The research question was formulated using the PICO framework (Population, Intervention/Exposure, Comparison, Outcome) as presented in Table 1.

A comprehensive literature search was conducted across nine electronic databases: Taylor & Francis, Scopus, ClinicalKey, ProQuest, Cochrane Library, ScienceDirect, PubMed, EBSCO, and AccessMedicine. The search terms included: “Type 2 Diabetes Mellitus” OR T2DM OR “Diabetes Mellitus”) AND (“All-Cause Mortality” OR Mortality OR “Death Rate” OR “Survival Rate”) AND (“Risk Factor*” OR Predictor* OR “Associated Factor*” OR Correlate* OR “Hazard Ratio” OR “Relative Risk”) AND (“Mental Health” OR “Mental Disorder*” OR Depression OR Anxiety OR “Psychological Distress” OR Stress OR “Psychiatric Condition*” OR “Psychological Factor*”. The search was performed using Boolean operators (AND, OR) to combine the search terms. Additional filters were applied to restrict the results to articles published between January 2020 and July 2025, written in English, and available in full text.

1.

Original research articles (prospective/retrospective cohort studies, case-control studies, or cross-sectional studies).

1.

Studies focused on type 1 diabetes mellitus or gestational diabetes.

The literature selection process followed the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines, consisting of four stages: 1.

Identification: The initial search across six electronic databases yielded a total of 3,494 records (Scopus: 1,384; ProQuest: 721; Cochrane Library: 295; ScienceDirect: 469; PubMed: 447; EBSCO: 178). A total of 302 duplicate records were removed prior to screening.

The PRISMA flow diagram illustrating the study selection process is presented in Figure 1.

Data were independently extracted by three reviewers using a standardized form that included: 1.

Study characteristics: author, year of publication, title, study design, and study setting.

Any discrepancies in data extraction were resolved through discussion among all reviewers.

Due to heterogeneity in study design, populations, and outcome measures, a narrative synthesis was performed. Findings were categorized based on types of mental health disorders (e.g., depression, anxiety, severe mental illness, suicidal behavior) and mortality outcomes (all-cause, cardiovascular, and suicide-related mortality). Effect estimates (HR, OR, RR) from individual studies were summarized in the study characteristics table ( Table 2). Subgroup patterns were described based on sex, age group, and type of mental health condition.

The assessment using the Joanna Briggs Institute (JBI) Critical Appraisal Tools indicates that most of the included studies were at low risk of bias, particularly those employing cohort designs ( Table 3). These studies demonstrated several methodological strengths, including valid and reliable measurement of both exposure and outcomes, the use of appropriate statistical analyses, and adequate identification and control of potential confounding factors. However, some limitations were identified across the studies, such as incomplete follow-up, insufficient reporting on the handling of missing data, and the potential for misclassification of exposure, particularly in studies relying on self-reported or administrative data. In addition, studies with cross-sectional designs exhibited a moderate risk of bias due to their inherent limitations in establishing causal relationships and controlling confounding variables. Despite these limitations, the overall body of evidence can be considered methodologically robust and provides consistent findings.

Depression consistently emerged as an independent predictor of all-cause mortality in individuals with T2DM. Study Salinero demonstrated a 40% increased risk of mortality (HR 1.40) in patients with baseline depression ( Salinero-Fort, Gómez-Campelo and Cárdenas-Valladolid, 2021). Study Chen, in a large-scale cohort, observed the highest mortality risk among patients with both T2DM and depression (HR 2.47), compared to those with T2DM alone (HR 1.95) or depression alone (HR 1.23) ( Chen, Wang and Hsieh, 2022). Prigge confirmed these findings, reporting an HR of 2.16 for all-cause mortality. ( Feng et al., 2023) Identified an elevated risk (HR up to 1.67) associated with moderate-to-severe depression, particularly in men and older adults. Most recently, Cai reported a hazard ratio (HR) of 2.09, indicating a synergistic effect on mortality for patients with comorbid depression and T2DM ( J Cai et al., 2024b). Taken together, the evidence underscores that depression substantially amplifies mortality risk in T2DM, primarily through its adverse impact on disease management and underlying biological mechanisms, including insulin resistance.

