Impact of medication side effects on medication adherence and implications for treatment outcomes in patients with borderline personality disorder: a pilot study Authors: Sarah Wazlawik, Claudia Oppenauer Published: 2026 ABSTRACT AbstractBackgroundBorderline Personality Disorder (BPD) is a severe mental illness characterized by unstable interpersonal relationships, affective instability, and impulsivity. While psychotherapy is the recommended first-line treatment, over 80% of BPD patients use psychotropic medication and often report side effects. This pilot study examines whether perceived medication side effects influence treatment outcomes in BPD, anxiety, and depression symptoms indirectly through their effect on medication adherence, with adherence conceptualized as a potential mediator of this relationship.MethodsThis retrospective study included patients from the personality disorder unit at University Hospital Eggenburg. BPD diagnoses were confirmed using the Borderline Symptom List (BSL-23). Perceived side effects and medication adherence were assessed via self-report questionnaires at admission. Three mediation analyses examined the association between side effects and treatment outcomes (using the BSL-23, PHQ-9, and GAD-7 for BPD, depression, and anxiety symptoms, respectively), with adherence as a mediator and gender and baseline symptom severity as covariates.Results84% of patients (n = 48) reported side effects, most commonly depression, fatigue, dizziness, and tremor. The leading cause of nonadherence was forgetting to take medication. The mediation analyses showed no indirect effect of X on Y via M (n = 31; BPD: indirect effect = 0.0011, 95% CI = [-0.0209, 0.0169], anxiety: indirect effect = 0.0038, 95% CI = [-0.0989, 0.1075], depression: indirect effect = 0.0002, 95% CI = [-0.0813, 0.0946]).ConclusionsDespite frequent side effects, adherence remained high—likely due to the structured inpatient setting and voluntary admission. The lack of significant associations suggests other factors may influence treatment outcomes. Future research should involve larger samples, outpatient settings, follow-ups, and objective adherence measures. Targeted interventions addressing specific side effects may help improve adherence and clinical outcomes.Trial registrationNot applicable (retrospective observational study, no healthcare intervention trial).Keywords: Borderline personality disorder, Medication side effects, Medication adherence, Psychotropic medication, Treatment outcomes Background Borderline Personality Disorder (BPD) is a severe mental illness that is characterized by unstable and intense interpersonal relationships, an unstable sense of self, affective instability, frantic efforts to avoid real or imagined abandonment and impulsivity in areas like substance abuse or binge eating [1]. BPD has a prevalence of 1.6% in the general population and 6.4% among adult primary care visits [2]. Patients with BPD often exhibit comorbid psychological disorders, with lifetime comorbidity rates of 84.5% for anxiety disorders, 82.7% for mood disorders, and 78.2% for substance use disorders [3]. In terms of treatment, current practice guidelines recommend psychotherapy as the first-line treatment to achieve lasting improvements in patients’ overall functioning and personality [4]. European guidelines broadly agree that outpatient psychotherapy should be the primary treatment for BPD with inpatient care recommended when necessary [5]. However, there is no consensus on the use of pharmacotherapy, and no medication is officially approved for BPD treatment [6]. APA guidelines suggest using pharmacotherapy to manage acute decompensation and trait vulnerabilities [4], while NICE guidelines recommend pharmacotherapy only for BPD patients with severe comorbid disorders and only for short periods of time [7]. The German AWMF guidelines recommend psychopharmacotherapy as a second-line treatment used in addition to psychotherapy and for a limited time and advise against polypharmacy due to the lack of benefit over single medication [8]. Swiss, Finnish and Dutch guidelines suggest pharmacotherapy for specific symptoms, such as anger or impulsivity [5]. Despite these recommendations, over 80% of BPD patients in Europe are prescribed medication, with more than half receiving three or more drugs simultaneously [6]. A study by Mehlum et al. [9] found that 86% of adolescent patients with BPD features in Barcelona and 67% in the United States were using psychotropic medications, likely due to the broad range of symptoms presented in BPD patients [7]. Moreover, BPD patients may be more sensitive to medication, potentially leading to new symptoms and additional prescriptions [10]. Clinicians frequently prescribe antidepressants, such as sertraline or venlafaxine [1], although randomized trials show minimal benefit over placebo [11]. Mood