Socioenvironmental determinants of polypharmacy in mental health disorders: a systematic review Authors: Gobezie Temesgen Tegegne, Hossein Tabatabaei-Jafari, Theophile Niyonsenga, Saba Ghiasvand, Yassamin Bagheri, Nasser Bagheri Published: 2026 ABSTRACT BackgroundOver the past decades, the role of socioenvironmental determinants (SEDs) on shaping mental health outcomes has gained increasing attention. However, their relationship with antipsychotic and psychotropic polypharmacy (PP) remains unknown.AimThis systematic review aimed to synthesise evidence on the association between SEDs and PP, as well as summarising PP prevalence in patients with mental health disorders (MHDs).MethodA systematic search was conducted in PubMed, Web of Science, CINHAL, PsycINFO and Scopus databases. Literature was also sourced from Google, Google Scholar and the reference lists of included studies. The review included all studies reporting on sociodemographic and SEDs on psychotropic and antipsychotic PP in patients with MHDs. The prevalence rates were computed using medians and interquartile ranges (IQR), while factors associated with PP were summarised in a table.ResultsA total of 65 findings from 57 studies were analysed, 45 examining antipsychotic PP and 20 studied psychotropic PP. The median prevalences of psychotropic and antipsychotic PP were 30.5% (IQR: 18.0—39.5) and 28.1% (IQR: 18.7—42.9), respectively. Additionally, studies targeting inpatients, patients with schizophrenia and those utilising medical records had higher median prevalence rates. Moreover, our review found that family income to poverty ratios, family adverse consequences, Childhood Opportunity Index, social deprivation and area of residence (e.g. rural, metropolitan) were reported to be significant factors of PP in patients with MHDs.ConclusionThough the reported PP prevalence estimates varied, our systematic review showed that psychotropic and antipsychotic PPs are widespread in clinical practices. Additionally, a few SEDs were reported to be significantly associated with PP in patients with MHDs. This highlights the importance of investigating these modifiable factors. Hence, further in-depth exploration of the relationships between SEDs and polypharmacy is required. Evidence generated through such studies informs population-level strategies to reduce PP, improve equity in mental healthcare, and ultimately enhance patient outcomes.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12889-026-27658-w. Introduction Mental health disorders are among the top ten chronic conditions affecting over 11% of the world’s population [1]. MHDs result in significant morbidity, increased disability-adjusted life years, higher mortality rates, and have negative impacts on the global economy [2–4]. Moreover, MHDs worsen the prognosis of chronic diseases, such as cardiovascular disease and diabetes [5]. The management of MHDs is complex, involving multifaceted approaches integrating both pharmacotherapeutic and non-pharmacotherapeutic interventions. Pharmacotherapy is the cornerstone of treatment to optimise patients’ clinical, psychosocial and behavioural outcomes. However, higher rates of comorbidity [6, 7], relapse, poor treatment outcomes and adverse drug events [8–12] in patients with MHDs complicate pharmacotherapy utilisation, leading to multiple medication use. The simultaneous use of multiple medications is commonly referred to as polypharmacy though it lacks a standardised cut-off point to define. PP is a commonly seen in patients with chronic diseases, multimorbidity, and in the aging population [13]. However, this practice is significantly associated with inappropriate medication use, increased possibility of drug interactions and adverse events [14]. Such events then result in an increased rate of hospitalisation [15], higher healthcare costs [16] and mortality [17]. Complications related to PP can also contribute to depression [18], dementia [18], reduced quality of life [19] and cognitive impairment for patients [20]. These consequences are detrimental particularly for those impacted by MHDs as they exacerbate psychiatric symptoms, and impose additional burden. Several factors have been shown to have a significant association with medication use and thereby affect PP; such as, age, gender, education, severity of illness, comorbidity, length of hospital stay and previous hospitalisation [21–24]. However, the association between socioenvironmental determinants