A systematic review and meta-analysis of the association between low socioeconomic status and all-cause mortality in patients with diabetes Authors: Joanna Konkol, Chunrong Pang, Cheng Pu Published: 2026 ABSTRACT ObjectiveThe aim of this study was to evaluate the association between socioeconomic status (SES) and all-cause mortality among individuals with diabetes. We also examined how individual SES components, including income, employment, education, and housing conditions, were associated with mortality risk.MethodsFollowing PRISMA 2020 guidelines, we searched PubMed, Embase, the Cochrane Library, and Web of Science through September 2025. Eligible studies included adults with type 1 or type 2 diabetes reporting associations between SES indicators (income, education, occupation, or area-level deprivation) and all-cause mortality. Pooled odds ratios (ORs) with 95% confidence intervals (CIs) were calculated using fixed- or random-effects models based on heterogeneity (I2). Publication bias was assessed using funnel plots and Egger’s test.ResultsNineteen studies were included. Low SES was associated with a higher risk of all-cause mortality (OR = 1.67; 95% CI: 1.49–1.88; p < 0.00001). Similar associations were observed across income, education, housing, and employment domains. Heterogeneity was substantial (I2 = 99%), indicating considerable variability across studies. Sensitivity analyses showed that no single study materially influenced the pooled estimate, and publication bias appeared minimal.ConclusionLow socioeconomic status is associated with increased mortality among individuals with diabetes. Addressing socioeconomic inequalities through improved access to education, employment opportunities, healthy environments, and equitable healthcare may help reduce survival disparities and mitigate the overall burden of diabetes.Systematic review registrationhttps://www.crd.york.ac.uk/prospero/display_record.php?RecordID=1248984, identifier PROSPERO (CRD420251248984). Introduction An estimated 589 million adults worldwide are living with diabetes mellitus (DM), with projections indicating an increase to 853 million by 2050 (1), making it a major public health concern. Despite improvements in diagnosis and treatment, this disease continues to contribute to premature all-cause mortality. Large population-based studies show that individuals with diabetes have a significantly higher risk of all-cause mortality than those without diabetes (2, 3). This is mainly due to the increased risk of cardiovascular disease, kidney disease, respiratory disorders, and infections (4). Similar observations have been reported in different populations worldwide (5). The increase in diabetes prevalence has been particularly rapid in low- and middle-income countries, where the increase in incidence has outpaced that observed in high-income settings (6). In line with this, Global Burden of Disease analyses show that mortality attributable to diabetes is disproportionately higher in regions with lower levels of economic development (7), highlighting the potential contribution of socioeconomic disadvantage to inequalities in both diabetes burden and survival outcomes. Additionally, type 2 diabetes is associated with a high burden of comorbid conditions, with cardiovascular, musculoskeletal, and mental health disorders among the most common, and population-based studies indicate that multimorbidity is highly prevalent in people with diabetes, which may considerably complicate disease management (8). The social determinants of health (SDOH), which the World Health Organization defines as “the conditions in which people are born, grow, live, work, and age, and the wider set of forces and systems shaping these conditions,” have a significant impact on health disparities. These determinants contribute to preventable health inequalities and are estimated to account for approximately 30–55% of health outcomes (9). Socioeconomic status (SES), typically measured by education, income, and occupation, represents one of the key determinants (10). Lower SES has been linked to a higher prevalence of diabetes, limited access to preventive care, and poorer disease management outcomes (10–12). Individuals from disadvantaged backgrounds more often face financial hardship, inadequate housing, limited healthcare access, and lower levels of education, all of which have been linked to difficulties in both prevention and treatment (10). Large-scale analyses further indicate that lower education, income, and occupational status are associated with a 30–40% higher