Studies investigating SMI, such as schizophrenia and bipolar disorder, consistently report an elevated mortality risk. Han found that patients with T2DM and SMI exhibited a 1.92-fold increased risk of all-cause mortality and a 2.24-fold increased risk of cardiovascular mortality compared to those without SMI ( Han et al., 2021). Similarly, Bakkedal reported a twofold higher mortality risk in individuals with schizophrenia compared to a non-SMI cohort, despite comparable diabetes management, suggesting potential undertreatment of cardiovascular conditions ( Bakkedal et al., 2024). Alba demonstrated that substance use disorders (OR 2.18) and schizophrenia (OR 1.82) significantly elevated the 4-year mortality risk, highlighting the profound impact of SMI on both all-cause and cause-specific mortality outcomes ( Guerrero Fernández de Alba et al., 2020).

Two studies specifically addressed the risk of suicide among patients with T2DM and comorbid mental disorders. In a nationwide cohort of 875.671 individuals, Baek reported the highest suicide risk among patients with bipolar disorder (hazard ratio [HR] 3.00), followed by schizophrenia (HR 2.51) and depression (HR 1.92) ( Baek, Lee and Kang, 2025). In Ethiopia, Gebreselassie documented a prevalence of suicidal behavior of 30.8% among T2DM patients, with 15.8% reporting suicidal ideation, 14.4% attempting suicide, and 15.1% having a suicide plan ( Gebreselassie, Metekiya and Gebrehiwot, 2020). These findings underscore the substantial burden of suicide risk in T2DM patients with mental disorders, particularly those with severe mental illness, and highlight the urgent need for targeted mental health interventions.

This review emphasizes a mortality gradient associated with the comorbidity burden of mental health disorders. Kim reported that patients with T2DM and a single mental disorder had a hazard ratio (HR) for mortality of 1.27, which rose to 1.57 in those with two or more disorders ( Kim, Kim and Cho, 2024). Ho et al further demonstrated that depression increased mortality risk in T2DM patients, even among individuals with a low complication burden ( Ho, Chan and Lo, 2025). Similarly, Alba observed that substance use disorders and schizophrenia significantly elevated four-year mortality risk. Taken together, these findings suggest that multiple mental disorders exacerbate mortality risk, most likely through cumulative effects on disease management and limited access to healthcare services ( Guerrero Fernández de Alba et al., 2020).

Several studies have highlighted disparities in healthcare provision for patients with T2DM and comorbid mental disorders. Study Bakkedal and Ho reported that patients with SMI were less likely to receive cardiovascular medications despite comparable levels of clinical monitoring, suggesting potential undertreatment ( Baek, Lee and Kang, 2025; Ho, Chan and Lo, 2025). Study Han observed higher rates of emergency hospitalizations but fewer elective admissions for cardiac conditions among individuals with SMI, indicating gaps in proactive care ( Han et al., 2021). These findings underscore the need for integrated care models that address both physical and mental health requirements in the management of T2DM.

This review identified 15 studies published between 2020 and 2025, encompassing a broad geographical scope, ranging from high-income countries (United Kingdom, United States, Italy, Denmark) to middle-income countries (China, South Korea, Taiwan, Ethiopia). This regional diversity provides a more representative global perspective, in contrast to prior reviews that were predominantly limited to developed nations ( Farooqi et al., 2019; Wu, Hsu and Wang, 2020). The majority of studies employed prospective or retrospective cohort designs, which are robust for assessing temporal relationships, while a smaller subset utilized cross-sectional or nested case-control designs, which, despite certain limitations, contributed to the diversity of evidence ( Han et al., 2021). Sample sizes varied widely, ranging from hundreds of participants to large-scale population studies exceeding 875.000 individuals ( Chen, Wang and Hsieh, 2022), enhancing the validity of findings while highlighting contextual heterogeneity. Mental disorders investigated included depression, anxiety, schizophrenia, bipolar disorder, and substance use disorders, with diagnoses based on ICD codes or validated instruments such as the PHQ-9 and the HADS ( Salinero-Fort, Gómez-Campelo and Cárdenas-Valladolid, 2021; J Cai et al., 2024b). The primary outcomes were all-cause mortality, cardiovascular mortality, and suicide, predominantly measured using national registries ( Feng et al., 2023). Heterogeneity persisted due to variations in diagnostic methods and data sources; however, the consistent pattern of increased risk across studies underscores the robustness of these findings ( van der Feltz-Cornelis et al., 2021). A novel contribution of this review is the application of the PRISMA methodology with multiple databases (Scopus, ProQuest, Cochrane Library, ScienceDirect, PubMed, EBSCO), resulting in a more rigorous study selection process compared to prior reviews limited to a single database ( Farooqi et al., 2019).