and PP remains poorly understood. The SEDs consist of circumstances, where people live, work and age, including neighbourhood and built environmental conditions, such as deprivation, economic stability, education access, healthcare access and other social and community contexts [25]. These modifiable factors are proven to be significantly associated with population health outcomes and overall well-being by creating disparities and inequities [26–29]. Hence, understanding their relationships with PP facilitates the design of targeted population-level interventions, and enables policymakers to make informed decisions on optimal resource distribution. Aim This systematic review, therefore, aims to analyse the existing body of knowledge on the relationships between SEDs and PP, as well as estimating the prevalence of antipsychotic and psychotropic PP among patients with MHDs. Method This systematic review followed the Preferred Reporting Items for Systematic Review and Meta‐Analysis guidelines (PRISMA) [30], and the protocol was registered on PROSPERO (CRD-42024585832). (supplementary file_1). Databases and search strategy The search employed PubMed, Web of Science, CINHAL, PsycINFO and Scopus electronic databases from inception to the end of August 2025. In addition, Google, Google Scholar and the reference list of included studies were explored to ensure all scientific literature available is captured. Two systematic search strategies were developed based on four core concepts: social factors, built environmental factors, mental disorders and polypharmacy. For each concept, a comprehensive list of database-specific terms and relevant keywords was prepared. The search strings used were: ‘Social factors AND Polypharmacy AND Mental disorder’ and ‘Environmental factors AND Polypharmacy AND Mental disorder’. Boolean operators ‘OR’ and ‘AND’ were also applied to refine our search. An example of search strategy used in PubMed was as follows: (mental disorders [MeSH] OR mental illness [keyword]) AND (Social Determinants of Health [MeSH] OR Geography, Medical [MeSH] OR social determinant [keyword] OR environmental determinant [key word]) AND (Polypharmacy [MeSH] OR poly medication [key word]). The detail of our search strategy is shown in the supplementary file (supplementary file_2). All retrieved articles were imported into Covidence software for systematic deduplication, title/abstract screening and full-text assessment. Eligibility criteria Inclusion criteria All studies reporting on the relationship between sociodemographic, SEDs, and PP in patients with MHDs were included. The SEDs include neighbourhood social and built-environmental factors, such as social deprivation, social fragmentation, living area, availability of green space and others. Moreover, MHDs were considered as psychiatric disorders based on the World Health Organization (WHO) definition [31], whilst psychotropic and antipsychotic PP were considered for the PP definition. Exclusion criteria Studies reporting on PP in neurological and developmental disorders, PP not explicitly defined as either psychotropic or antipsychotic PP, PP and patient outcomes, and prescribers’ issues in relation to prescription patterns were excluded. All peer-reviewed studies published in English language and conducted in human population fulfilling the above-mentioned criteria qualified for inclusion. Systematic reviews, academic reviews, case reports, case series, abstracts with no full information, studies not published in English, letters to the editors and opinion articles were excluded. Article screening and assessment Title and abstract screening were independently conducted by two reviewers (G.T. and S.G.). Articles marked ‘Yes’ or ‘Maybe’ by both reviewers were considered for full-text review, while those marked as ‘No’ were excluded. Disagreements were resolved by consensus of the reviewers. The same reviewers also independently assessed the full-text article against the predefined eligibility criteria. Studies agreed upon for inclusion were retained for the systematic review. Those excluded at this stage were documented along with the reasons for exclusion. Disagreements were resolved through discussion between the two reviewers. Data extraction and study quality assessment A data extraction template was developed using Microsoft Excel, which was modified following pilot study. The following information was extracted from each study: authors’ name, year of