Differential Patient Advocacy Diagnostic Overshadowing Clinical Reasoning Diagnostic History Research Studies

Medication, Polypharmacy and FND: What the Research Reveals and What Remains Unanswered

Author profile image FND Nope Moderator
0 comments

What happens when a patient experiencing unexplained neurological symptoms is prescribed multiple medications, some of which can themselves produce neurological adverse effects? How thoroughly are those medications investigated before symptoms are attributed to functional neurological disorder (FND)? And when patients accumulate prescriptions over years of medical investigation, how confidently can clinicians distinguish the original illness from the potential consequences of its treatment?

These are not hypothetical concerns invented by critics of FND. In 2019, O'Connell and colleagues published a study in General Hospital Psychiatry examining medication prescribing in 322 patients diagnosed with motor FND at a large UK mental health service. The researchers compared their prescribing patterns with those of 644 psychiatric controls and identified substantial medication exposure, frequent polypharmacy and concerns about potential iatrogenic harm. The findings do not establish that medications caused the patients' neurological symptoms, but they raise questions that deserve considerably more attention than a simple discussion of diagnostic labels. [1]


What the Study Actually Found

The researchers conducted a retrospective case-control study using clinical records covering 2006 to 2016. Their analysis examined the number and categories of medications prescribed to patients with motor FND, alongside prescribing patterns in a psychiatric comparison group. This was not a trial of medication safety or an investigation designed to determine whether particular drugs caused functional neurological symptoms. It was an examination of prescribing practices in a defined clinical population. [1]

The findings were substantial:

  • 68% of medicated motor FND patients were prescribed antidepressants.
  • 34% were prescribed antiepileptic medications.
  • 31.2% were prescribed opioid analgesics.
  • 43.7% of medicated motor FND patients received five or more medications.
  • The mean number of medications was 4.77 in the motor FND group, compared with 2.98 among psychiatric controls.

These figures describe prescribing patterns rather than inappropriate treatment. Some medications may have been prescribed for independently established psychiatric disorders, pain, epilepsy or other medical conditions. The patients were also drawn from a specialist mental health service, meaning the results cannot automatically be generalised to everyone diagnosed with FND. Nevertheless, the differences in medication burden are important, particularly when considering the complexity of symptoms and the potential consequences of prescribing several centrally acting medications simultaneously. [1]

The researchers also identified associations between higher medication counts and factors including physical comorbidity and previous psychiatric admissions. Such findings illustrate how difficult it can be to disentangle underlying illness, accompanying conditions and treatment exposure in patients with complicated clinical histories. [1]


Polypharmacy Is More Than a Prescription Count

Polypharmacy is commonly understood as the use of multiple medications, often operationally defined in research as five or more. It is not inherently inappropriate. A patient with several established medical conditions may require several treatments, each with a defensible indication. The concern arises when medication burden increases without sufficiently clear benefit, when interactions or adverse effects are overlooked, or when prescriptions continue without regular reassessment. Medication safety guidance recognises polypharmacy as an area requiring careful review because avoidable medication-related harm can occur even when individual prescribing decisions initially appear reasonable. [2][3]

Several medication classes identified in the study can produce effects relevant to neurological assessment. Depending on the specific drug and clinical circumstances, these may include dizziness, sedation, tremor, impaired coordination, movement abnormalities, cognitive changes or withdrawal-related symptoms. These are recognised pharmacological possibilities, not evidence that such reactions occurred in the patients examined by O'Connell and colleagues. Establishing that would require individual exposure histories and clinical assessment.

The diagnostic difficulty is obvious. A patient presenting with weakness, abnormal movements, dizziness or other neurological complaints may already have been exposed to several medications capable of affecting nervous system function. If additional treatments are introduced as symptoms persist, the clinical picture can become increasingly difficult to interpret. The original disorder, coexisting illnesses, treatment effects and medication withdrawal may all require consideration.

When neurological symptoms coexist with extensive medication exposure, reviewing the prescriptions is not an optional courtesy. It is part of investigating what might be contributing to the patient's condition.

One particularly important finding of the O'Connell paper was the authors' acknowledgement of limited evidence supporting pharmacological treatment specifically for motor FND, alongside their concern about possible iatrogenic harm. That does not mean every prescription was unnecessary. It means the burden of medication deserves scrutiny, especially where treatments have accumulated without a clear understanding of their overall effects. [1]


What the Research Cannot Tell Us

The study has important limitations, and acknowledging them does not diminish its findings. Its retrospective design cannot establish causation. The researchers could not determine from prescribing patterns alone whether medication exposure preceded the onset of neurological symptoms, whether adverse effects were mistaken for functional symptoms or whether medication changes altered patients' clinical presentations. Nor could the study establish that a particular prescription was inappropriate without examining the full clinical indication and circumstances. [1]

There is also a distinction between prescribed medication and medication actually taken. Prescription records do not necessarily establish adherence, dosage exposure or the duration of use. Those details matter when investigating adverse drug reactions and potential treatment-related harm.

