health 11 (2011): 450.

21. Ludvigsson J. F., Almqvist C., Bonamy A. K., et al., “Registers of the Swedish Total Population and Their Use in Medical Research,” European Journal of Epidemiology 31, no. 2 (2016): 125–136.

22. Ludvigsson J. F., Svedberg P., Olen O., Bruze G., and Neovius M., “The Longitudinal Integrated Database for Health Insurance and Labour Market Studies (LISA) and Its Use in Medical Research,” European Journal of Epidemiology 34, no. 4 (2019): 423–437.

23. Teljas C., Bostrom I., Marrie R. A., et al., “Validating the Diagnosis of Multiple Sclerosis Using Swedish Administrative Data in Varmland County,” Acta Neurologica Scandinavica 144, no. 6 (2021): 680–686.

24. McKay K. A., Manouchehrinia A., Zhu F., et al., “Higher Healthcare Use Before Paediatric Multiple Sclerosis Onset: A Nationwide Cohort Study,” Brain Communications 7, no. 3 (2025): fcaf180.

25. Manouchehrinia A., Zhu F., Hillert J., et al., “Prodromal Phase of Multiple Sclerosis: Evidence From Sickness Absence Patterns Before Disease Onset—A Matched Cohort Study,” Journal of Neurology, Neurosurgery & Psychiatry 96 (2025): 960–965.

26. Hogg T., Wijnands J. M. A., Kingwell E., et al., “Mining Healthcare Data for Markers of the Multiple Sclerosis Prodrome,” Multiple Sclerosis and Related Disorders 25 (2018): 232–240.

27. Marrie R. A., Patten S. B., Tremlett H., et al., “Sex Differences in Comorbidity at Diagnosis of Multiple Sclerosis: A Population‐Based Study,” Neurology 86, no. 14 (2016): 1279–1286.

28. Berger J. R., Pocoski J., Preblick R., and Boklage S., “Fatigue Heralding Multiple Sclerosis,” Multiple Sclerosis 19, no. 11 (2013): 1526–1532.

29. Pakpoor J., Seminog O. O., Ramagopalan S. V., and Goldacre M. J., “Clinical Associations Between Gout and Multiple Sclerosis, Parkinson's Disease and Motor Neuron Disease: Record‐Linkage Studies,” BMC Neurology 15 (2015): 16.

30. Simonsen C. S., Flemmen H. O., Broch L., et al., “Reduced Childbirth Rates in Multiple Sclerosis From the Prodromal Phase: Evidence From a Population‐Based Cohort Study,” Multiple Sclerosis (Houndmills) 31 (2025): 398–407.

31. Yusuf F. L. A., Wijnands J. M. A., Karim M. E., et al., “Sex and Age Differences in the Multiple Sclerosis Prodrome,” Frontiers in Neurology 13 (2022): 1017492.

Associated Data

Supplementary Materials

Appendix S1: ene70464‐sup‐0001‐AppendixS1.pdf.

Data Availability Statement

The data that support the findings of this study are provided by the Swedish MS Registry, National Board of Health, and Statistics Sweden. Data‐sharing agreements prohibit public availability, although the relevant data steward may grant access.