Abstract
Background:
The potential of aripiprazole to reverse amisulpride-induced hyperprolactinemia has yet to be established.
Objectives:
To investigate whether adjunctive aripiprazole treatment can reverse amisulpride-induced prolactin (PRL) elevation.
Design:
This study was a single-arm retrospective cohort study.
Methods:
We conducted a retrospective analysis of medical records from patients treated with amisulpride at a tertiary specialized psychiatric hospital, between June 1, 2022, and December 31, 2025, who subsequently received adjunctive aripiprazole. Data were collected via the electronic medical record system. A paired t-test was used to compare PRL values before (⩽7 days, PRL1) and after (⩾7 days, PRL2) aripiprazole initiation.
Results:
The difference between PRL1 and PRL2 was not statistically significant (mean difference = 4.115 ng/ml, 95% CI (−5.121, 13.351) ng/ml, SD = 15.284, p = 0.351, Cohen’s d = 0.269, 95% CI for d (−0.290, 0.818)). Stratified analyses by sex and diagnosis yielded results consistent with the main analysis across subgroups (p > 0.05). None of the leave-one-out analyses altered the study’s conclusions (all p-values > 0.05), indicating that the results were not driven by any single extreme value.
Conclusion:
In individuals with amisulpride-related hyperprolactinemia, this retrospective study suggests that adjunctive aripiprazole has an insignificant therapeutic effect in reversing PRL elevation. Therefore, until more comprehensive research evidence becomes available, a cautious approach should be taken in clinical practice regarding the use of adjunctive aripiprazole to reverse amisulpride-induced PRL elevation.
Plain language summary
Amisulpride is a widely used atypical antipsychotic in clinical practice. However, it is associated with the adverse effect of inducing hyperprolactinemia, which can lead to serious consequences and poses significant health risks for patients. There is insufficient clinical evidence regarding whether adjunctive aripiprazole can reduce the prolactin elevation caused by amisulpride. Using real-world data, this study answered a question commonly asked in clinical practice. We preliminarily used a retrospective study to demonstrate that adjunctive aripiprazole fails to reverse amisulpride-induced hyperprolactinemia, making this the largest-scale study to date supporting the conclusion. This study is submitted as a ‘brief report’ to provide a preliminary yet significant warning regarding the potential inefficacy of adjunctive aripiprazole in amisulpride-induced hyperprolactinemia, aiming to facilitate more prudent clinical decision-making and inform future large-scale research.
Introduction
Hyperprolactinemia (HPRL) is a common neuroendocrine adverse reaction induced by antipsychotic drugs, which may lead to a range of risks such as menstrual disorders in women, gynecomastia in men, galactorrhea, sexual dysfunction, infertility, reduced bone mineral density, and increased risk of breast cancer. 1 Previous studies have shown that both typical and atypical antipsychotics may cause hyperprolactinemia. 2 Among these, amisulpride is considered the antipsychotic most strongly associated with elevated prolactin (PRL) levels, 3 surpassing risperidone and olanzapine.4,5 A dose–response meta-analysis published in 2025 by Lin et al. 6 suggested that antipsychotic-induced HPRL is dose-dependent; however, due to insufficient evidence from related studies, this conclusion does not include amisulpride. The study by Kopecek et al. 5 suggested that all patients treated with amisulpride at doses as low as 50 mg experience HPRL; Glatard et al. 7 mentioned in their review that amisulpride-induced HPRL does not appear to have a dose–response relationship; Paparrigopoulos et al. 8 found that the significant prolactin-elevating effect of amisulpride was not influenced by dosage or duration of administration.