risk of developing type 2 diabetes, a pattern observed across multiple national datasets, including NHANES (12, 13). In addition, longitudinal studies have shown that income inadequacy during childhood is associated with an increased risk of diabetes in adulthood (14). Once diabetes develops, socioeconomic differences continue to influence disease management and, consequently, patient survival. Previous studies have shown that lower SES is associated with higher all-cause mortality among individuals with diabetes, even after adjustment for lifestyle and clinical factors (15). For instance, in the UK, patients from the most deprived quintile had a 40–50% higher risk of premature death compared with those from the least deprived areas (15), with similar patterns reported in Scotland, particularly among men (16). Among the main SES components—income, education, employment, and housing—each is linked to higher mortality in individuals with diabetes. In Korea, very low income more than doubled the risk of mortality, particularly among younger men (17). A similar but less pronounced pattern was observed for education, with lower educational attainment being associated with approximately 1.5–1.7 times higher mortality (18). Employment instability is also associated with worse outcomes. Job loss or unemployment increased mortality risk by up to tenfold among women with diabetes (19). In addition, studies on housing deprivation have shown that mortality was highest among tenants and individuals living in low-quality housing, even after adjustment for income and education (20). Given existing evidence suggesting an association between SES and mortality in individuals with diabetes, we aimed to quantitatively synthesize available studies to quantify the strength of the association between low SES and all-cause mortality. In addition, we explored the relative contribution of different SES components and examined potential moderators, including sex, SES domain, study design, and geographic context. In this context, our findings may provide evidence to support policies aimed at reducing socioeconomic inequalities in diabetes outcomes. Methods Literature search This study was registered in PROSPERO (CRD420251248984) and complied with PRISMA 2020 guidelines (21) (see Supplementary Table S1). Four databases (PubMed, Cochrane, Embase, and Web of Science) were used for the literature search. The keywords “Diabetes Mellitus,” “Socioeconomic Factors,” and “Mortality” were used to find eligible records. Studies published up until September 11, 2025, were included in the search. We provided the full search strategy in Supplementary Table S1. Two independent reviewers screened all titles and abstracts to identify eligible studies. Discrepancies were resolved through discussion and consensus. Inclusion and exclusion criteria Eligibility criteria were defined according to the PECOS framework: - Population (P): adults (≥18 years) with type 1 or type 2 diabetes mellitus. - Population (P): adults (≥18 years) with type 1 or type 2 diabetes mellitus. - Exposure (E): low socioeconomic status (income, education, or occupation). - Exposure (E): low socioeconomic status (income, education, or occupation). - Comparator (C): higher socioeconomic status groups. - Comparator (C): higher socioeconomic status groups. - Outcome (O): all-cause mortality (hazard ratio, risk ratio or odds ratio). - Outcome (O): all-cause mortality (hazard ratio, risk ratio or odds ratio). - Study design (S): prospective or retrospective cohort studies, case–control studies, or randomized controlled trials reporting SES–mortality associations. - Study design (S): prospective or retrospective cohort studies, case–control studies, or randomized controlled trials reporting SES–mortality associations. Inclusion criteria: Original observational studies (prospective or retrospective cohort, case–control) Original observational studies (prospective or retrospective cohort, case–control) Reported the association between low SES and mortality (providing HR/RR/OR with 95% CI) Reported the association between low SES and mortality (providing HR/RR/OR with 95% CI) Study population consisted of patients with diabetes Study population consisted of patients with diabetes Exclusion criteria: Reviews, commentaries, abstracts, or other non-original studies Reviews, commentaries, abstracts, or other non-original studies Studies without mortality outcomes or without SES exposure data Studies without mortality outcomes or without SES exposure data Studies with incomplete data or of low methodological