stabilizers, such as valproate semisodium and lithium, are used cautiously due to their adverse effects [4]. More recently, atypical antipsychotics like quetiapine or aripiprazole have shown promising results [12], improving multiple symptom areas with relatively mild adverse effects [7]. The use of multiple medications and combinations can lead to numerous side effects including weight gain, tremor, nausea, fatigue, and headaches [4, 10]. However, most of those studies focused on different patient populations, such as epilepsy and depression (e.g. [13, 14]). A review by Bozzatello et al. compared various psychotropic drugs for patients with BPD, and also listed side effects reported by different patient populations. While adverse effects from medication in patients with BPD were significantly underreported, common side effects included nausea, fatigue, tremor, dizziness, weight gain, and headaches [7]. Research indicates that severe side effects contribute to medication nonadherence, thereby affecting drug efficacy [13, 14]. Some reported side effects may also be attributable to patients’ expectations of experiencing side effects (known as the nocebo effect) rather than being actual pharmacological properties of the drug [15]. Individuals with heightened anxiety may be more susceptible to the nocebo effect, as they are more likely to attribute pre-existing symptoms to medication side effects [16]. Anxious people are also shown to be more prone to somatization, which can increase the reports of side effects [15]. Nonadherence can range from minor deviations from the treatment plan to complete abandonment of therapy [17]. Common reasons for medication nonadherence include fear of dependence and adverse effects [17, 18]. Side effects like weight gain, cognitive impairment, and agitation are linked to higher medication nonadherence and treatment discontinuation, resulting in increased healthcare utilization [13]. In one of the few studies on medication adherence in patients with BPD, Mirhaj Mohammadabadi et al. [17] identified side effects, poor therapeutic alliance, and fear of medication dependence as key predictors of medication nonadherence. Cook et al. [19] found that patient self-reports often overestimate adherence due to guilt, desire for approval, or lack of awareness of nonadherence severity. Health care practitioners also tend to believe patients are more adherent than they are [19]. In a study on antiepileptic drugs, severe side effects led to early treatment discontinuation in 25% of the patients [20]. Treatment dropout, or nonadherence in general, is linked to various adverse effects in multiple psychiatric disorders, such as worse treatment outcomes [21, 22]. Similarly, a study on patients with schizophrenia demonstrated that medication nonadherence was associated with a three-to fivefold increased risk of relapse or (re)admission [23]. In order to maximize treatment efficacy, several studies have tried to identify additional factors beyond medication adherence that predict poor treatment outcomes. Higher pretreatment symptom severity in a sample of patients with BPD has been found to predict poorer treatment outcomes, although some studies suggest the opposite effect [24, 25]. Moreover, the complexity of medication regimens has been identified as a barrier to medication adherence in these studies. A review examining a wide range of conditions such as diabetes or psychiatric disorders, found that a 50% reduction in the risk of medication nonadherence was observed when medication regimens were simplified to once-daily dosing [26]. Additionally, patient information and education were found to be significant predictors of medication adherence [27]. Although psychotherapy remains the first-line treatment for BPD, pharmacotherapy is frequently prescribed for specific symptom domains and, in particular, for comorbid psychiatric disorders [28]. This is especially relevant given the high prevalence of comorbidities among individuals with BPD [29]. Comorbid disorders can, in turn, decrease medication adherence, which can lead to worse treatment outcomes, such as increased rehospitalization. Therefore, it stands to reason that adjunctive pharmacotherapy may still be beneficial in this group. However, Gunderson and Choi-Kain [30] argue that medication should be prescribed with healthy scepticism. Based on existing research, we hypothesize that patients who experience severe side effects are less likely to adhere to their prescribed medication, which may negatively affect treatment outcomes for BPD, depression and anxiety. Hypotheses: Subjective perceived medication side effects influence treatment outcomes in patients with BPD at discharge. Subjective perceived medication side effects influence treatment outcomes in patients with BPD at discharge. This association is mediated by medication adherence. Subjective