Depression consistently emerges as the mental disorder most strongly associated with increased all-cause mortality in patients with T2DM. Large cohort studies report hazard ratios (HRs) ranging from 1.40 to 2.47, confirming depression as an independent risk factor for premature mortality ( Salinero-Fort, Gómez-Campelo and Cárdenas-Valladolid, 2021; Chen, Wang and Hsieh, 2022; J Cai et al., 2024b). These findings align with the meta-analysis by ( Wu, Hsu and Wang, 2020) but indicate a higher risk compared to the review by Farooqi, which was limited to South Asia. This discrepancy may reflect variations in diagnostic tools, cultural differences, and access to mental health services ( Farooqi et al., 2019). Biological mechanisms include insulin resistance, chronic inflammation, and glucose dysregulation, while behavioural factors, such as poor treatment adherence and inadequate self-care, exacerbate prognosis in depressed patients ( van der Feltz-Cornelis et al., 2021).

However, variations by gender and age are evident. Feng and Ho indicate that men with T2DM and depression face a higher mortality risk, particularly from cardiovascular causes, whereas reported a higher prevalence of depression in women, while found that T2DM onset before age 40 increases relative mortality risk nearly threefold compared to later-onset diagnoses ( Mansour, Golden and Yeh, 2020; Khubchandani et al., 2023; Ho, Chan and Lo, 2025). These variations underscore the complex interplay of depression with biological, social, and demographic factors ( Culman et al., 2013). Although the mortality risk associated with depression is lower than that of SMI, its population-level impact remains substantial due to its significantly higher prevalence among T2DM patients. Thus, depression worsens clinical control and significantly contributes to mortality at the population level ( Han et al., 2021; Baek, Lee and Kang, 2025).

Studies investigating SMI, such as schizophrenia and bipolar disorder, exert a more pronounced impact on mortality in patients with T2DM compared to depression. A cohort study by Han reported a 92% increase in all-cause mortality and a 124% increase in cardiovascular mortality among patients with both T2DM and SMI ( Han et al., 2021). These findings are consistent with Bakkedal, who noted that, despite declining mortality trends in the general population, patients with schizophrenia or bipolar disorder continue to exhibit elevated mortality rates. further identified a risk gradient, with the highest cardiovascular mortality hazard ratio (HR) for schizophrenia (2.38), followed by bipolar disorder (1.70) and major depression (1.84).

These results underscore that SMI not only worsens the prognosis of T2DM but also substantially increases the risk of cause-specific cardiovascular mortality. The underlying mechanisms are multifactorial: second-generation antipsychotics (e.g., clozapine, olanzapine) have been shown to exacerbate insulin resistance and dyslipidemia, while systemic barriers, such as delayed diagnosis of physical illnesses and limited access to cardiovascular care, further aggravate outcomes ( Mitchell, Malone and Doebbeling, 2009; Baldessarini, Innamorati and Erbuto, 2017; Campbell and Campbell, 2020; Knudsen, Scheuer and Diaz, 2023).

SMI also exhibits risk variations based on demographic characteristics. Men with T2DM and SMI face higher mortality rates than women, particularly for cardiovascular causes, whereas women are more frequently diagnosed with depression but do not exhibit a comparable increase in relative risk (Feng et al., 2023b; Ho, Chan and Lo, 2025). Regarding age, elderly patients with SMI show higher absolute mortality rates, while those with early-onset T2DM face a greater relative risk, indicating a strong interaction between age and disorder type ( Mansour, Golden and Yeh, 2020). Additionally, comparisons across disorder types reveal a risk gradient, with schizophrenia consistently showing the highest HR, followed by bipolar disorder and major depression ( Han et al., 2021; Fleetwood et al., 2023). Thus, variations by age, gender, and disorder type highlight the heterogeneous nature of mortality risk in T2DM patients with SMI, necessitating stratified approaches in clinical practice.

Suicide risk represents a critical yet often overlooked outcome in patients with T2DM and comorbid mental disorders. A large cohort study by Baek Hs, reported the highest suicide risk in patients with schizophrenia (HR 3.24), followed by bipolar disorder (HR 2.47), depression (HR 2.08), and anxiety (HR 1.63) ( Baek, Lee and Kang, 2025). Study Kim reinforced these findings, demonstrating that diabetic patients are more likely than the general population to experience suicidal ideation, plan suicide, and attempt suicide ( Kim et al., 2022b).