For antipsychotic-induced PRL elevation, the primary recommendation is to switch to an alternative agent. If discontinuation of the medication is not possible, adjunctive pharmacotherapy may be considered, such as adding one or more drugs like aripiprazole, metformin, Chinese herbal preparations, or dopamine agonists. 9 Among these, aripiprazole is a high-potency dopamine D2 receptor partial agonist. It acts as a D2 receptor antagonist in hyperdopaminergic states and as an agonist in hypodopaminergic states. This unique pharmacological mechanism allows it to treat psychiatric symptoms without increasing serum prolactin levels. 10 Aripiprazole is known as an antipsychotic with prolactin-lowering effects and even reduces prolactin levels below those observed in placebo groups. 11 Substantial evidence indicates that adjunctive aripiprazole can be effective in treating antipsychotic-induced hyperprolactinemia.12–17 Of these studies, three meta-analyses included randomized controlled trials in which the antipsychotics causing HPRL did not include amisulpride.12–14 Another network meta-analysis included amisulpride, but did not conduct a subgroup analysis on whether adjunctive aripiprazole could be used to treat amisulpride-induced HPRL. 15 A 6-month retrospective cohort study by Yang et al. 17 showed that among 12 patients with amisulpride-induced HPRL, only five experienced remission of HPRL after adjunctive treatment with aripiprazole. Paulzen et al. 18 also reported a case in which aripiprazole failed to reverse amisulpride-induced HPRL. Therefore, the efficacy of adjunctive aripiprazole for amisulpride-induced PRL elevation remains unclear.
This study aims to investigate, through a retrospective single-arm design, whether amisulpride-induced PRL elevation demonstrates a dose–response relationship and to examine whether adjunctive aripiprazole treatment can reverse PRL elevation induced by amisulpride.
Materials and methods
Study design and ethics
This study is a single-center, retrospective, observational, single-arm cohort investigation. We retrospectively analyzed the medical records of patients who were treated with amisulpride and subsequently received adjunctive aripiprazole therapy at a tertiary specialized psychiatric hospital, between June 1, 2022, and December 31, 2025. The reporting of this study conforms to the STROBE statement (Supplemental Material). 19
Study population
We initially screened hospitalized patients within the study period who had received amisulpride and subsequently initiated adjunctive aripiprazole treatment via the hospital’s electronic medical record system. The study reduces selection bias through a consecutive enrolment design: every patient identified from June 2022 to December 2025 who met our inclusion criteria was included; exclusions were determined in advance. The inclusion criteria were as follows: (1) Patients were adults or adolescents, of either sex. (2) The primary diagnosis was within the category of psychiatric disorders (according to ICD-10 codes, primarily F00-F99). (3) Patients were treated with amisulpride monotherapy prior to the addition of aripiprazole (excluding anticholinergics, benzodiazepines, and non-benzodiazepine sedative-hypnotics). (4) Aripiprazole was added to the existing amisulpride regimen. (5) Serum PRL levels were measured within 7 days before aripiprazole administration (PRL1) and at least 7 days after aripiprazole administration (PRL2). Only the last result was recorded for duplicate tests. (6) The plasma concentration of amisulpride was measured on the same day as PRL1. (7) Complete clinical medical records were available. Exclusion criteria included the following: concurrent use of other antipsychotics (besides aripiprazole), mood stabilizers, or antidepressants during amisulpride treatment; comorbid conditions or physiological factors potentially causing HPRL, such as pregnancy, lactation, prolactinoma, sellar region tumors, hypothyroidism, polycystic ovary syndrome, or severe renal or hepatic diseases; concurrent use of non-psychiatric medications known to potentially induce HPRL, including peripheral dopamine receptor blockers, proton pump inhibitors, H2 receptor antagonists, or calcium channel blockers. 15 Blood concentration monitoring was used to ensure patient medication adherence.
Data collection
Patients’ details were collected through the electronic medical record system, including gender, age, disease diagnosis, daily dose of amisulpride and its corresponding blood concentration, maximum administered dose of aripiprazole, and PRL1 and PRL2 values. The following quantitative variables were analyzed as continuous measures without categorization: age (in years), dose and blood concentration of amisulpride, PRL1, and PRL2. Prolactin level was considered normal if it was <25 ng/ml for female patients and <20 ng/ml for male patients.