perceived medication side effects influence medication adherence in patients with BPD at admission.Medication adherence influences treatment outcomes (BPD, depression, and anxiety symptoms) at discharge. This association is mediated by medication adherence. Subjective perceived medication side effects influence medication adherence in patients with BPD at admission.Medication adherence influences treatment outcomes (BPD, depression, and anxiety symptoms) at discharge. Subjective perceived medication side effects influence medication adherence in patients with BPD at admission. Subjective perceived medication side effects influence medication adherence in patients with BPD at admission. Medication adherence influences treatment outcomes (BPD, depression, and anxiety symptoms) at discharge. Medication adherence influences treatment outcomes (BPD, depression, and anxiety symptoms) at discharge. Methods Participants and study design Patients from the personality disorder treatment unit of University Hospital Eggenburg treated between 2021 and 2022 were included in this retrospective data analysis. All patients were self-referred or referred by mental health professionals to the clinic. Patients with acute psychotic symptoms, suicidal behaviour, acute intoxication, or multiple substance use are not admitted to treatment at this hospital. Psychological outcomes were routinely monitored using the Computer-based Health Evaluation System (CHES) [31] at preadmission, treatment start, and discharge. Patients were assigned to the personality disorder treatment unit based on case history, a treatment preparation screening form, and an unstructured clinical interview by a psychiatrist. Baseline Borderline Symptom List (BSL-23) results were used to validate the diagnosis of BPD, with only data from patients scoring ≥ 0.28 included. This resulted in a sample of 178 patients. However, because not all participants were given certain questionnaires (e.g., the MARS) relevant to our research questions, for reasons unknown to us, the final sample decreased to 57 patients. The patients were treated over a span of 8 to 12 weeks with a combination of individual and group therapy. The therapy involved a structured schedule of one to two individual psychotherapy sessions and five group psychotherapy sessions per week. Measures Borderline symptom list – short version (BSL-23) [32] The BSL-23 is a self-report questionnaire with 23 items measuring thought patterns and behaviors associated with BPD in the last week. Answers are given on a five-point Likert scale plus a subjective wellbeing item scoring from 1 to 100% [32]. BSL-23 mean scores were classed into six severity categories, with a severity of “none or low” corresponding to mean scores of 0–0.28, “mild” corresponding to scores of 0.28–1.07, “moderate” corresponding to scores of 1.07–1.87, “high” corresponding to scores of 1.87–2.67, “very high” corresponding to scores of 2.67–3.47 and “extremely high” corresponding to scores of 3.47–4 [33]. The BSL-23 has good to excellent psychometric properties, with high internal consistency (Cronbach’s Alpha = 0.97) and high test-retest reliability (r = .82; p = .0001) [31]. Furthermore, correlations between the BSL-23 and similar scales were moderate to high, substantiating convergent validity [32]. Questionnaire for the assessment of adverse effects of antiepileptic drugs (FENAT) [34] The FENAT lists 28 side effects experienced by patients on anticonvulsant medication. The patients answer either yes or no for each side effect and if they answered yes, they need to rate the side effect’s perceived severity (light, medium, or strong). Patients also need to indicate the most impairing side effect. Scores range from 0 (no side effects), to 27. Although originally designed for individuals with epilepsy, the questionnaire includes a broad range of potential side effects that overlap with those commonly experienced by people with mental illnesses taking medication. Therefore, the FENAT seems well suited for assessing side effects across a wide range of psychopharmacological treatments. The FENAT demonstrates high internal consistency (Cronbach’s Alpha = 0.89 – 0.92) and validity (significant correlations with similar scales) [34]. Medication adherence report scale (MARS) [35] The MARS is a self-report questionnaire with five items to measure medication nonadherence on a five-point Likert scale. It is based on the MARS-10 by Horne & Weinman [36]. A score of 22 or above indicates good medication adherence [37]. Reliability and validity were tested for different patient groups, and internal consistency was acceptable/good, with Cronbach’s Alpha coefficients ranging from 0.67 to 0.89 [35]. Moreover, criterion-related validity and construct validity were also good [35]. Patients health questionnaire 9 depression (PHQ-9) [38] The PHQ-9 is a self-report