Assessment of long-term mental health effects of mefloquine use during pregnancy: a secondary analysis of a randomized clinical trial in Benin

Abstract

Background

Mefloquine (MQ) was evaluated as intermittent preventive treatment in pregnancy (IPTp) in a randomized clinical trial, as part of the ongoing search for alternatives to sulfadoxine–pyrimethamine (SP). Although MQ was not recommended for IPTp due to poor tolerability, no psychiatric differences were found between trial arms. Nevertheless, widespread concerns about MQ's neuropsychiatric effects persist and remain clinically relevant — particularly for pregnant travelers to malaria-endemic areas for whom MQ prophylaxis is still indicated. Long-term data on MQ's effects on maternal mental health and infant neurodevelopment are lacking.

Methods

Data come from the Benin subsample of the MiPPAD randomized, three-arm clinical trial (N = 1183), conducted across antenatal care clinics in three study sites (Allada, Sékou, and Attogon) between 2009 and 2013. Women received either SP, MQ full dose (15 mg/kg), or MQ split dose (same total dose over two days). At one year postpartum, 752 mother–child pairs provided information on depression and anxiety (Edinburgh Postnatal Depression Scale (EPDS)), maternal-child interactions (Home Observation for Measurement of the Environment (HOME)), and 745 infants had information on neuropsychological outcomes (Mullen Scales of Early Learning (MSEL)). Logistic and linear regressions estimated odds ratios (OR) and beta coefficients (β) with 95% Confidence Intervals (CI).

Findings

Mothers were approximately 26 years of age at inclusion. No significant associations were found between IPTp regimen and maternal depressive symptoms (score ≥9: MQ full dose OR 0.85 [95% CI 0.60, 1.20]; MQ split dose OR 0.94 [95% CI 0.66, 1.34]), severe depression (score ≥12: MQ full dose OR 0.98 [95% CI 0.63, 1.54]; MQ split dose OR 1.26 [95% CI 0.82, 1.95]), or anxiety (MQ full dose OR 0.91 [95% CI 0.59, 1.40]; MQ split dose OR 1.03 [95% CI 0.67, 1.59]), all compared with SP. No associations were observed for infant MSEL composite scores (MQ full dose β −1.21 [95% CI –3.61, 1.18]; MQ split dose β −1.17 [95% CI –3.59, 1.25]) or total HOME scores (MQ full dose β 0.29 [95% CI –0.10, 0.69]; MQ split dose β 0.34 [95% CI –0.06, 0.74]). Subscale analyses suggested positive associations between MQ full dose and maternal involvement (β 0.12 [95% CI 0.02, 0.22]) and learning materials (β 0.21 [95% CI 0.02, 0.39]).

Interpretation

We found no evidence that MQ compared with SP affected maternal depression, anxiety, or infant development at one year. The associations with maternal involvement and learning materials warrant further investigation into possible indirect or behavioral effects of MQ.

Trial registration

ClinicalTrials.gov NCT 00811421; Pan African Clinical Trials Registry PACTR 2010020001429343.