Sample collection and laboratory analysis
We included only fasting venous blood samples collected between 6:00 AM and 8:00 AM. Samples without a clear record of collection time in the medical records or laboratory reports, or with collection times outside this window, were excluded.
All serum PRL concentrations and amisulpride plasma concentrations in this study were measured by the Department of Clinical Laboratory of our hospital. During the study period (June 2022–December 2025), PRL was measured by direct chemiluminescence. Serum concentrations of amisulpride were analyzed by an ultra-performance liquid chromatography-tandem mass spectrometry (UPLC-MS) method. The assay range of amisulpride was from 15 to 1800 ng/ml. Interday relative standard deviations and intraday precision were both within 15%, and accuracy was within the range 85%–115%.
Sample size determination
The primary outcome of this paper is the difference between PRL1 and PRL2. We conducted a sample size estimation using G*Power 3.1 (Kiel, Germany). Assuming an average effect size (Cohen’s d = 0.5), with a one-sided alpha of 0.05 and a power of 80%, the required number of subjects would be approximately 27.
However, as this is a retrospective, single-arm study, all eligible patients who met the inclusion criteria during the study period were included. Within this period, we identified 13 patients. Therefore, the entire cohort under study is fully represented within this specific time frame, and it is not a random sample. No additional eligible patients were available, as this was a single-center study.
Statistical analysis
Statistical analysis was performed using R 4.3.3 for Windows (Vienna, Austria). Continuous variables were reported as means ± standard deviation (SD) or medians (P25, P75), while categorical variables were reported as numbers (percentages).
Correlations between PRL1 and daily dose of amisulpride (normally distributed) were calculated using the Pearson correlation coefficient, whereas correlations between PRL1 and plasma concentration of amisulpride were calculated using the Spearman correlation coefficient, as the data were not normally distributed.
A paired t-test was performed to examine the differences between PRL1 and PRL2. An independent samples t-test or Mann–Whitney U test was used to compare various indices between male and female groups, as well as between patients with schizophrenia and those with non-organic mental disorders. Significance was set at p < 0.05. Because of the limited sample size, individual data points could have a substantial impact on the results. Therefore, data were also subjected to non-parametric statistical analyses, including Spearman’s rank correlation coefficient, Wilcoxon signed-rank test, and Mann–Whitney U test. The results of the non-parametric tests were consistent with those of the parametric tests and are therefore omitted. With the small sample size and the increased risk of type II errors, a post hoc analysis of statistical power will be conducted. Using the observed effect size, the statistical power to detect a paired difference at α = 0.05 (two-tailed) was calculated. In addition, 95% confidence intervals were calculated for the primary outcome, allowing a reasonable range of effect sizes to be assessed rather than relying solely on p-values for inference. Given the small sample size, it is possible that individual responses could influence the results. Therefore, a “leave-one-out” analysis was performed, sequentially omitting each participant’s data to assess the impact on the primary outcomes. The effects of gender and diagnoses on the primary outcome were also examined. There was no missing data in the database, so all analyses were conducted on the full dataset.
Results
The Anding medical database included data from 219 inpatients who received prescriptions for both amisulpride and aripiprazole during a single hospitalization between June 1, 2022, and December 31, 2025, of whom 13 patients were ultimately eligible for this study (Figure 1). There were no missing data for any of the variables of interest.
Diagram about including/excluding the study population.
Among these 13 patients, 5 (38.46%) were males. The patients’ ages ranged from 16 to 54 years (mean ± SD, 31.15 ± 10.01). Ten patients (76.92%) had been diagnosed with schizophrenia, while the remaining 3 (23.08%) had been diagnosed with non-organic mental disorders.