questionnaire with nine items measuring depression symptoms over the last two weeks on a four-point Likert scale. All items are summed to a total score. Cutoffs were defined as “none-minimal” for scores of 0–4, “mild” for scores of 5–9, “moderate” for scores of 10–14, “moderately severe” for scores of 15–19, and “severe” for scores of 20–27 [39]. The PHQ-9 demonstrates good internal consistency (Cronbach’s Alpha = 0.86 – 0.89) and criterion- as well as construct validity [40]. Generalized anxiety disorder-7 (GAD-7) [41] The GAD-7 measures anxiety symptoms over two weeks in seven items on a four-point Likert scale summed to a total anxiety score. Scores were classed into the categories “none-minimal” for cutoffs of ≤ 4, “mild” for cutoffs of 5–9, “moderate” for cutoffs of 10–14 and “severe” for cutoffs of ≥ 15 [42]. The GAD-7 shows excellent internal consistency (Cronbach’s Alpha = 0.92) and good test-retest reliability (intraclass correlation = 0.83) [41]. Criterion validity and construct validity were also good [41]. The GAD can also be used to screen for other anxiety disorders, showing acceptable sensitivity and specificity [43]. Statistical analysis We conducted three mediation analyses using PROCESS for SPSS [44] to examine the impact of the patients’ subjective experienced medication side effects on therapy outcome, with adherence as a mediator and gender and baseline symptom severity as covariates. Power analyses were conducted for regression models using G*Power [45]. Assuming a medium effect size (f² = 0.15), α = 0.05, and power (1-β) = 0.80, the required sample size for each analysis was calculated to be 55 participants. Results Sociodemographic and clinical description of the sample The study included data from 57 patients (14 male, 43 female), with a mean age of 33.12 years (SD = 11.06) treated at the unit for personality disorders. Due to organizational issues that led to missing data, the maximum N fluctuated depending on the variable investigated. The largest proportion of participants had a secondary school degree (e.g. Matura), 42%, followed by 23% who had completed general compulsory education. Out of n = 31, 48% of patients were either unemployed, receiving a rehabilitation allowance or permanently retired, compared with 23% of patients who were employed full-time and 10% who were employed part-time. Moreover, more than half (58%) were currently on sick leave or were considered unemployable. Most of the patients (86%) had already experienced at least one previous inpatient mental health care stay, with 32% having been in more than five previous inpatient treatments. For 68% of patients, it was their first stay at University Hospital Eggenburg. Additionally, 58% of patients had been diagnosed with BPD for more than two years. Table 1 illustrates sociodemographic variables and baseline assessments of BPD, depression, and anxiety. Table 1Sociodemographic variables and baseline assessmentsVariables N Meanor %SDAge5733.1211.06Gender, female5775%Education (secondary education)5742%Previous inpatient mental health care223.861.94Chronicity (> 2 years)3158% Clinical Evaluation N Mean SD BSL-23572.080.98PHQ-95716.885.67GAD-75713.214.55FENAT578.636.07MARS5721.963.35Note: BSL-23 Borderline Symptom List, PHQ-9 Patient Health Questionnaire (module depression), GAD Generalized Anxiety Disorder (questionnaire), FENAT Questionnaire for the Assessment of Adverse Effects of Antiepileptic Drugs, MARS Medication Adherence Report Scale, SD Standard deviation Sociodemographic variables and baseline assessments Note: BSL-23 Borderline Symptom List, PHQ-9 Patient Health Questionnaire (module depression), GAD Generalized Anxiety Disorder (questionnaire), FENAT Questionnaire for the Assessment of Adverse Effects of Antiepileptic Drugs, MARS Medication Adherence Report Scale, SD Standard deviation Treatment outcomes We conducted paired student’s t tests of the BPD, depression, and anxiety scores from admission to discharge (see Table 2). Since not all patients completed the outcome questionnaires at the end of their stay, the sample size decreased from N = 57 to n = 31 for the BPD, depression, and anxiety comparisons. However, patients improved significantly in BPD, depression, and anxiety symptoms during their stay. The mean score for the BSL-23 improved from “high” to “moderate”. For the PHQ-9 questionnaire, the mean score improved from a “moderately severe” classification to a severity level between “mild” and “moderate”. Additionally, the mean score for the GAD-7 improved from a “moderate” to a “mild” severity classification. Table 2Treatment outcomes of BSL-23, PHQ-9, and GAD-7 scores from T1 to T2T1T2Mean differenceStatistical testMSDMSDM95% CItdf p dBSL-232.080.981.511.05-0.57[0.16; 0.85]3.01300.0030.54PHQ-916.885.679.424.53-7.46[4.93; 8.62]7.5130< 0.0011.35GAD13.214.558.264.79-4.95[2.72; 6.25]5.1830< 0.0010.93Note: BSL-23 Borderline Symptom List, PHQ-9 Patient Health Questionnaire (module depression), GAD Generalized Anxiety Disorder (questionnaire), T1 & T2 admission and discharge, M Mean, SD Standard deviation, 95% CI 95% confidence interval, t t-value, df degrees of freedom, p Statistical significance, d Cohen’s effect size Treatment outcomes of BSL-23, PHQ-9, and GAD-7 scores from T1 to T2 Note: BSL-23 Borderline Symptom List, PHQ-9 Patient Health Questionnaire (module depression), GAD Generalized Anxiety Disorder (questionnaire), T1 & T2 admission and discharge, M Mean, SD Standard deviation, 95% CI 95% confidence interval, t t-value, df degrees of freedom, p Statistical significance, d Cohen’s effect size Medication side effects and medication adherence The number of side effects reported by the patients is illustrated in Fig. 1. Distribution of the number of patients reporting medication side effects Out of a list of 27 different possible side effects, 16% of the patients (n = 9) reported not experiencing any of these side effects (see Table 3). Two third (65%) of patients reported difficulties concentrating, followed by 63% experiencing fatigue and 56% facing inner unrest. Other very common side effects were depression (54%), memory problems (54%), and mood changes (53%). Very few patients encountered regurgitation (4%), weight loss (7%), or changes of the gums (9%). Additionally, patients were also asked to rate from mild to severe how much they were impaired by each symptom. Patients felt most impaired by skin problems (M = 2.45, SD = 0.82), gait unsteadiness (M = 2.40, SD = 0.70), and changes of the gums (M = 2.40, SD = 0.89). The symptoms reported to be the least impairing were regurgitation (M = 1.50, SD = 0.71), tremor (M = 1.60, SD = 0.68), and slowed reactions (M = 1.72, SD = 0.75). Table 3Prevalence and impairment of specific medication side effectsSymptom N %Impairment M SD Difficulties concentrating3765%2.270.73Fatigue3663%2.190.62Inner unrest3256%2.380.61Depression3154%2.390.67Memory problems3154%2.320.70Mood changes3053%2.370.67Slowed thinking2646%2.080.80Headaches2442%1.790.78Stomach issues2340%2.260.75Sleep disorders2340%2.170.72Tremor2035%1.600.68Slowed reactions1832%1.720.75Dizziness1730%1.880.86Weight gain1730%2.350.70Speech difficulties1628%2.000.73Appetite loss1526%1.870.74Changes in sex drive1526%2.330.72Sleeplessness1425%2.290.73Hair loss1221%1.830.83Skin problems1119%2.450.82Gait unsteadiness1018%2.400.70Changes in potency or pain during sex1018%2.100.74Double vision712%2.000.82Disturbed body coordination611%2.170.98Changes of the gums59%2.400.89Weight loss47%1.750.96Regurgitation24%1.500.71Note: N Number of patients, % percentages, M Mean, SD Standard deviation Prevalence and impairment of specific medication side effects Note: N Number of patients, % percentages, M Mean, SD Standard deviation 14 patients provided free-text responses specifying which symptoms impaired them the most. They most frequently mentioned depression, fatigue, dizziness, and tremor. We also split the patients into four groups depending on their anxiety levels (using the GAD-7 severity classification) and compared means and standard deviations of the number of side effects the patients reported experiencing. We found that with higher anxiety scores, patients reported experiencing more side effects from their medication (Table 4). Table 4Means and standard deviations of number of medication side effects depending on anxiety severityAnxiety severity n M SD None-minimal11.00-Mild125.253.84Moderate188.616.00Severe2610.506.33Note: n Number of patients, M Mean, SD Standard deviation Means and standard deviations of number of medication side effects depending on anxiety severity Note: n Number of patients, M Mean, SD Standard deviation Regarding adherence levels (see Fig. 2), 38 patients (66%) demonstrated good medication adherence, with 16 patients (28%) achieving the maximum score on the MARS questionnaire, indicating complete adherence to the prescribed medication regimen. Distribution of the patients’ adherence scores Concerning the medication adherence items (see Table 5), the most common reason for medication nonadherence was forgetting to take the medication, with a mean of 3.79 (SD = 1.21). For this item, 21% of the patients selected the response “often”. Table 5Means and standard deviations of the MARS itemsItem M SD I forget to take the medication.3.791.21I change the dose.4.321.04I take less than prescribed.4.491.02I purposefully skip a dose.4.670.79I discontinue the medication for a while.4.700.65Note: M Mean, SD Standard deviation Means and standard deviations of the MARS items Note: M Mean, SD Standard deviation Associations between medication side effects and treatment outcomes, mediated by medication adherence We conducted three mediation analyses to test whether there was a significant association between