Prediagnostic prescription patterns in multiple sclerosis: a UK-based longitudinal linked data study

Baoyue Zhang; Oleg Blyuss; Kirsten D Arendse; Luke TA Mounce; Ruth Dobson; Garth Funston

doi:10.3399/BJGP.2025.0210

Published eLetters

eLetter submission guidelines

Submit a Response to This Article

Compose eLetter

Title *

Author Information

Contributors

First Name and Middle Initial *

First or given name, e.g. 'Peter'.

Last Name *

Your last, or family, name, e.g. 'MacMoody'.

Email Address *

Your email address, e.g. higgs-boson{at}gmail.com. PLEASE NOTE: your email address will be published.

Occupation *

Your role and/or occupation, e.g. 'Orthopedic Surgeon'.

Affiliation *

Your organization or institution (if applicable), e.g. 'Royal Free Hospital'.

Statement of Competing Interests

Competing interests? *

Yes

Please describe the competing interests

CAPTCHA

This question is for testing whether or not you are a human visitor and to prevent automated spam submissions.

Vertical Tabs

Jump to comment:

Prediagnostic prescriptions in multiple sclerosis: signals, behaviour, and clinical reality
Jinxiang Chen, Beibei Chen and Shasha Chen

Published on: 27 April 2026

Published on: (27 April 2026)
Page navigation anchor for Prediagnostic prescriptions in multiple sclerosis: signals, behaviour, and clinical reality
Prediagnostic prescriptions in multiple sclerosis: signals, behaviour, and clinical reality
Jinxiang Chen, Doctor, Zhejiang Chinese Medical University
Other Contributors:
Beibei Chen, Doctor, Zhejiang Chinese Medical University

Shasha Chen, Doctor, Zhejiang Chinese Medical University
We read with great interest the recent study by Zhang and colleagues, which used UK primary care data to elegantly demonstrate altered prescribing patterns up to 10 years before multiple sclerosis (MS) diagnosis.¹ This well-designed case–control study provides important evidence for extended prodromal windows in MS, highlighting gastrointestinal, urological, migraine, mood, and sexual dysfunction prescriptions as potential early markers. We offer three complementary perspectives to strengthen the interpretation and guide future research.

First, regarding the methodology for change-point detection, routine primary care electronic health record (EHR) data are inherently noisy, non-stationary, and autocorrelated. The authors’ use of Poisson regression with fixed 3-month intervals helps identify inflection points but remains vulnerable to random fluctuations and seasonal trends. Advanced time-series approaches—such as Bayesian structural time series (BSTS) or hidden Markov models (HMM)—could better isolate genuine prodromal signals from background structural noise, yielding more robust and dynamic estimates of prodromal onset.²

Second, consultation rate bias and psychosomatic pathways require careful adjustment. Patients in the MS prodrome often experience unexplained somatic symptoms, leading to more frequent GP contacts. Higher consultation frequencies artificially inflate the probability of receiving sympto...
Show More

We read with great interest the recent study by Zhang and colleagues, which used UK primary care data to elegantly demonstrate altered prescribing patterns up to 10 years before multiple sclerosis (MS) diagnosis.¹ This well-designed case–control study provides important evidence for extended prodromal windows in MS, highlighting gastrointestinal, urological, migraine, mood, and sexual dysfunction prescriptions as potential early markers. We offer three complementary perspectives to strengthen the interpretation and guide future research.

First, regarding the methodology for change-point detection, routine primary care electronic health record (EHR) data are inherently noisy, non-stationary, and autocorrelated. The authors’ use of Poisson regression with fixed 3-month intervals helps identify inflection points but remains vulnerable to random fluctuations and seasonal trends. Advanced time-series approaches—such as Bayesian structural time series (BSTS) or hidden Markov models (HMM)—could better isolate genuine prodromal signals from background structural noise, yielding more robust and dynamic estimates of prodromal onset.²

Second, consultation rate bias and psychosomatic pathways require careful adjustment. Patients in the MS prodrome often experience unexplained somatic symptoms, leading to more frequent GP contacts. Higher consultation frequencies artificially inflate the probability of receiving symptomatic prescriptions, creating a confounded association.³ Furthermore, the surge in anxiety and depression observed close to diagnosis may represent secondary psychosomatic responses to unexplained symptoms and diagnostic uncertainty, rather than primary demyelination. Future models should include annual consultation frequency as a time-varying confounder to disentangle these effects.

Third, clinical translation faces critical base-rate challenges. Urinary tract infections, migraines, gastrointestinal symptoms, and anxiety are ubiquitous in primary care, whereas MS is rare. Even with strongly elevated odds ratios, the positive predictive value (PPV) of these prescribing patterns alone will remain exceptionally low.⁴ Widespread screening based on these non-specific features risks overdiagnosis, unnecessary neuroimaging, and iatrogenic anxiety. Consequently, these prodromal markers may be better suited for risk stratification within known high-risk cohorts (e.g., those with a family history) rather than as standalone screening tools for the general population.

We commend the authors for advancing our understanding of the MS prodrome. These methodological considerations may help refine early detection strategies and support safe, effective translation into primary care practice.

References
1. Zhang B, Blyuss O, Arendse KD, Mounce LT, Dobson R, Funston G. Prediagnostic prescription patterns in multiple sclerosis: a UK-based longitudinal linked data study. Br J Gen Pract 2026; doi:10.3399/BJGP.2025.0210.
2. Gianacas C, Liu B, Kirk M, et al. Bayesian structural time series, an alternative to interrupted time series in the right circumstances. J Clin Epidemiol 163:102-110. doi:10.1016/j.jclinepi.2023.10.003.
3. Agniel D, Kohane IS, Weber GM. Biases in electronic health record data due to processes within the healthcare system: retrospective observational study. BMJ 2018;361:k1479. doi:10.1136/bmj.k1479.
4. Grimes DA, Schulz KF. Uses and abuses of screening tests. Lancet 2002; 359(9309):881-4. doi:10.1016/S0140-6736(02)07948-5.

Show Less

Competing Interests: None declared.

Main menu

User menu

Search

Prediagnostic prescription patterns in multiple sclerosis: a UK-based longitudinal linked data study

Published eLetters

Plain text

Vertical Tabs

Jump to comment:

In this issue

Citation Manager Formats

Keywords

More in this TOC Section

Related Articles

Cited By...

NAVIGATE

RCGP

MY ACCOUNT

NEWS AND UPDATES

AUTHORS & REVIEWERS

CUSTOMER SERVICES

CONTRIBUTE

CONTACT US

Main menu

User menu

Search

Prediagnostic prescription patterns in multiple sclerosis: a UK-based longitudinal linked data study

Published eLetters

Plain text

Vertical Tabs

Jump to comment:

In this issue

Citation Manager Formats

Jump to section

Keywords

More in this TOC Section

Related Articles

Cited By...

NAVIGATE

RCGP

MY ACCOUNT

NEWS AND UPDATES

AUTHORS & REVIEWERS

CUSTOMER SERVICES

CONTRIBUTE

CONTACT US