publication, country of study, study population, study design, healthcare setting, definition and prevalence of PP and contributing factors (Supplementary file_3). In addition, the internal validity of included studies was assessed using the Joanna Briggs Institute (JBI) critical appraisal checklists [32]. The checklists include eight questions for cross-sectional study design and eleven questions for cohort study design to assess selection of the study participants, exposures status, confounders, biases, outcomes and statistical analysis. Each question has “Yes” (scored as 1), and “No,” “Unclear,” or “Not applicable” (scored as 0) responses. Then, the overall study’s risk of bias was assessed as low (≥ 70%), moderate (50—69%) or high (≤ 49%). The score is calculated by dividing the total number of ‘Yes’ responses by the total number of items. As JBI has no scoring system, scores were based on previously published study [33] (Supplementary file_4). Data analysis The data was analysed using STATA version 18. A meta-analysis with a random-effect model using the Freeman-Tukey technique was planned to estimate the prevalence of psychotropic and antipsychotic PP. A subgroup analysis was also planned based on study characteristics (e.g. study design, definition of PP, etc.). However, due to the substantial variation observed among included studies (for psychotropic PP: pooled effect size (29.0%), I2 = 99.98%, τ2 = 0.23, p < 0.001), and for antipsychotic PP: pooled effect size (28.0%), I2 = 99.98%, τ2 = 0.15), the meta-analysis could not be computed. The protocol was amended accordingly. Consequently, the overall psychotropic and antipsychotic PP estimates were summarised using median and interquartile ranges (IQR). In addition, studies were categorised and discussed according to their study method, study participants and definition they employed. Further, factors associated with PP were summarised in a table. Results Study selection A total of 5,631 articles were identified, and 963 duplicates were removed. Of the remaining articles, 4,532 and 79 were excluded during title and abstract screening and full-text review for not meeting the eligibility criteria, respectively. Finally, 57 unique studies were included in our systematic review (Fig. 1).Fig. 1PRISMA flow diagram PRISMA flow diagram Study characteristics Our review found 65 study findings from 57 unique studies. Four studies 4 [24, 34–36] reported more than one finding. 45 of the findings focused on antipsychotic PP (n = 39 studies), while the remaining 20 addressed psychotropic PP (n = 18 studies). Two studies reported both psychotropic and antipsychotic PP issues [36, 37]. All included studies, except for Sohn M et al. [38], considered sample size as the number of individual participants. Sohn M et al., however, considered the total number of patient visits. Considering the number of visits, our review included a total of 19, 987, 049 observations, ranging from 120 [39] to 16,131,721 [38]. Studies were published between 2004 and 2024, with the highest number of studies (n = 6) published in 2018. The majority of studies (n = 53) were conducted within a single country, spanning 28 different countries. The remaining four studies [40–43] were conducted in more than one country. One focused on European countries [43], while the remaining studies were conducted in various Asian countries [40–42]. The United States accounted for the highest number of studies (n = 20), followed by Italy (n = 4) (Fig. 2).Fig. 2Distribution of psychotropic and antipsychotic PP estimates globally Distribution of psychotropic and antipsychotic PP estimates globally The majority of included studies (n = 48) used a cross-sectional design. Three studies [43–45] employed a prospective cohort design, and six remaining studies [34, 46–50] utilised a retrospective cohort design. Over half of the included studies (n = 31) targeted an adult population, while ten studies focused on paediatric population and three studies on the elderly population. The remaining 13 studies included participants across all age groups. Moreover, 30 studies focused on a single type of MHD, while the remaining studies investigated comorbid MHDs. Schizophrenia was the most frequently studied (n = 22) MHD, followed by bipolar disorder (n = 4) [37, 44, 51, 52] and depressive disorder (n = 3) [23, 45, 53]. In terms of study setting, 21 and 12 studies were carried out in the inpatient and outpatient settings, respectively. The