quality Studies with incomplete data or of low methodological quality Data extraction Two independent researchers (JK and CP) conducted the data extraction. Any disagreements were resolved through discussion to ensure consistency. Extracted data included the following: Study characteristics: first author, year of publication, study country, and study design. Study characteristics: first author, year of publication, study country, and study design. Population details: sample size, gender, and age. Population details: sample size, gender, and age. Outcomes and effect measures: adjustment variables, outcomes, and effect sizes (OR, RR, HR) with 95% confidence intervals (CIs). Outcomes and effect measures: adjustment variables, outcomes, and effect sizes (OR, RR, HR) with 95% confidence intervals (CIs). Quality assessment The methodological quality and risk of bias of each included study were evaluated using standardized tools according to study design. For randomized controlled trials (RCTs), we applied the Revised Cochrane Risk-of-Bias Tool (RoB 2), which assesses bias across five domains: the randomization process, deviations from intended interventions, missing outcome data, measurement of the outcome, selection of the reported result. For non-randomized studies, including self-controlled studies and retrospective studies, the Risk of Bias in Non-Randomized Studies of Interventions (ROBINS-I) tool was used. Seven domains of bias were assessed, including: confounding, selection of participants, classification of interventions, deviations from intended interventions, missing data, measurement of outcomes, and selection of reported results. The robvis web application1 was used to visualize risk-of-bias assessments. It produced summary bar charts and traffic-light plots to show which studies classified as having a low, unclear, or high risk of bias within each domain. Statistical analysis All of the analyses were performed using Review Manager version 5.4.1 (Cochrane Collaboration, Oxford, United Kingdom). Odds ratios (OR) with 95% confidence intervals (CIs) were used to represent effect sizes. The I2 statistic was used to assess statistical heterogeneity between studies; significant heterogeneity was defined as p < 0.05 or I2 > 50%. When there was substantial heterogeneity, a random-effects model was utilized; otherwise, a fixed-effect model was employed. Funnel plots were used to visually assess publication bias when at least 10 studies were included. In addition, Egger’s regression test was conducted in Stata version 15.0 (StataCorp, College Station, TX, United States) to evaluate asymmetry. All statistical tests were two-sided, and p < 0.05 was considered statistically significant. Results Study characteristics and results of the screening process A total of 3,479 records were identified: 1,317 from PubMed, 1,794 from Web of Science, 182 from Embase (Ovid), and 186 from the Cochrane Library. After removing 471 duplicates, 2,989 records were excluded based on ineligible article types, irrelevant topic, lack of full text, or missing data. Finally, 19 studies met the inclusion criteria and were included in the meta-analysis (Figure 1) (17, 19, 20, 22–37). The included studies were conducted across multiple regions, including Europe, North America, Asia, and Australia, with the largest number originating from Korea, the United Kingdom, China, and the United States. Most studies were observational cohort studies, whereas only one randomized controlled trial was included. The analyzed populations included both type 1 and type 2 diabetes, although type 2 diabetes predominated across studies. The key characteristics of the included studies are summarized in Table 1. PRISMA flow diagram of study selection. Table 1Basic information of literature characteristics.StudyYear of publicationCountryStudy regionsDiseasesStudy designFactorsGroupSample (total/case/control)Age of CasesAge of ControlsGender of cases(F/M)Gender of Control(F/M)Duration of study(year)OR, RR, HR, RH(95% CI)OutcomesSaydah 2010I2010USAUSADiabetesCross-sectionalEducational level 15 years139,681/62059/1654155.2 ± 8.50/na0/na0–2, 2–4, 4–6, >61.21(1.12,1.31)All-cause mortalityDalsgaard 2015II_F2015DenmarkDanishType 2 diabetesCohortEducational level≤10 years vs. > 15 yearsna/0na/00–2, 2–4, 4–6, >61.29(1.17,1.42)All-cause mortalityDalsgaard 2015III_M2015DenmarkDanishType 2 diabetesCohortIncome level20 vs. 80 percentile139,681/33850/214190/na0/na0–2, 2–4, 4–6, >61.42(1.30,1.56)All-cause mortalityDalsgaard 2015IV_F2015DenmarkDanishType 2 diabetesCohortIncome