the patients’ reported medication side effects and treatment outcomes via medication adherence (see Table 6). There was no significant main effect of medication side effects on BPD (p = .6447), depression (p = .9701), and anxiety symptoms (p = .3787). Furthermore, there were no significant indirect effects of side effects on treatment outcomes via adherence (BPD: CI = [-0.0211; 0.0161], depression: CI = [-0.0757; 0.0751], anxiety: CI = [-0.0989; 0.1075]). Table 6Mediation model resultsPathEstimate SE t p CI BPD symptoms (BSL-23) Side effects – Adherence0.07260.09450.76800.4491− 0.1213 – 0.2665Side effects –BPD symptoms0.01470.03150.46660.6447− 0.0500 – 0.0794Adherence –BPD symptoms0.01460.06340.23040.8196− 0.1157 – 0.1449Indirect effect0.00110.0092− 0.0211 – 0.0161Total effect0.01570.03060.51480.6109− 0.0470 – 0.0785 Depression symptoms (PHQ-9) Side effects – Adherence0.07510.10030.74950.4600− 0.1306 – 0.2808Side effects –Depression symptoms0.00560.14790.03780.9701− 0.2978 – 0.3090Adherence –Depression symptoms0.00260.28760.00910.9928− 0.5875 – 0.5927Indirect effect0.00020.0354− 0.0757 – 0.0751Total effect0.00580.14410.03990.9684− 0.2894 – 0.3009 Anxiety symptoms (GAD-7) Side effects – Adherence0.10480.08811.19060.2442− 0.0758 – 0.2855Side effects –Anxiety symptoms0.14550.16240.89570.3787− 0.1884 – 0.4793Adherence –Anxiety symptoms0.03620.34600.10450.9176− 0.6751 – 0.7474Indirect effect0.00380.0488− 0.0989 – 0.1075Total effect0.14930.15540.96060.3453− 0.1696 – 0.4681Note: SE Standard error, t t value, p Statistical significance, CI Confidence interval Mediation model results Side effects – BPD symptoms Adherence – BPD symptoms Side effects – Depression symptoms Adherence – Depression symptoms Side effects – Anxiety symptoms Adherence – Anxiety symptoms Note: SE Standard error, t t value, p Statistical significance, CI Confidence interval Discussion This study aimed to investigate the impact of perceived medication side effects on treatment outcomes, and whether this relationship was mediated by medication adherence, in patients with BPD. Contrary to our initial hypotheses, the results did not show significant effects of medication side effects on treatment outcomes for BPD, depression, and anxiety scores via medication adherence. Given the limited research on the experiences of patients with BPD regarding medication side effects, our study contributes valuable insights into the prevalence and impact of side effects commonly experienced by this population when taking psychotropic medications. Medication Side Effects and Medication Adherence One key finding was the high prevalence of subjective side effects, with 84% of patients reporting at least one side effect. Nevertheless, medication adherence levels were generally high, with 66% demonstrating good adherence. Moreover, the mediation analyses did not reveal a significant association between side effects and medication adherence. These results contrast with previous research, which has consistently shown that medication side effects are a major barrier to medication adherence across various patient populations, including those with mental health disorders [13, 14, 46, 47]. Thus, these results might be influenced by the characteristics of our patient group which was compromised of inpatients with a highly structured and supporting therapeutic environment. In such settings, medication adherence might be improved by regular monitoring and the availability of immediate clinical and psychotherapeutic support. This assumption is supported by Janssen et al. [48], who found a significant association between voluntary hospital admission and increased medication adherence. However, there are also other studies in inpatient settings who report high medication nonadherence and elevated hospital (re)admissions rates [49, 50]. Additionally, self-report measures of adherence may lead to overestimation as suggested by Cook et al. [19]. Oftentimes, patients indicate high adherence because they are not aware of the extent of their nonadherence [19]. In other cases, they might intentionally give a false report, because they are afraid of repercussions or due to a desire for approval [51]. However, it needs to be noted that these findings pertain to different patient populations, such as patients with schizophrenia, depression, breast cancer and hypertension. To our knowledge, there are almost no studies focusing on medication adherence in BPD patients. Concerning the types of side effects experienced, the most common ones in this study were difficulties concentrating, fatigue and inner unrest. Fatigue was also a side effect of antiepileptic drugs reported by other studies [18], as well as memory problems, tremor, gastrointestinal issues, depression, irritability, dizziness, and drowsiness [18, 52, 53]. Most of these adverse effects were also reported by the patients in our