These limitations define the questions that further research needs to answer. How often did symptoms begin after medication initiation or dose escalation? Were recognised neurological adverse effects systematically investigated? Did withdrawal or dose reduction coincide with changes in symptoms? Were medication-related explanations considered before clinicians settled on a functional diagnosis? And how often did a structured medication review materially change a patient's treatment or diagnostic assessment?

The O'Connell study does not answer those questions. But neither does its inability to answer them make the questions irrelevant. The difference between an unanswered clinical question and a disproven concern is fundamental.


Does Scientific Evidence Have an Expiry Date?

One FND proponent has reportedly questioned the relevance of this study because it is seven years old. That is an odd objection in isolation, particularly in a field where some clinical examination signs have histories stretching back more than a century. Charles Franklin Hoover described his sign in 1908, building on an earlier tradition of examining discrepancies between voluntary and associated movements. Joseph Babinski and other neurologists had already investigated related phenomena, and contemporary reviews continue to examine the diagnostic performance of several of these historically rooted techniques. [4][5][6]

The comparison is not an argument that Hoover's sign is invalid because it is old, or that the O'Connell study must be accepted because it is newer. Both require scrutiny. Modern research can validate historical clinical observations, refine their interpretation or expose limitations. Likewise, subsequent prescribing research can confirm, qualify or supersede findings from 2019. The relevant question is what the accumulated evidence demonstrates, not whether an arbitrary number of years has passed.

A 2025 review of techniques used to assess functional motor weakness illustrates this principle. It examined the validation evidence for several examination signs, including Hoover's sign, and identified limitations in the available studies and the consistency of examination methods. The appropriate scientific response is to examine those limitations and pursue better validation, not simply accept or reject the signs because of their historical origins. [6]

The prescribing study deserves the same treatment. Its clinical records are now considerably older than the publication itself, and prescribing practices may have changed since the period examined. That limits what can be concluded about current practice. It does not erase the prescribing patterns observed in the study or resolve the concerns its authors raised. Those concerns should be assessed against subsequent evidence, not dismissed by reference to a calendar.


The Missing Research Matters

Perhaps the most important issue is not the medication burden itself, but how little can be concluded about its consequences from prescription counts alone. The study documented extensive medication exposure among motor FND patients. It did not establish how many experienced medication-related neurological adverse effects, whether those effects influenced their diagnosis or whether potentially avoidable prescribing contributed to their symptoms. Those are distinct questions requiring different research methods. [1]

A useful next step would be prospective studies following patients from their initial neurological presentation, recording medication exposure, dose changes, adverse effects, withdrawal, diagnostic findings and clinical outcomes over time. Research could also examine whether structured medication reviews identify previously unrecognised adverse effects or interactions, and whether such reviews alter clinical management. These investigations would be particularly valuable in patients whose symptoms developed or changed around medication exposure.

Another important question concerns prescribing indications. A medication prescribed for confirmed epilepsy, for example, cannot be evaluated in the same way as one prescribed empirically for an unexplained symptom. Similarly, antidepressants may be prescribed for several different conditions, and their presence in a patient's record does not establish that clinicians regarded the neurological symptoms as psychiatric. Understanding why medications were prescribed is essential to interpreting whether treatment was appropriate.

There is also the question of diagnostic closure. A positive functional examination sign may contribute to a diagnosis, but it does not automatically rule out medication-related complications or coexisting neurological disease. A careful assessment should consider relevant competing and contributing explanations rather than treating a diagnostic label as a substitute for reviewing the patient's clinical history. This is especially important when medication exposure and symptom chronology suggest a plausible relationship that has not yet been investigated.

The issue is not whether medication-related explanations should automatically take precedence over FND. They should not. The issue is whether they are being evaluated with sufficient care, and whether patients receive the same attention to potential treatment-related harm that would be expected in any other area of medicine.


What Should Happen Next?

The O'Connell study provides evidence of substantial prescribing and polypharmacy in a particular motor FND population. Its authors raised concerns about the limited evidence for pharmacological treatment and the possibility of iatrogenic harm. The findings warrant follow-up research into prescribing indications, adverse effects, medication interactions and the relationship between treatment exposure and neurological symptoms. They do not justify claiming that medications cause all FND presentations, but they certainly do not justify ignoring medication-related possibilities. [1]

Ultimately, the most important question is what happens to patients when those possibilities are not investigated. If medication-related symptoms are overlooked, patients may continue receiving treatments that contribute to their difficulties. If necessary medications are incorrectly blamed, patients may also be harmed. Both outcomes illustrate why individualised assessment, careful prescribing and properly supported diagnostic reasoning matter.

What the researchers documented was a significant medication burden. What remains to be established is how often that burden contributed to symptoms, complicated diagnosis or caused preventable harm. Until those questions are answered, the appropriate response is further investigation, not certainty in either direction.

A diagnosis should help clinicians investigate and treat a patient's condition. It should never become a reason to stop asking whether the treatment itself might be part of the problem.


References