level20 vs. 80 percentilena/0na/00–2, 2–4, 4–6, >61.40(1.26,1.56)All-cause mortalityKim 20162016KoreaKoreadiabetes mellitusCohortSocioeconomic statusLowest 30% vs. highest 30%20,220/4903/830256.1 ± 11.453.8 ± 11.72972/53302393/25107.91.31(1.10,1.55)mortalityRawshani 2016I2016SwedenSwedenType 2 diabetesCohortIncome quintileLowest vs. highest115,029/23734/2694259.0 ± 10.757.3 ± 8.124,567/171419361/38891102.12(2.01,2.24)MortalityRawshani 2016II2016SwedenSwedenType 2 diabetesCohortEducational level<9 years vs. college/university87,781/29525/2115160.4 ± 8.657.2 ± 9.230,117/4574514,907/24322101.23(1.18,1.30)MortalityShin 2016I_M2016KoreaKoreaType 2 diabetesCohortIncomeLow vs. high6156/2016/226444–7544–750/20162264/071.56(1.18,2.05)MortalityShin 2016II_F2016KoreaKoreaType 2 diabetesCohortIncomeLow vs. high7705/2164/23912164/00/239171.61(1.16,2.23)MortalityShin 2016III_M2016KoreaKoreaType 2 diabetesCohortJob statusUnemployed vs. maintain Job7156/2793/30190/27933019/073.78(2.81,5.09)MortalityShin 2016IV_F2016KoreaKoreaType 2 diabetesCohortJob statusUnemployed vs. maintain Job7705/6049/9636049/00/96379.78(2.39,39.98)MortalityShin 2016V_M2016KoreaKoreaType 2 diabetesCohortResidenceRural areas vs. Seoul7156/1528/8050/1528805/071.11(0.86,1.45)MortalityShin 2016VI_F2016KoreaKoreaType 2 diabetesCohortResidenceRural areas vs. Seoul7705/1605/8801605/00/88071.30(0.91,1.85)MortalityAguilar-Palacio 2017I_M2017SpainSpanishDiabetesCohortIncomeDeprivation vs. affluent census tractsna16–29nana101.08(0.59,1.98)MortalityAguilar-Palacio 2017II_F2017SpainSpanishDiabetesCohortINCOMEDeprivation vs. affluent census tractsna16–29nana100.62(0.27,1.41)MortalityBlomster 20172017AustraliaAustraliaType 2 diabetesRCTsEducationLow vs. higher education11,140/4024/711667.1 ± 6.365.0 ± 6.32748/12761985/513151.34(1.18,1.52)MortalityShin 2018I_M2018KoreaKoreaDiabetesCohortIncomeLow vs. high34,403/12813/2159056(48–64)12,813/00/21590121.32(1.23,1.41)MortalityShin 2018II_F2018KoreaKoreaDiabetesCohortIncomeLow vs. high21,036/9854/111820/985411,182/0121.21(1.10,1.33)MortalityShulman 20182018CanadaOntarioType 1 diabetesCohortIncomeMost vs. least deprivedna<19nana6.6 (3.0–10.8) and 6.5 (3.0–10.7)2.03(1.13,3.63)MortalityCampbell 2020I_M2020UKScotlandType 1 diabetesCohortSIMD20% most vs. least deprived areas274,095/55217/504120–39, 40–59, 60–79 and ≥80 years120,527/81368101.21(0.98,1.49)MortalityCampbell 2020II_F2020UKScotlandType 1 diabetesCohortSIMD20% most vs. least deprived areas101.55(1.20,1.99)MortalityChen 2022I_M2022ChinaChangshu and HuaianType 2 diabetesCohortSESLow vs. high6971/2098/2651nana0/20980/26515.7 ± 0.91.87(1.55,2.25)All-cause mortalityChen 2022II_F2022ChinaChangshu and HuaianType 2 diabetesCohortSESLow vs. high10,582/6158/2219nana6158/02219/01.53(1.26,1.86)All-cause mortalityLusk 20222022DurhamDurhamType 1 and 2 DiabetesCohortADIADI 86-100(low SES) vs. ADI 1-15(high SES)130,100/19300/19800>65>6566,507/6364431.22(1.11,1.34)30-day mortalityLee 20232023KoreaKoreaType 2 diabetesCohortIncomeLowest vs. highest income status1,943,354/401061/594593nananana50.89(0.80,0.98)All-cause mortalityLiao 20232023ChinaTaiwanType 2 diabetesCohortEducational levelElementary or below vs. College or above1,307,963/667419/118596nana620,798/687165161.89(1.85,1.92)P4P participation on mortalityRosella 20232023CanadaOntarioDiabetesCohortIncome20% lowest vs. highest income1,741,098/409403/27795758.06 ± 14.76204,996/204407122,579/155378251.05(1.04,1.06)All-cause mortalityLi 20252025ChinaUSDiabetesCohortHousehold incomeLow vs. high household income2069/477/425nananana182.78(1.89,4.17)All-cause mortality Basic information of literature characteristics. Assessment of study quality Quality assessment revealed that among the 19 included studies, 1 was rated as low risk of bias, 13 as moderate risk, and 5 as high risk overall. Among 18 non-randomized studies assessed with ROBINS-I, 4 showed serious risk of bias, mainly due to confounding (19, 23, 28, 32). One study (37) was judged as low risk of bias, while the remaining studies were classified as moderate risk. The single randomized controlled trial (29), assessed using the RoB 2 tool, was rated as high risk of bias, mainly due to deviations from the intended intervention. Because only one RCT was included, and the vast majority of studies were observational, the RCT did not materially influence the pooled effect. Detailed risk-of-bias assessments are presented in Figures 2, 3. Risk of bias assessment for the randomized controlled trial using the RoB 2 tool. Risk of bias assessment for non-randomized studies using the ROBINS-I