study. Furthermore, antipsychotics are often shown to be associated with sedation, weight gain, hypotension, and sexual dysfunction [54]. There is also an association between antidepressants and gastrointestinal issues, weight change, sleep problems, anxiety, and tremor [54]. Anxiety, gastrointestinal issues, and sleep problems were also frequently reported in our study. Patients with BPD experienced in particular fatigue, inner unrest/anxiety, nausea, tremor, dizziness, weight gain, and headaches [7]. However, when patients had to indicate which side effects they found most impairing, a different pattern emerged. In our sample, dermatological issues, gait instability, and gingival changes were identified as the most burdensome. In contrast, a study by Hunkeler et al. [55] described drowsiness, sexual dysfunction, and dry mouth as the most commonly reported side effects with weight gain, anxiety, and rash/itching as the most burdensome. Lam et al. [56] investigated the impact of antidepressant side effects on patients’ occupational functioning. Their findings indicated that daytime sedation, insomnia, headache, and anxiety/agitation were the most disruptive side effects in the workplace [56]. These discrepancies highlight the variability in patients’ experiences with and perceptions of medication side effects across different studies. Moreover, our findings emphasize the distinction between the prevalence of side effects and their perceived severity. The analysis of our study data also revealed a correlation between patients’ anxiety scores and the number of side effects reported. Specifically, higher anxiety levels were associated with a greater number of side effects compared to patients with lower anxiety levels. This finding could suggest a nocebo effect, as research indicates that patients with higher anxiety are more sensitive to bodily sensations and are more likely to attribute these sensations to medication side effects [7]. Furthermore, anxious individuals tend to be more concerned about experiencing medication side effects [15, 16, 57]. Medication side effects, adherence and treatment outcomes Our data revealed significant improvements for the BPD, depression, and anxiety symptoms after treatment. However, no significant relationship was found between medication side effects and treatment outcomes via adherence. This contrasts with a substantial body of literature indicating that better adherence is associated with better treatment outcomes [21, 22]. One reason for this discrepancy could be the comparably more important role of psychotherapeutic interventions in BPD compared to the more supportive and symptom-targeted impact of medication. Moreover, we have only investigated the period from admission to discharge, and we have no follow up data about medication adherence and treatment outcomes after the intense multidisciplinary treatment. Our detailed analysis of patients’ medication adherence demonstrated that forgetting to take the medication was the most common reason for medication nonadherence. This aligns with findings from two studies on antidepressants, which identified forgetfulness as a common cause for nonadherence [58, 59]. The use of pharmacotherapy in individuals with BPD remains characterized by a fundamental clinical tension. Although psychotherapy is widely regarded as the evidence-based first-line treatment for BPD, psychotropic medications are frequently prescribed in routine clinical practice, particularly to address comorbid psychiatric disorders and specific symptom domains such as affective instability, impulsivity, anxiety, or sleep disturbances [28, 30]. At the same time, evidence supporting the efficacy of pharmacotherapy for the core symptoms of BPD remains limited and inconsistent, and several treatment guidelines caution against routine or excessive pharmacological treatment [4, 5, 8]. This creates a dilemma for clinicians, as medications may provide symptomatic relief for some patients while also exposing them to considerable side effects and polypharmacy. Given the high prevalence of psychiatric comorbidity in BPD, pharmacotherapy may often be clinically indicated; however, the potential benefits of medication must be carefully weighed against its risks and the possibility that adverse effects may negatively affect treatment engagement and overall outcomes. Limitations This study has several limitations. First, the small sample size and retrospective design limit the generalizability of the findings. As the data was obtained from routine procedures, we had no control over which questionnaires were administered or how much dropout occurred. Larger and more diverse samples are needed to explore the relationships between side effects, medication adherence, and treatment outcomes in more depth. The lack of a significant association between side effects and medication adherence