Before the addition of aripiprazole, the daily dosage of amisulpride ranged from 200 to 1200 mg/d (mean ± SD, 684.61 ± 321.06 mg/d), and the blood level was 443.58 ± 253.51 ng/ml. Adjunctive aripiprazole was administered at doses of 5–30 mg/d (median (P25, P75), 10 (5, 15) mg/d). The PRL1 measurement before aripiprazole administration was conducted 0–6 days prior to its use, with an average level of 106.76 ± 49.66 ng/ml. The PRL2 test after aripiprazole treatment was performed 7–38 days after the drug was started, with an average concentration of 110.87 ± 51.51 ng/ml. The PRL levels of all patients were abnormal at both time points, before and after adjunctive aripiprazole treatment.
No correlation was found between amisulpride blood concentration or dosage and PRL1 levels (r = 0.049, p = 0.873, n = 13, 95% CI (−0.516, 0.585); ρ = −0.290, p = 0.337, n = 13, 95% CI (−0.734, 0.327), respectively; see Figures 2 and 3).
Relationship between PRL1 and amisulpride serum concentration.
Relationship between PRL1 and amisulpride daily dose.
The difference between PRL1 and PRL2 was not statistically significant (mean difference = 4.115 ng/ml, 95% CI (−5.121, 13.351) ng/ml, SD = 15.284, p = 0.351, Cohen’s d = 0.269, 95% CI for d (−0.290, 0.818); see Figure 4). Post hoc power analysis for this sample (n = 13) and effect size (d = 0.269) indicates a maximum power of 23%. Leave-one-out sensitivity analysis showed that after sequentially excluding each patient, the range of changes in prolactin levels was between −20.15 ng/ml and 34.08 ng/ml. None of the leave-one-out analyses altered the study’s conclusion (all p-values > 0.05), so we can rule out the influence of any single outlying observation.
The dot plot with lines of PRL1 and PRL2 levels in 13 patients.
This study investigated the influence of gender and diagnosis on the primary outcome. The results indicated no statistically significant differences in age, daily amisulpride dose, or blood concentration between male and female patients (p > 0.05). Although differences in PRL1 and PRL2 levels between genders were statistically significant (p < 0.05, with higher values in females than in males), the difference in the change between PRL1 and PRL2 (ΔPRL) was not statistically significant (p > 0.05). There were no statistically significant differences in any of the above indicators between patients with schizophrenia and those with non-organic mental disorders (see Table 1).
Patient demographics and characteristics of different genders and diagnoses.
Due to the small sample size of this subgroup (n = 3), the data are presented as median (minimum-maximum).
A significant difference was observed in male compared to female patients (*p < 0.05).
PRL, prolactin.
Discussion
This study retrospectively analyzed the clinical data of 13 patients treated with amisulpride who received adjunctive aripiprazole therapy at a tertiary mental health center. Based on these preliminary findings, no relationship was found between the blood concentration or dose of amisulpride and prolactin (PRL) levels after treatment, within the therapeutic dose range of amisulpride. Furthermore, for patients whose prolactin levels were elevated due to amisulpride, the addition of aripiprazole did not demonstrate significant effectiveness.
Bressan et al. in their 2004 study 20 found no correlation between PRL levels and either the daily dose or plasma concentration of amisulpride after administration, which is consistent with the findings of this study. This outcome may be due to differences in D2 receptor occupancy between the central and peripheral systems with amisulpride. The anterior pituitary is located outside the blood–brain barrier, so antipsychotic drugs can have a direct effect on the D2 receptors in this area without actually crossing the blood-brain barrier. 21 P-glycoprotein has a very high affinity for amisulpride and therefore selectively removes it from the brain while increasing its concentration in the hypothalamic effluent blood. As a result, a more potent dopamine receptor blockade is achieved in the prolactin-producing cells of the anterior pituitary. 22 Therefore, it is hypothesized that the low-potency drug amisulpride, even at low doses, may effectively bind to pituitary D2 receptors and cause an increase in PRL. This is also why almost all patients on 50 mg or lower doses of amisulpride develop HPRL.
