39 — Iatrogenic Psychiatry: When Treatment Produces New Symptoms

Abstract

Treatments intended to improve psychiatric illness can themselves cause new mental or behavioral symptoms. Medication-induced agitation, insomnia, withdrawal, delirium or cognitive effects can be misread as disease progression. The paper's thesis is that iatrogenic causation must remain visible in any biological psychiatry framework, because chronology and intervention exposure may explain a syndrome more directly than a newly proposed underlying metabolic disorder.

The treatment timeline is a causal instrument

A medication started days before new agitation provides a different clue from a long-standing, unchanged treatment. Dose changes, combinations, renal or hepatic function and abrupt discontinuation can all alter exposure and effect. Yet temporal association does not always establish causation: patients often receive treatment because their condition is already worsening. The correct analysis asks what changed first, which adverse-effect patterns are documented, and what alternative processes were active at the same time. Unsupervised abrupt medication withdrawal can be dangerous and should not be used as a casual experiment.

An oxalate-related attribution must survive the same scrutiny. If symptoms begin after a psychoactive drug change and no oxalate disturbance is measured, a metabolic narrative has little explanatory advantage. Conversely, serious renal disease may alter drug handling and create an interaction between the metabolic and iatrogenic pathways. The interesting mechanism could be combined causation rather than a simple one-agent story.

The akathisia example

Akathisia involves an often intensely distressing internal restlessness with observable movement or difficulty remaining still. It is associated with antipsychotic medications and has also been reported with other psychotropic classes. A patient may describe panic, agitation or unbearable internal discomfort rather than use the term akathisia. Increasing medication under the assumption that agitation represents worsening psychosis could worsen an adverse drug effect in some situations. Standardized evaluation and recognition of timing are important.

Activation, sedation and withdrawal

Medication changes can alter sleep, autonomic state and cognitive performance. Activation after initiation or dose escalation requires distinction from mania, anxiety and ordinary treatment adjustment. Sedation can be mistaken for depressive slowing; emotional blunting for persistent anhedonia. Abrupt withdrawal of some medications can produce distressing physiological and psychiatric symptoms. The relationship between onset and prescription changes is informative but not sufficient by itself: worsening illness can happen at the same time.

A chronology-centered method

A useful medication timeline records agent, dose, start, titration, discontinuation, concurrent drugs and the first appearance of each symptom. The clinician asks whether symptoms were present before exposure, whether they are recognized adverse-effect patterns and whether an alternative medical explanation is plausible. A clinically supervised medication adjustment may clarify cause, but unsafe abrupt discontinuation or unmonitored changes are not appropriate experiments.

Why the idea matters for medical theory

The phrase “treatment resistance” can conceal an adverse-effect phenotype, while the phrase “drug toxicity” can conceal illness progression. Both are forms of premature causal closure. A serious systems model examines interventions as new biological exposures with intended and unintended effects, and accepts that the safest action may sometimes be simplification rather than intensification.

The treatment can become part of the presenting syndrome

A patient may develop insomnia, agitation, cognitive slowing or perceptual disturbance during treatment for an unrelated condition. The causal possibilities include a direct drug adverse effect, a withdrawal phenomenon, an interaction, a dose-related exposure change or progression of the original illness. Corticosteroids, some stimulants, certain neurological medications and many other drug classes can produce clinically important neuropsychiatric changes in susceptible people. Recognizing this possibility does not mean that prescribed psychiatric medicines are generally harmful or unnecessary. It means that a careful chronology of exposures is part of diagnosis.

A common error is to interpret every new symptom as evidence that the underlying psychiatric disorder has worsened. If the symptom appeared after a dose increase, withdrawal or addition of an interacting agent, medication effects should be considered. However, temporal sequence alone is not enough. Illness severity can prompt medication changes immediately before natural deterioration, creating confounding by indication. A structured assessment asks whether the adverse effect is biologically plausible, consistent with known pharmacology and reproducible, while maintaining appropriate safety during changes in treatment.

The pharmacological terrain

Drug exposure depends on absorption, distribution, hepatic metabolism and renal clearance. Systemic inflammation or organ dysfunction can alter these pathways, and sleep or nutritional disturbances may change tolerability. These observations help explain why an intervention that was previously tolerated may become problematic during acute medical illness. A patient with kidney impairment may accumulate certain drugs or metabolites, while a new interacting prescription changes enzyme activity. The result may look psychiatric even when the immediate cause is pharmacological.