remaining studies were conducted in mixed and non-specified settings. Moreover, 36 studies utilised medical records, while 22 relied on datasets, and the remaining studies collected data prospectively through interview and patients’ medication containers [45, 54]. Study quality assessment using JBI checklists All included studies were within moderate to low risk of bias. More than half of the cross-sectional studies (n = 34) were rated as low-risk, and the remaining had moderate-risk of bias. Most of the cross-sectional-based studies did not explicitly define confounders even though they computed logistic regression. Considering cohort studies, three were assessed as low risk, while the rest had a moderate risk of bias. Similar to the cross-sectional studies, though regression analysis was computed, four studies did not clearly define their confounders. Additionally, the outcome status of the study participants at the beginning of the study was not clearly reported in five studies (Supplementary file_4). Prevalence of psychotropic PP A total of 18 unique studies [34–37, 39, 44, 45, 52, 54–63] reported 20 study findings about psychotropic PP. Two of them reported more than one study findings; Medhekar et al. [35] reported two findings on psychiatric patients who were treated by a single prescriber and those treated by multiple prescribers. Meanwhile, Lohr et al.’ s [34] study found two reports on patients with mood and bipolar disorders. Most of the studies (n = 12) were conducted on patients with comorbid MHDs, while others targeted specific MHD type, such as post-traumatic stress disorder [60], bipolar [37, 44, 52, 62] and depression [45]. Moreover, the included studies employed different study design, population and PP definitions. Considering their definition, studies defined psychotropic PP depending on the number medications and duration of patient-exposure to these medications. Half of the studies defined psychotropic PP as the concurrent use of two or more psychotropic medications, while four studies each operationalised their definition as the use of three or more, or four or more different types of psychotropic medications over a different period, respectively. One study [63] used an atypical definition, defining psychotropic PP as the use of more than the median number of prescribed psychotropic medications. Furthermore, two studies [34, 36] with three findings investigated factors of psychotropic PP without reporting prevalence estimates. Hence, three studies [34, 36, 63] mentioned above were excluded from median prevalence computations. Based on 17 prevalence estimates, the overall median prevalence of psychotropic PP was 30.5% (IQR: 18.0—39.5) with the prevalence ranging from 2.7% [54] to 79.9% [60]. The prevalence rates varied depending on the study’s method, setting and population, as well as psychotropic PP definition used. Studies conducted using a cross-sectional design showed a higher median prevalence (31.1%, IQR: 18.6—39.3) than those using cohort design (21.0%, IQR: 18.5—30.3). Two cohort studies defined psychotropic PP using higher thresholds (≥ 3 and ≥ 4 psychotropic medications) [44, 45], and another cohort study targeted paediatric population [34]. These factors might have contributed to the lower prevalence estimates observed among cohort studies. Regarding study participants, a higher median prevalence rate (35.6%, IQR: 26.6—43.7) was found among all age-targeted studies than studies conducted on paediatric patients only (23.6%, IQR: 9.6—30.8). In addition, studies focused on patients with one type of MHD showed a 35.6% (IQR: 20.7—55.8) median proportion, which is higher than studies investigating comorbid MHDs (29.9%, IQR: 13.0—32.0). Most studies investigating comorbid MHDs utilised datasets which potentially reduce the reported prevalence. Depending on study setting and source of data, those studies utilising medical records as source of data (32.1%, IQR:29.2—40.5) and those conducted at the inpatient settings (38.09%, IQR: 33.0—63.4) showed higher rates than studies conducted using datasets (24.2%, IQR: 13.0—37.2), and at mixed settings (29.2%, IQR: 18.0—31.7). Furthermore, varying levels of PP were found among studies employing different psychotropic PP definitions. Studies defined psychotropic PP as the concurrent use of two or more psychotropic medications showed a higher median prevalence (31.7%, IQR:29.2—46.9) followed by those considering three or more (28.3%, IQR: 20.8—36.8) and