Pediatric Psychosis and Catatonia: Exploring Infection-Associated Neuropsychiatric Symptoms

A longitudinal case of severe childhood psychosis, catatonia, and functional regression explores possible post-infectious mechanisms, recurring responses to antibiotic treatment, and the role of integrative care when diagnostic uncertainty persists.

Background: For naturopathic physicians, complex pediatric neuropsychiatric illness calls for whole-person assessment and attention to patterns when diagnostic testing is incomplete or difficult to access. Longitudinal observations can generate hypotheses while uncertainty remains explicit.

Case Presentation: A previously high-functioning girl developed anxiety, paranoia, hallucinations, disorganized language, catatonic features, and marked functional regression beginning in late 2020. She was hospitalized in February–March 2021 while SARS-CoV-2 positive. Symptoms fluctuated for more than two years and included reduced spontaneous speech, pacing, pica, incontinence, sleep disruption, altered eating, and severe agitation.

Intervention: Care included inpatient psychiatric treatment, lorazepam, antipsychotic trials, medical and neuroinflammatory evaluation, nutritional support, iron supplementation, complementary and alternative therapies, and several antibiotic courses prescribed amid concern for an infection-associated syndrome.

Outcomes: Substantial recovery emerged during 2023, including return to school, spontaneous speech and writing, independent daily activities, and resolution of pica and restrictive eating. By June 2025, no recurrence of psychosis or catatonia was reported; escitalopram had been discontinued and trazodone was used occasionally for sleep.

Conclusion: The course raises the possibility of an infection-associated or post-infectious contribution alongside other vulnerabilities. Repeated temporal improvement during antibiotic courses is a clinically relevant observation. In selected cases, empiric treatment may be considered when the plausible benefit outweighs foreseeable risks, with monitoring and reassessment; this case cannot establish treatment efficacy, but does provide observations and questions for future research.

Keywords: pediatric psychosis; catatonia; infection-associated neuropsychiatric symptoms; PANS/PANDAS; neuroinflammation; integrative medicine

“Missing biomarkers or incomplete testing do not make clinical observation meaningless.”

Introduction

New-onset psychosis with catatonia in a child calls for broad assessment of psychiatric, neurologic, infectious, immune, metabolic, nutritional, medication-related, and psychosocial contributors. Catatonia may accompany several conditions; investigation and symptom relief may need to proceed together. Pediatric autoimmune encephalitis can include psychiatric or cognitive changes, movement abnormalities, seizures, sleep disruption, or regression, but diagnosis depends on the clinical picture and available investigations.1-3,6,7

PANS/PANDAS and other infection-associated neuropsychiatric processes have been proposed for selected acute-onset presentations. Mechanisms and treatment remain debated, and ASO alone cannot establish active infection or causality.1,2 Still, missing biomarkers or incomplete testing do not make clinical observation meaningless. This case uses a naturopathic, pattern-based lens: symptom timing, caregiver observations, treatment responses, tolerability, sleep, nutrition, environment, and functional recovery are considered together when formulating a patient centered, collaborative treatment plan.

Case Presentation

Patient Information and Presenting Concern

The patient was approximately 10 years old at initial hospitalization. History included early separation anxiety, trauma associated with witnessing domestic violence, bilateral clubfoot correction, and prior restrictive eating/anorexia and pica. She had been described as high-functioning before illness. Progressive symptoms began in fall 2020 with anxiety and paranoia, progressing to auditory hallucinations, impaired communication, disorganized behavior, and catatonic features. She was hospitalized in February–March 2021 while SARS-CoV-2 positive. Symptoms persisted and fluctuated after discharge, impairing communication, eating, toileting, sleep, social engagement, and school participation.

Clinical Course and Treatment Responses

During hospitalization, quetiapine was stopped and lorazepam was used for catatonia, with documented improvement in motor and mental-status findings. Olanzapine was initially associated with increased agitation and disorganization and was stopped; it was later reintroduced and tolerated with gradual, incomplete improvement in psychosis. The caregiver stopped medications about two weeks after discharge. A later lorazepam retrial initially appeared to improve communication but was subsequently associated by the caregiver with agitation, facial or mouth movements, pacing, pica, and reduced appetite, prompting dose reduction and taper.