In a patient with oxalate-associated renal disease, impaired filtration might change medication handling, providing a plausible indirect route to altered behavior. This would not mean oxalate directly affected neurotransmission. It would mean that oxalate-related illness contributed to a chain of consequences involving renal function and drug exposure. Such a chain is medically important precisely because it can be identified and managed through ordinary clinical pharmacology rather than unsupported detoxification claims.

Interventions themselves create inference problems

Suppose a patient improves when a medication is discontinued under appropriate supervision. That improvement supports a role for the medication but may not distinguish adverse effects from withdrawal resolution, normal symptom variation or simultaneous changes in care. Abrupt discontinuation of some psychiatric drugs can itself cause serious symptoms, so the investigation of adverse effects cannot be conducted through unsafe self-experimentation. The clinician should assess indication, risk, tapering considerations where appropriate and alternative therapy.

For research, medication timelines should be analyzed alongside symptoms, kidney and liver function, interacting exposures and objective functional outcomes. A new psychiatric presentation in someone with significant systemic illness deserves a broad assessment including iatrogenic possibilities. The central terrain proposition is that treatment acts within a changing organism. Recognizing that fact strengthens pharmacology and patient safety; it should not become a blanket argument against evidence-based medical treatment.

A timeline is often more informative than a broad symptom label

Consider a patient who becomes agitated and sleepless several days after starting systemic corticosteroids. A medication effect is biologically plausible and should enter the differential, although the medical illness requiring treatment may also influence mood and sleep. Another patient develops sensory symptoms and anxiety after abruptly discontinuing a medication known to produce withdrawal effects. The mechanisms are different, and their management should not be conflated. A rigorous assessment records start dates, dose changes, discontinuation, co-prescriptions, organ function and relevant clinical outcomes.

The hazard of unstructured self-experimentation

Patients frustrated by adverse effects may stop several agents at once or make radical diet changes simultaneously. If symptoms then improve or worsen, the cause becomes difficult to identify; some abrupt medication changes can be dangerous. Appropriate clinical supervision is therefore part of good causal inquiry, not a barrier to independent thinking. A researcher examining iatrogenic syndromes should account for withdrawal, rebound symptoms and the original reason the medication was prescribed. Otherwise a treatment-related problem can be mistaken for proof that the underlying medical diagnosis never existed.

Why a terrain hypothesis still needs pharmacology

The organism's internal state matters to drug distribution and response. Kidney or liver dysfunction, inflammation and nutrition can modify treatment exposure. But knowing that a state variable matters in principle is not enough to recommend altering a prescription. The actual pathway must be defined for the drug in question. This principle applies directly to claims about oxalate: confirmed oxalate-related renal failure may alter relevant drug handling, but high-oxalate foods in a healthy patient do not thereby constitute a demonstrated cause of medication-induced psychosis.

The same reasoning applies to adverse effects attributed to supplements. A person may be using multiple products marketed for sleep, detoxification or neurotransmitter support, sometimes with poorly characterized ingredients or doses. A new symptom should prompt an accurate inventory of these exposures alongside prescription medicines. However, stopping a product and feeling better is not definitive evidence that its advertised active ingredient was the cause; contaminants, interactions and concurrent behavior changes may matter. Causal confidence rises when timing, known pharmacology and a consistent exposure–response relation align. Patient safety takes precedence over experimental re-challenge when the suspected effect is serious.

Investigators should also avoid assuming that a medicine is the sole cause whenever symptoms follow treatment. The underlying disease may be worsening, or the drug may have been started precisely because the patient was already deteriorating. This is confounding by indication. A rigorous assessment weighs the expected adverse-effect profile, timing, alternative diagnoses and results of medically appropriate changes. Stronger causal inference requires more than an appealing sequence of events.

Conclusion

Iatrogenic psychiatry challenges the assumption that all new symptoms reveal more underlying disease. An exposure–time–effect analysis should precede speculative causal reconstruction.

Selected readings

Pringsheim et al., The Assessment and Treatment of Antipsychotic-Induced Akathisia: https://pmc.ncbi.nlm.nih.gov/articles/PMC6299189/

Bousman et al., Antidepressant metabolism and tolerability: https://pmc.ncbi.nlm.nih.gov/articles/PMC10564324/