may be due to the high medication adherence rates observed in our population. In the intensive multidisciplinary setting, patients may have continued their medication despite experiencing side effects. Individual perceptions of side effects can vary, and highly adherent patients may cope differently, potentially leading to an underestimation of the impact on medication adherence. Future research should include outpatient populations to better understand how medication adherence affects treatment outcomes in settings where patients face more challenges in maintaining medication regimens. Additionally, the reliance on self-report instruments for both side effects and medication adherence may have introduced bias. Future studies should use more objective measures of medication adherence, such as pharmacy refill records or electronic monitoring. Lastly, the cross-sectional design, with some data collected only at hospital admission, is a limitation. Follow-up data is essential to gain a more comprehensive understanding of medication adherence throughout multidisciplinary and long-term treatment. This would provide insights into how medication adherence evolves and would suggest strategies to sustain medication adherence post-treatment. Conclusions Despite the high prevalence and burden of side effects, medication adherence in this sample of BPD patients remained generally high. No significant associations were found between side effects and treatment outcomes via medication adherence. These findings suggest a complex relationship between medication side effects, medication adherence, and treatment outcomes in BPD, likely influenced by multiple factors. Given the limited literature on medication adherence in BPD, future research should explore additional variables that may impact medication adherence and outcomes, such as the therapeutic alliance, patient motivation, and the integration of psychotherapy with pharmacotherapy. While no significant associations emerged, this study underscores important considerations for clinical practice. The high prevalence of side effects, particularly fatigue, concentration difficulties, and mood changes, emphasizes the need for standardized monitoring and management. It also shows that pharmacotherapy in BPD should be used sparingly and with caution. Clinicians should only prescribe medication when necessary (such as for comorbid disorders or specific symptoms) and should maintain open communication about side effects and consider dosage adjustments or medication changes when necessary. Additionally, while medication adherence may be higher in inpatient settings, this may not translate to outpatient care, where support is limited. Interventions such as psychoeducation, medication reminders, and regular follow-ups may be crucial for improving medication adherence in outpatient settings. Acknowledgements We would like to thank Prim. Dr. Friedrich Riffer and Dr. Elmar Kaiser at University Hospital Eggenburg for providing the data used in our study. We would also like to thank Dr. Rachel Dale for her valuable insights and advice throughout the revision process. Abbreviations American psychological association Borderline personality disorder Borderline symptom list- short version Computer-based health evaluation system Questionnaire for the assessment of adverse effects of antiepileptic drugs Generalized anxiety disorder 7 Medication adherence report scale Patients health questionnaire depression Psychosomatisches zentrum waldviertel (= university hospital eggenburg) Selective serotonin reuptake inhibitors Author contributions SW and CO designed and conceptualized the study. CO assured the data access. SW conducted the data curation and data analyses. SW prepared the first draft of the manuscript. CO wrote the discussion section. CO thoroughly revised the manuscript and finalized it for submission. All authors have read and agreed to the published version of the manuscript. Funding We acknowledge support by Open Access Publishing Fund of Karl Landsteiner University of Health Sciences, Krems, Austria. Data availability The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request. Declarations Ethics approval and consent to participate The study retrospectively analyzed data collected within routine clinical care. All participants consented to the use of their data. The analysis was approved by the ethics commission of the Karl Landsteiner University of Health Sciences (Nr: 1045/2024) and complies with the ethical standards of the Declaration of Helsinki. Consent for publication Not applicable. Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Contributor Information Sarah Wazlawik, Email: sarah.wazlawik@donau-uni.ac.at. Claudia Oppenauer, Email: claudia.oppenauer@kl.ac.at. References Associated Data Data Availability Statement The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.