During 2021–2022, symptoms included hallucinations, reduced spontaneous speech, echolalia or repetitive language, pacing, incontinence, pica, disturbed sleep, reduced intake, and weight loss followed by weight gain. Symptoms worsened again in September–December 2022 after a year abroad, amid housing instability, school disruption, and caregiver burden. A household viral illness coincided with acute worsening in December 2022.

The caregiver reported improvement during or after several antibiotic courses, including amoxicillin and cefdinir, with recurrence after some courses ended. This repeated sequence raises the possibility that infection, microbial ecology, or immune signaling played a role and supports further inquiry. Other explanations include natural fluctuation and concurrent changes. The pattern does not prove PANDAS or antibiotic efficacy, but neither should it be dismissed simply because testing was incomplete or controlled evidence is limited.

During the most impaired period, observations included self-talk or jargon, limited response to name, reduced eye contact, restricted affect, pacing, impaired spontaneous speech, urinary and stool incontinence, vomiting/regurgitation, and loss of daily functioning. By June 2025, psychiatric follow-up described normal-rate speech, organized thought, appropriate affect, no reported perceptual abnormalities, and no observed response to internal stimuli.

Diagnostic Assessment

MRI in February 2021 showed nonspecific findings of limited significance. EEG in March 2021 showed mild generalized slowing while the patient was taking benzodiazepines. CSF contained 133 red blood cells/µL, 2 nucleated cells/µL, protein 19 mg/dL, and glucose 49 mg/dL. The CSF autoimmune encephalitis panel was incomplete because of insufficient sample; tested CASPR2, GABA-B, GAD, LGI-1, and NMDA receptor antibodies were negative. Serum pediatric autoimmune testing was negative. Other reported negative tests included MOG, AQP4, thyroid antibodies, ceruloplasmin, lead, copper, zinc, HIV, syphilis, and tuberculosis. Further CSF testing was considered, but results for all proposed studies are unavailable.

ASO was elevated in February and September 2021 (231 and 211 IU/mL; reference 0–150), mildly elevated in October 2022 (155), and within range in January and September 2023 (135 and 120). Throat cultures were reported negative. ESR was 17 mm/hr in February 2021, 20 in April 2023, and 24 in September 2023 (reference 0–20). TSH was low once in October 2022 (0.57 mIU/L; reference 0.70–4.17), with decreased T3 also reported but no value available. September 2023 iron studies showed iron 30, TIBC 497, and saturation 6%, consistent with an iron-deficiency pattern. Vitamin B12 values were 916 and 1,076 pg/mL. WBC reached 3.1 in July 2021 and later recovered; hemoglobin was 11.1 g/dL in September 2023.

Clinical Assessment

The working formulation is severe pediatric psychosis with prominent catatonia and a possible infection-associated or post-infectious contribution; primary psychiatric vulnerability remains in the differential. PANS/PANDAS was considered because of abrupt change, prior restrictive eating, elevated early ASO, and reported improvement during antibiotic courses. The record does not establish acute streptococcal infection at onset; negative throat cultures and incomplete CSF evaluation limit certainty. The available information neither confirms a specific inflammatory diagnosis nor excludes every infection-associated process.

Interventions, Follow-up, and Safety

Interventions included lorazepam, quetiapine, olanzapine, and later escitalopram, trazodone, and clonidine; antibiotic courses of penicillin V, cefdinir, cephalexin, and amoxicillin; and complementary and alternative therapies including probiotics, multivitamin, zinc, magnesium, melatonin, low-dose lithium orotate, botanical anti-microbial drops, chamomile, fiber, and dietary strategies. Calorie-dense and iron-rich nutrition, weight monitoring, iron supplementation, school supports, and home-based educational services were also documented. The effects of individual interventions cannot be isolated due to concurrent use. Expanded CSF testing, repeat EEG if symptoms recurred, and possible IVIG after specialist assessment were considered; IVIG was not offered because the neuroinflammatory assessment remained incomplete and medical necessity could not be clearly established.