Biological mechanisms linking these conditions involve activation of the hypothalamic-pituitary-adrenal (HPA) axis in chronic depression, which triggers norepinephrine secretion, exacerbates insulin resistance, and heightens susceptibility to suicidal ideation ( Detka et al., 2013; Chung et al., 2014; Kim, Um and Kim, 2022b). Social factors, including advanced age, low socioeconomic status, insulin use, and mental health stigma, further elevate risk ( Lofman et al., 2012; Kim, Um and Kim, 2022b). Several studies also confirm the mediating role of depression, whereby T2DM indirectly increases suicide risk through depressive symptoms ( Myers et al., 2013; Sher, 2022). This review provides a novel contribution by highlighting that suicide risk in T2DM patients is not solely associated with depression but is also significant in schizophrenia and bipolar disorder, with higher HRs than depression. This expands on prior evidence that primarily focused on depression as the main determinant ( Sartorius, 2021; Sher, 2023).

Subgroup analyses reveal variations in suicide risk by disorder type, gender, and age. The highest risks are observed in patients with bipolar disorder and schizophrenia, followed by depression and anxiety, indicating that SMI imposes a greater risk burden than depression alone ( Baek, Lee and Kang, 2025). Demographically, men with T2DM and depression exhibit higher suicide mortality, despite depression being more prevalent in women ( Khubchandani et al., 2023). Additionally, younger T2DM patients are more prone to suicidal ideation and attempts compared to older groups, although absolute mortality rates remain higher in older age groups ( Conti et al., 2017; Kim, Um and Kim, 2022b). Thus, suicide risk in T2DM patients is a multidimensional phenomenon influenced by age, gender, and disorder type, underscoring the need for prioritized risk screening in vulnerable groups.

This review highlights significant disparities in healthcare delivery for patients with T2DM and comorbid mental disorders. While overall mortality rates have declined, patients with SMI have not experienced comparable reductions, indicating structural inequities. Study Han noted that patients with SMI are more likely to require emergency care but less likely to undergo elective hospitalization for cardiovascular conditions, suggesting inadequate preventive care ( Han et al., 2021). Study Knudsen further reported poorer achievement of HbA1c and cholesterol targets in this group, reflecting disparities in diabetes management ( Knudsen, Scheuer and Diaz, 2023).

Collaborative care models have proven effective in reducing depressive symptoms and improving glycemic control, supporting the integration of physical and mental health services as a strategic approach ( van der Feltz-Cornelis et al., 2021). However, implementation in developing countries remains limited due to resource constraints and mental health stigma ( Westling et al., 2012). This review underscores the importance of routine mental health screening in primary diabetes care and integrated interventions, which can reduce both cardiovascular and suicide-related mortality ( Baek, Lee and Kang, 2025; Ho, Chan and Lo, 2025).

A novel contribution of this review is its emphasis on the need for a cross-sectoral approach to T2DM management, extending beyond pharmacological therapy to include psychosocial interventions and public health policies. This broadens the perspective of prior literature, which often focused on clinical relationships without addressing health system dimensions ( Knudsen, Scheuer and Diaz, 2023).

Disparities in health system interactions also vary by disorder type, gender, and age. Patients with SMI are less likely to receive preventive cardiovascular care despite comparable metabolic monitoring, with more pronounced disparities among men, who experience delayed diagnoses and worse outcomes ( Knudsen, Scheuer and Diaz, 2023). Regarding age, elderly patients with T2DM and mental disorders exhibit the highest absolute mortality due to accumulated comorbidities, while those with early-onset T2DM face greater relative risks of complications and failure to meet clinical targets ( Mansour, Golden and Yeh, 2020). Variations by disorder type are also significant, as patients with schizophrenia and bipolar disorder encounter more substantial service gaps compared to those with depression. Therefore, health system improvements must integrate considerations of gender, age, and disorder type to enhance the effectiveness of complication prevention and mortality reduction strategies.

This review is limited by substantial heterogeneity across included studies in terms of design, populations, diagnostic criteria, and outcome definitions, which may affect comparability of findings. Differences in healthcare systems and sociodemographic contexts may also influence the reported effect estimates. In addition, several methodological limitations of the review process should be acknowledged. The search was restricted to English-language publications and did not include grey literature, which may introduce publication bias. Due to heterogeneity, no meta-analysis or formal assessment of reporting bias was conducted, and findings were synthesized narratively. Furthermore, although study selection and data extraction were performed systematically, the process may still be subject to potential reviewer bias.