four or more (10.2%, IQR: 3.9—20.0) of these medications, respectively. Prevalence of antipsychotic PP A total of 39 unique studies reported 45 study findings on antipsychotic PP. One study reported more than one findings [24] on patients with schizophrenia, depression, bipolar and personality disorders. More than half of the studies (n = 22) focused only on schizophrenia, two on [64, 65] severe mental illness, one on bipolar disorder [37], and the remaining studies were on comorbid MHDs. Among the included studies, six findings by four studies [24, 36, 37, 50] reported only factors associated with antipsychotic PP without mentioning prevalence estimates. Unlike that of psychotropic PP, studies in this category used a relatively similar definition of antipsychotic PP, concurrent use of two or more antipsychotics, except for different periods of medication exposure. The overall median prevalence of antipsychotic PP was 28.1% (IQR: 18.7—42.9) based on 39 prevalence estimates. The reported prevalence ranged from 1.7% [47] to 70.5% [66]. Similarly with psychotropic PP, rates of antipsychotic PP utilisation varied depending on methodology used, study setting and participants, and how PP was defined. Based on study setting, those conducted within inpatient setting showed a higher median prevalence (39.3%, IQR: 29.0—45.2) than those done on mixed (24.2%, IQR: 22.3—34.9) or outpatient settings (21.3%, IQR: 17.8—28.2). Method-wise, cross-sectional studies reported a higher median prevalence (28.2%, IQR: 19.2—43.6) than studies using a cohort design (22.3%, IQR: 13.0—23.0). Three retrospective cohort studies [47–49] employed datasets which might lower prevalence reports. Regarding study population, studies targeting adult patients (28.2%, IQR: 21.7—45.2) found a higher prevalence compared to those conducted on paediatric patients (7.5%, IQR: 6.3—19.1). Also, studies targeting patients with Schizophrenia only (30.1%, IQR:22.8—42.9) revealed a higher median prevalence than those conducted on patients with other MHDs (21.7%, IQR: 12.5—30.7). Furthermore, medical record-based studies (30.7%, IQR: 22.7—44.1) found a higher median prevalence compared to dataset-based studies (22.0%, IQR: 12.6—25.5). Those studies defined antipsychotic PP as the simultaneous use of two or more antipsychotics for at most two weeks (42.8%, IQR: 31.4—43.2) also reported a higher proportion, followed by those studies that did not specify duration of exposure to medication (28.2%, IQR:18.2—38.1) and studies using a duration of at least two weeks (23.0%, IQR:16.0—31.4). Socioenvironmental and sociodemographic determinants of psychotropic and antipsychotic PP Most studies reported that age, gender, marital status, race, having insurance and employment were significantly associated with psychotropic PP. Additionally, studies identified that lower neighbourhood Childhood Opportunity Index, family adverse consequences and lower proportion of family income to poverty were significantly associated with psychotropic PP. Moreover, studies reported that patients living in metropolitan areas, capital or central cities had a higher risk of PP in comparison to their counterparts (Table 1).Table 1Sociodemographic and socio-environmental determinants of polypharmacyVariablesDeterminants of psychotropic PPDeterminants of antipsychotic PPSocio-demographicsAge [34–37, 52, 55–58, 60, 61]Gender [34–36, 44, 56, 57, 62, 63, 67]Marital status [52, 55, 63]Race [34–36, 44, 56, 58, 59, 61, 63, 68]Age [38, 40, 42, 47, 48, 50, 65, 69–74]Gender [24, 37, 46, 48–51, 65, 66, 69, 73, 75–78]Race [38, 48, 50, 71, 73, 74]Marital status [24, 71]Lonely [43, 69]Relationship status [47]Education [24]Socioeconomic statusMedical insurance [36, 56, 61, 67]Employment [59]Medicare eligibility [58, 68]Employment [24, 79]Neighbourhood statusNeighbourhood Childhood Opportunity Index [61]Family adverse experiences [54]Ratio of family income to poverty [56]Social deprivation [47]Living areasLiving in Metropolitan or non-metro areas [34]Living in different regions of a country [58]Living in urban or non-urban or rural areas [24, 48]Living in different countries [40, 41, 70] Table 1Sociodemographic and socio-environmental determinants of polypharmacyVariablesDeterminants of psychotropic PPDeterminants of antipsychotic PPSocio-demographicsAge [34–37, 52, 55–58, 60, 61]Gender [34–36, 44, 56, 57, 62, 63, 67]Marital status [52, 55, 63]Race [34–36, 44, 56, 58, 59, 61, 63, 68]Age [38, 40, 42, 47, 48, 