An earlier small-sample (n = 5) study by Chen et al. 23 showed that co-administration of aripiprazole might alleviate the PRL elevation caused by amisulpride to a certain extent. However, such a reduction should not be considered of major clinical relevance as the drop in PRL levels was very small (from 144.9 ± 24.4 ng/ml to 127.5 ± 21.7 ng/ml). Similarly, Paulzen et al. 18 noted that, in their opinion, although aripiprazole has a greater affinity for the dopamine D2 receptor than amisulpride, the previously mentioned difference between central and peripheral D2 receptor occupancy by amisulpride makes it highly unlikely that aripiprazole would bind to tuberoinfundibular dopamine D2 receptors to such an extent that it could counteract amisulpride-induced hyperprolactinemia. As this is a case report, the evidence merely describes a clinical phenomenon and serves as a basis for hypothesis generation rather than generalizable conclusions. The efficacy of aripiprazole in amisulpride-induced hyperprolactinemia requires further validation in higher-level studies.
Based on the above discussion, the present study concluded that, in the absence of more robust research evidence, caution is warranted in clinical practice when considering adjunctive aripiprazole to reverse amisulpride-induced PRL elevation. This suggestion may not only spare patients additional financial burden and risk of side effects but also serve as a clinical caution. However, post hoc power calculation indicates that, given the observed effect size of Cohen’s d = 0.269, the true power of the study would have been 23%, which is well below the conventional threshold of 80%. Therefore, even if aripiprazole does have some reversing effect in amisulpride-induced HPRL, this study had a 77% chance of failing to detect any actual effect due to the small sample size. A more appropriate conclusion of our study is that no statistically significant effects of aripiprazole can be detected based on these findings alone, as the study is limited and requires more data. Thus, the results should be considered hypothesis-generating rather than conclusive.
Limitations
This research is significantly limited in several aspects. First, it used a retrospective single-arm cohort design, whose main drawback is the absence of a concurrently randomized control group. As a result, it could not control for the effects of both known and unknown confounding variables, such as the influence of stress on PRL. Furthermore, the small sample size and the single-center experience further restrict the generalizability of the findings. In addition, in our study, the schedule of PRL evaluations depended on clinical needs rather than a prearranged research plan, and the follow-up period after the introduction of adjunctive aripiprazole was relatively short, which could compromise the study’s internal validity. In addition, due to the small sample size, we could not conduct stratified analyses according to aripiprazole dose. Hence, it is not known whether there were different treatment effects of aripiprazole at different doses. It is also unable to address important confounders such as the severity and duration of the illness, as well as the effects of prior treatments. In the future, larger prospective studies are needed with adequate controls, including the use of multivariate regression, propensity scores, and mixed-effects models. In summary, future randomized controlled trials with sufficient statistical power are needed to determine whether aripiprazole can reverse amisulpride-induced hyperprolactinemia.
Conclusion
In patients with HPRL due to amisulpride, the present retrospective study suggests that adjunctive treatment with aripiprazole shows limited efficacy. Therefore, until more comprehensive research evidence becomes available, a cautious approach should be taken in clinical practice regarding the use of adjunctive aripiprazole to reverse amisulpride-induced PRL elevation. However, as this study was retrospective and has the limitations mentioned above, this decision needs to be verified by prospective, randomized controlled trials in the future.
Supplemental Material
sj-docx-1-tpp-10.1177_20451253261450772 – Supplemental material for Adjunctive aripiprazole for amisulpride-induced hyperprolactinemia: a preliminary retrospective study showing no reversing effect
Supplemental material, sj-docx-1-tpp-10.1177_20451253261450772 for Adjunctive aripiprazole for amisulpride-induced hyperprolactinemia: a preliminary retrospective study showing no reversing effect by Xiaoqian Lan, Hongyan Zhuang and Pengfei Li in Therapeutic Advances in Psychopharmacology
Footnotes
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