tool. Results of meta-analysis The meta-analysis of 19 studies demonstrated that low socioeconomic status (SES) was significantly associated with higher all-cause mortality among patients with diabetes (OR = 1.67; 95% CI: 1.49–1.88; p < 0.00001). The association was consistently observed across all included studies, with no evidence of protective effects. However, heterogeneity between studies was substantial (I2 = 99%), which may reflect differences in SES measures (e.g., education, income), study populations, analytical approaches, as well as other unmeasured factors. Despite the high heterogeneity, an association between socioeconomic disadvantage and mortality risk in diabetes was observed across studies. Detailed results are shown in Figure 4. Forest plot of the association between low socioeconomic status and all-cause mortality in patients with diabetes. Income By country: Significant differences were observed (p < 0.0001). Low income was associated with increased mortality risk in most populations, with the strongest effects in China (OR = 2.18; 95% CI: 1.51–3.14) and Sweden (OR = 2.12; 95% CI: 2.01–2.24). Significant but smaller associations were observed in Denmark and the United Kingdom, whereas no significant association was found in Spain (OR = 0.88; 95% CI: 0.52–1.49). The pooled effect confirmed an elevated mortality risk among low-income groups (OR = 1.49; 95% CI: 1.24–1.79; I2 = 99%). Detailed results are shown in Supplementary Figure S1A. By gender: Low income increased mortality risk in both sexes, with comparable effect sizes: men (OR = 1.35; 95% CI: 1.28–1.43; I2 = 6%) and women (OR = 1.31; 95% CI: 1.23–1.40; I2 = 62%). The difference between sexes was not statistically significant (p = 0.50). Detailed results are shown in Supplementary Figure S1B. By study design: Cohort studies demonstrated an overall effect of OR = 1.63 (95% CI 1.36–1.95; I2 = 98%), while the single cross-sectional study showed a stronger effect (OR = 2.46; 95% CI 2.15–2.81; p = 0.0003). Given the limited number of cross-sectional studies, this difference should be interpreted with caution. Detailed results are presented in Supplementary Figure S1C. Education By country: Low educational attainment was associated with increased mortality across most countries, particularly in Belgium (OR = 4.17; 95% CI: 1.45–11.97) and the USA (OR = 3.00; 95% CI: 1.20–7.50). Moderate but significant effects were also observed in Australia, Denmark, Sweden, and the UK. Detailed results are presented in Supplementary Figure S3A. By gender: Lower education increased mortality risk in both men (OR = 1.37; 95% CI 1.16–1.61; I2 = 78%) and women (OR = 2.24; 95% CI 1.37–3.64; I2 = 90%). The gender difference was not statistically significant (p = 0.06) but suggested a potentially stronger effect among women. Detailed results are presented in Supplementary Figure S3B. By study design: Cohort studies demonstrated an overall effect of OR = 1.63 (95% CI 1.36–1.95; I2 = 98%), while the cross-sectional study indicated a stronger effect (OR = 2.46; 95% CI 2.15–2.81; p = 0.0003). As only one cross-sectional study was available, results should be interpreted with caution. Detailed results are presented in Supplementary Figure S3C. Housing status Disadvantaged housing conditions were associated with a significantly higher risk of all-cause mortality (OR = 1.99; 95% CI 1.77–2.24; I2 = 89%). By country, the effect was strongest in Belgium (OR = 2.38) and the UK (OR = 2.58), and not statistically significant in Korea (OR = 1.17). By gender, the effect was significant in both men (OR = 1.66; 95% CI 1.01–2.74) and women (OR = 2.48; 95% CI 1.36–4.51), with no significant sex difference (p = 0.31). Detailed results are shown in Supplementary Figures S5A,B. Employment status Unemployment was associated with a substantially higher mortality risk (OR = 3.33; 95% CI: 1.55–7.17; I2 = 72%). Country-specific effects varied: no significant association was found in the United States (OR = 1.60; 95% CI: 0.80–3.20), whereas a strong effect was observed in Korea (OR = 4.69; 95% CI: 2.15–10.23; p = 0.0001). The between-country difference was statistically significant (p = 0.04). Detailed results are presented in Supplementary Figure S7. Publication bias Egger’s test indicated no significant publication bias for most analyses (p > 0.25), except for education by gender (p = 0.002). Funnel plots showed mild asymmetry in some subgroups, which likely resulted from heterogeneity rather than true publication bias. However, these findings should be interpreted with caution because some subgroup