In 2023, the patient returned to school, regained spontaneous speech and writing after more than two years, became independent in daily activities, and no longer had pica or restrictive eating. In January 2024, weight was 138.6 lb and BMI 23.24 kg/m². By June 2025, she had completed the school year without academic or behavioral concerns and had no reported recurrence of psychosis or catatonia. Escitalopram had been discontinued by May 2025 without persistent worsening; trazodone was used occasionally for sleep.

Medication tolerability concerns included initial agitation/disorganization during olanzapine treatment, caregiver-reported worsening during a later higher-dose lorazepam trial, and gibberish speech, pacing, and sleep disruption when escitalopram was increased to 7 mg/day, reportedly resolving after reduction to 5 mg/day. These associations do not establish causation. Antibiotics and complementary therapies also require attention to allergies, adverse effects, interactions, product quality, and appropriate follow-up.

Discussion and Therapeutic Rationale

This case illustrates the value of clinical pattern recognition when diagnostic access or testing is incomplete. Onset, symptom clusters, possible triggers, repeated treatment-response patterns, relapse after treatment cessation, sleep, nutrition, medication effects, and functional trajectory can guide hypotheses and next steps. The repeated caregiver reports of improvement during antibiotics and recurrence after some courses ended are a hypothesis-generating signal, not proof of a mechanism.

From an integrative perspective, empiric antibiotics may sometimes be considered when infection-associated illness is plausible, definitive testing is unavailable, delayed treatment could carry meaningful cost, and anticipated benefit outweighs risks. This is individualized judgment, not a rationale for indiscriminate or indefinite prescribing. When feasible, obtain targeted testing first; clarify the goal, select an appropriate agent and duration, review allergies and interactions, monitor response and adverse effects, and reassess the hypothesis. One small pediatric study reported fewer infections and neuropsychiatric exacerbations during penicillin or azithromycin prophylaxis than in the preceding baseline year,9 establishing a possible treatment approach to consider in cases where there is a strong clinical history supported by evidence of prior strep infection preceding neuropsychiatric symptoms.

A post-infectious or neuroinflammatory contribution remains possible given the timing around a SARS-CoV-2-positive hospitalization and prominent neuropsychiatric and motor symptoms. Available data do not demonstrate autoimmune encephalitis: MRI findings were nonspecific, CSF lacked pleocytosis and had normal protein, serum autoimmune testing was negative, and the CSF panel was incomplete.6,7 An incomplete evaluation leaves questions unresolved, but uncertainty cuts both ways: it neither confirms nor rules out an immune-mediated process. The course cannot be attributed to one infection or mechanism. Also, due to the timing of this case with the COVID-19 pandemic, there was limited medical literature available to guide treatment in the context of SARS-CoV-2 infection preceding catatonia.

Escitalopram may have immunomodulatory effects according to a small adult study reporting decreased inflammatory cytokines after treatment.8 This potential effect of low dose escitalopram is an interesting hypothesis to consider in the context of this case although the research is not robust enough to draw any conclusions about a possible mechanism of action in the pediatric population. Drawing any conclusion about treatment efficacy of escitalopram from this case study is impossible, given that the treatment occurred alongside changes in sleep, improved nutrition, school participation, use of complementary therapies, changes in family circumstances, and support.

A naturopathic approach can integrate psychiatric and neurologic care with attention to infection history, sleep, nutrition, micronutrients, medication effects, environmental exposures, psychosocial stressors, and caregiver knowledge. Where testing is inaccessible, it is important to document the rationale and alternatives, use the least risky reasonable intervention, and set monitoring and stopping criteria. Coordination with the child’s medical and mental-health team is important, particularly with psychosis or catatonia. Limitations include retrospective data, caregiver reports, incomplete testing, and concurrent interventions.

Conclusion

This child experienced more than two years of fluctuating psychosis and catatonia followed by substantial recovery. The course supports an idiopathic onset that includes—but does not establish—an infection-associated contribution. Repeated improvement reported during antibiotic courses is meaningful enough to document and investigate. Empiric therapy may be reasonable when individualized potential benefit outweighs risks, provided uncertainty is acknowledged and response, harms, and alternatives are reassessed.

Author: NDNR-emmabelle-2026.