50, 65, 69–74]Gender [24, 37, 46, 48–51, 65, 66, 69, 73, 75–78]Race [38, 48, 50, 71, 73, 74]Marital status [24, 71]Lonely [43, 69]Relationship status [47]Education [24]Socioeconomic statusMedical insurance [36, 56, 61, 67]Employment [59]Medicare eligibility [58, 68]Employment [24, 79]Neighbourhood statusNeighbourhood Childhood Opportunity Index [61]Family adverse experiences [54]Ratio of family income to poverty [56]Social deprivation [47]Living areasLiving in Metropolitan or non-metro areas [34]Living in different regions of a country [58]Living in urban or non-urban or rural areas [24, 48]Living in different countries [40, 41, 70] Sociodemographic and socio-environmental determinants of polypharmacy Age [34–37, 52, 55–58, 60, 61] Gender [34–36, 44, 56, 57, 62, 63, 67] Marital status [52, 55, 63] Race [34–36, 44, 56, 58, 59, 61, 63, 68] Age [38, 40, 42, 47, 48, 50, 65, 69–74] Gender [24, 37, 46, 48–51, 65, 66, 69, 73, 75–78] Race [38, 48, 50, 71, 73, 74] Marital status [24, 71] Lonely [43, 69] Relationship status [47] Education [24] Medical insurance [36, 56, 61, 67] Employment [59] Medicare eligibility [58, 68] Neighbourhood Childhood Opportunity Index [61] Family adverse experiences [54] Ratio of family income to poverty [56] Living in Metropolitan or non-metro areas [34] Living in different regions of a country [58] Living in urban or non-urban or rural areas [24, 48] Living in different countries [40, 41, 70] Regarding antipsychotic PP, age, gender, race, relationship status, education and employment were the frequently reported significant determinants. Most importantly, studies identified that deprivation level and area of residence (e.g. rural vs urban, Korea and Singapore) were significantly associated with antipsychotic PP (Table 1). Discussion This review aimed to synthesise evidence on the relationship between SEDs and PP, and estimating PP prevalences in patients with MHDs. To our knowledge, it is the first comprehensive systematic review addressing the association between SEDs and polypharmacy in MHDs. Considering heterogeneity across studies, the median prevalences of psychotropic and antipsychotic PP among patients with MHDs were 30.5% and 28.1%, respectively. The higher median psychotropic PP prevalence reflects the common practice of prescribing multiple medication classes to manage complex symptoms and/or comorbidities. It should be noted that actual PP prevalence is potentially higher than the above estimates due to the global rise in psychotropic medication consumption [80] which is driven by the growing number of people impacted by MHDs and the chronic nature of these diseases. Additionally, the increasing incidence of chronic illnesses such as diabetes and cardiovascular disease, which have a bidirectional relationship with MHDs, may further amplify the extent of PP [81, 82]. Furthermore, slow progress in developing safer and more effective medications [83] leads patients to use older drugs with lower efficacy and higher adverse effect profiles. These contribute to suboptimal treatment response which results in increased use of multiple medications [84]. Furthermore, the finding of our review is slightly lower than the pooled estimate among general population (37%, 95% CI: 31–43) [85]. Nevertheless, this finding shows that patients with MHDs experience a substantial burden of PP. This highlights the need for in-depth investigation of PP in this group of patients. Our review found that the prevalence estimates of both psychotropic and antipsychotic PP varied by study population, setting, data source and definition of polypharmacy used. Higher median prevalence rate in the inpatient setting is likely due to the complexity and severity of illness in this environment often necessitates the use of multiple medications [12]. Similarly, studies involving adult patients had a higher prevalence of PP, owing to increased chance of comorbidities and age-related decline in organ function. These factors expose this population to poor treatment outcomes which can increase the likelihood of PP [86]. Moreover, a relatively complete documentation of medication history in medical records could contribute to higher rates of PP than using administrative datasets. Further, the type and number of psychotropic medication use, as well as period of exposure affects prevalence estimates in patients with MHDs. Every disease state involves complex interactions of biological, psychological and socioenvironmental factors [87]. Consequently, medical care alone cannot adequately manage patients’ overall well-being, unless it is complemented by interventions addressing SEDs. Disparities in these factors significantly impact health outcomes [88], which can be closely linked to medication consumption. Therefore, it is essential to deepen our understanding of SEDs and their impact on medication consumption patterns. Our finding showed that some studies reported social factors which were significantly associated with PP. Social deprivation, family adverse consequences and low family-income to poverty ratio were significantly associated with antipsychotic PP [47]. Areas with high deprivation and low economic conditions usually have inadequate access to healthcare resources and professionals. This might affect early treatment initiation, and most patients seek professional help when their condition is more advance, resulting in the need for use of multiple medications [89]. Similarly, the lower opportunity of accessing treatment facility among patients with lower Childhood Opportunity Index scores [61] could contribute to higher rates of polypharmacy. However, association of other key social factors such as neighbourhood level social fragmentation and social vulnerability are yet to be explored. Although the relationship between built environmental factors and PP was not explicitly described across the included studies, some studies reported a significant relationship between broader built environmental contexts and PP. These include living in rural versus urban, metropolitan versus non-metropolitan regions, and different countries or geographical locations (e.g., north, east, west, south). However, interpreting these findings could be misleading as these parameters do not clearly explain the availability, accessibility and quality of built environmental features, which are critical factors affecting the prognosis of MHDs [90]. Future study is warranted to see the relationship between PP and specific built environmental factors, such as access to green and blue spaces, air pollution, parking and others. As most SEDs are modifiable, future population-level interventions should target such important area-level factors to enhance patients' overall well-being [91, 92]. With ongoing urbanisation, integrating these factors into urban planning and health policy is essential to fostering healthier and more productive population. This underscores the critical roles of health-promoting environments that can reduce reliance on medications. Hence, future studies should adopt an ecological modelling to thoroughly examine the impact of SEDs on medication utilisation. This perspective is essential to develop comprehensive and context-specific interventions to reduce PP. This systematic review has several strengths and a few limitations. To our knowledge, this is the first comprehensive systematic review involving a large number of studies and study participants, without publication year restriction, addressing the socioenvironmental drivers of PP in MHDs. Additionally, the review underwent a systematic search across multiple databases to incorporate as many studies as possible aiming to assess the effect of socioenvironmental determinants. However, our systematic review has also few limitations. First, the review was limited to studies published in English and did not include grey literature. Second, substantial variability across studies in PP definition, study population, data source and study settings limits to draw broad conclusion. Hence, we recommend caution when interpreting the findings. Third, the review did not evaluate the clinical appropriateness of polypharmacy. Conclusion Despite the diversity of included studies and varying reports of PP prevalence estimates, our review found that PP is a common clinical practice challenge. This review also identified that some SEDs were reported to have a significant association with PP in patients with MHDs. However, in-depth investigation is required to fully establish such association. Generating such evidence is crucial for developing feasible and context-based intervention strategies that can guide stakeholders, including policy makers in making informed decisions. Supplementary Information Supplementary Material 1. Supplementary Material 2. Supplementary Material 3. Supplementary Material 4. Supplementary Material 1. Supplementary Material 2. Supplementary Material 3. Supplementary Material 4.