analyses included only a small number of studies. Funnel plots and detailed statistical results are provided in Supplementary Figures S2–S8. Sensitivity analysis The results remained stable in leave-one-out sensitivity analyses, as exclusion of individual studies did not materially affect the pooled estimates for income or education (Supplementary Figures S9–S11). Discussion The results of our meta-analysis indicate that low SES is associated with a 67% higher risk of all-cause mortality among individuals with diabetes (OR = 1.67; 95% CI: 1.49–1.88). All of the studies showed the same direction of effect, with none showing a protective association. In subsequent analyses, subgroup analyses suggested that different SES components may contribute unequally to mortality risk, with the strongest association observed for employment status (OR = 3.33). Housing conditions (OR = 1.99) and educational attainment (OR ≈ 1.63), followed by income (OR = 1.49), also showed significant positive associations. This indicates that multiple dimensions of socioeconomic disadvantage may influence survival, although their relative contributions appear to differ. One possible explanation for this phenomenon is unequal access to healthcare. Effective diabetes management requires regular monitoring, timely treatment, and consistent access to medications and supplies. However, these resources are often more difficult to obtain for socioeconomically disadvantaged patients. Limited insurance coverage (38–42) and high healthcare costs (10, 32) may contribute to delayed diagnosis, poorer glycemic control, and reduced treatment adherence. This may partly explain our findings regarding employment status, which showed the strongest association with mortality in our analysis (OR = 3.33). This may be because employment status influences both access to insurance coverage and the ability to afford healthcare-related expenses. Similarly, lower income can further restrict access to healthcare resources and the affordability of treatment. In our analysis, income was also associated with increased mortality risk (OR = 1.49). This is in line with previous studies showing an inverse relationship between income level and mortality risk among individuals with diabetes (17, 33, 36), a pattern that persists even in countries with broadly universal healthcare coverage, such as Denmark and the UK (15, 25). Another potential explanation for higher mortality may be differences in the use of specialist diabetes care. Evidence suggests that individuals with lower SES are less likely to achieve optimal disease control, partly due to lower adherence to treatment and reduced follow-up care (43). This is reflected in less frequent attendance at diabetes centers, lower treatment intensity, and fewer therapy adjustments (44). Finally, differences between individuals with higher and lower SES are also reflected in a higher risk of diabetes-related complications among those in more disadvantaged situations. These include retinopathy, cardiovascular complications, and diabetic ketoacidosis (45, 46). However, the problem of increased mortality among individuals in disadvantaged socioeconomic conditions extends beyond access to medical care alone. Housing-related deprivation has been associated with poorer glycemic control, while unstable living conditions can make self-management, medication adherence, and healthy eating more difficult (47, 48). This pattern is consistent with our analysis, in which housing conditions demonstrated a strong association with mortality risk (OR = 1.99). In addition, individuals with lower SES often face restricted access to healthy food and transportation (49), which may further complicate disease management. Educational attainment represents another important dimension of socioeconomic disadvantage, which was also reflected in the results of our meta-analysis. Lower education and limited health literacy may reduce patients’ capacity for self-management, treatment adherence, and participation in preventive care (10). Furthermore, evidence from systematic reviews indicates that lower health literacy is associated with poorer glycemic control (50). New technologies that have the potential to support diabetes management are also less accessible to individuals with lower SES due to disparities in digital literacy, internet access, and financial constraints (51). In our analysis, the association between education and mortality appeared stronger among women, suggesting that educational inequalities may disproportionately affect disease management in female populations. Behavioral patterns that may influence mortality in individuals with diabetes are also linked to patients’ SES, and their effects may accumulate. For instance, those living in more deprived conditions are more likely to engage in unhealthy behaviors, including smoking, physical inactivity, poor diet, and prolonged sedentary time, which may contribute to differences in diabetes management and outcomes (15). In addition, chronic psychosocial stress, which is more common among lower-SES groups, may further reinforce unhealthy behaviors, including tobacco use, reduced physical activity, and poor dietary patterns (52). On the other hand, studies have shown that the protective effects of healthy lifestyle factors, such as non-smoking, regular physical activity, and a balanced diet, appear to be stronger among individuals with higher SES. This may indicate that adverse socioeconomic conditions can limit the beneficial effects of such behaviors (53). Importantly, the relationship between type 2 diabetes and socioeconomic status appears to be bidirectional. While lower SES increases the risk of developing diabetes, living with the disease may itself contribute to further socioeconomic disadvantage. Patients living with diabetes are exposed to increased healthcare expenditures and significant out-of-pocket costs. In addition, diabetes negatively affects productivity, contributes to higher rates of unemployment, and is associated with decreased educational attainment, all of which may adversely influence long-term SES. Taken together, these interconnected pathways may create a self-reinforcing cycle of disadvantage, making both disease management and social mobility increasingly difficult (54). Considering these mechanisms, socioeconomic disadvantage likely reflects a combination of interrelated conditions rather than isolated exposures. Therefore, composite indicators such as SES may be helpful in guiding strategies aimed at addressing overlapping determinants and improving survival outcomes in individuals with diabetes. Importantly, these findings should also be interpreted in the context of health inequalities. Low socioeconomic status may reflect social and economic barriers affecting access to healthcare, preventive services, and long-term diabetes management. In this context, reducing mortality among individuals with diabetes may require not only optimized clinical treatment but also public health strategies aimed at reducing socioeconomic disparities. However, several limitations should be considered when interpreting these findings. First, most included studies were observational, which limits causal inference. The inclusion of a single RCT among predominantly observational evidence may also have contributed to methodological heterogeneity. Second, substantial heterogeneity was observed between studies (I2 = 99%), likely reflecting differences in SES definitions, study populations, and analytical approaches. SES indicators varied across studies—spanning income, education, occupation, and area-level deprivation—and composite versus domain-specific measures were not consistently reported. As a result, the pooled estimate should be interpreted as a summary of the broadly adverse association observed across settings rather than a precise effect size. Third, reporting of key determinants—such as age, lifestyle factors, multimorbidity, and access to healthcare—differed to some extent between studies, which increases the likelihood of residual confounding. Reverse causality is also a relevant issue, as more advanced illness may contribute to declines in socioeconomic status. Finally, the use of all-cause mortality as the primary outcome limits the ability to distinguish diabetes-related deaths from those unrelated to diabetes-specific pathways. Future research should incorporate diabetes-attributed or cause-specific mortality, which may offer a clearer understanding of the mechanisms linking socioeconomic status with mortality among individuals with diabetes. Conclusion Lower socioeconomic status is significantly associated with increased all-cause mortality among individuals with diabetes. Higher mortality risk was consistently linked to lower income, education, occupational status, and housing conditions. These findings suggest that survival outcomes in diabetes may depend not only on biological and clinical factors but also on broader socioeconomic conditions. Addressing socioeconomic disadvantage and reducing structural barriers to healthcare may therefore help improve survival in this population.