Abstract
Purpose
To investigate the association between dipeptidyl peptidase-4 inhibitors (DPP4is) and the risk of primary open-angle glaucoma (POAG) and normal tension glaucoma (NTG) in patients with type 2 diabetes mellitus (T2DM).
Design
Retrospective cohort study.
Subjects
A total of 582 710 patients with T2DM treated with either DPP4i (exposure group) or non-DPP4i medications (control group) were analyzed between 2008 and 2021.
Methods
Patients were 1-to-1 matched by propensity scores on demographic and clinical characteristics. Cox proportional hazards models were applied to estimate hazard ratios for POAG and NTG, adjusting for age, sex, comorbidities, and concurrent antidiabetic medications.
Main Outcome Measures
Incidences of POAG and NTG.
Results
Dipeptidyl peptidase-4 inhibitor users demonstrated a significantly lower risk of POAG (adjusted hazard ratio [aHR], 0.53; 95% confidence interval [CI], 0.50–0.56) and NTG (aHR, 0.55; 95% CI, 0.50–0.62) compared to non-DPP4i users on first-generation diabetes medication. Subgroup analysis revealed a consistent risk reduction across all age groups (18–39: aHR, 0.56; 95% CI, 0.51–0.62; 40–64: aHR, 0.52; 95% CI, 0.47–0.57; ≥65 years: aHR, 0.51; 95% CI, 0.47–0.56) and among patients with or without diabetic-related complications, including diabetic retinopathy, diabetic kidney disease, and diabetic neuropathy (DN) (aHR: without vs. with diabetic retinopathy [0.54 vs. 0.43], without vs. with diabetic kidney disease [0.53 vs. 0.49], without vs. with DN [0.54 vs.0.43]), with all comparisons showing statistical significance (P < 0.001). Cumulative incidence analyses revealed a sustained lower risk for DPP4i users throughout the study period (log-rank P < 0.001).
Conclusions
Exposure to DPP4i was associated with a reduced risk of developing POAG and NTG compared with users of first-generation diabetes medication. Further research is needed to explore the underlying mechanisms and their implications for glaucoma prevention and management.
Financial Disclosure(s)
Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.
Keywords: Dipeptidyl peptidase-4 inhibitors, Normal tension glaucoma, Ocular hypertension, Primary open-angle glaucoma, Type 2 diabetes mellitus
Diabetes poses a significant public health concern, affecting 10.5% of adults worldwide, with type 2 diabetes mellitus (T2DM) accounting for >90% of cases. The high prevalence of T2DM significantly contributes to morbidity, including an increased risk of glaucoma. Glaucoma, particularly primary open-angle glaucoma (POAG) and normal tension glaucoma (NTG), further exacerbates the chronic disease burden associated with T2DM.1, 2, 3
Antidiabetic medications provide cardiovascular, renal, and neurotrophic benefits beyond glucose control.4 Metformin—the preferred first-line therapy for T2DM5—has shown geroprotective properties and a potential to reduce the risk of open-angle glaucoma compared with other first-generation diabetes drugs.6 Recent retrospective studies have extended these findings to second-generation diabetes drugs like glucagon-like peptide-1 receptor agonists (GLP-1 RAs)7,8 and sodium-glucose cotransporter 2 inhibitors,9 both of which are associated with lower glaucoma risk. Furthermore, GLP-1 RAs, which act on incretin system, have shown various systemic advantages, including potential neuroprotective properties.10,11
Dipeptidyl peptidase-4 inhibitors (DPP4is), another second-generation diabetes drug, prolong endogenous incretin activity by inhibiting the rapid degradation of incretins.10 While DPP4is share incretin-mediated effects with GLP-1 RAs, emerging evidence suggests that DPP4is may confer additional benefits through mechanisms independent of incretins.12, 13, 14 These may involve modulation of reactive oxygen species (ROS) and glutamate signaling, potentially contributing to optic nerve health in ways that differ from GLP-1 RAs.12, 13, 14, 15 Rodent studies have demonstrated DPP4i benefits in inflammatory, microvascular, and neurodegenerative disorders, such as Parkinson disease11 and cerebral ischemia.16 However, limited studies have explored the association between DPP4is and glaucoma. A recent study found lower glaucoma risk with sodium-glucose cotransporter 2 inhibitors compared with DPP4is, but it lacked comparisons with other antidiabetic medications and did not adjust for concurrent use of other glucose-lowering agents, which may affect glaucoma risk.17
Given the emerging evidence linking antidiabetic agents to glaucoma risk reduction, DPP4is represent a promising candidate due to its incretin-mediated and independent effects. This study aims to investigate the association between DPP4i use and the risk of developing POAG or NTG in patients with T2DM. To minimize confounding, we excluded users of other second-generation diabetes drugs previously associated with glaucoma risk. Our analysis specifically focused on evaluating the association between DPP4i and first-generation diabetes medication (including metformin, sulfonylureas, thiazolidinediones, α-glucosidase inhibitors, and glinide).
Methods
Data Source
This retrospective cohort study analyzed data from Taiwan's National Health Insurance Research Database, a database covering >99% of the population. Diagnoses in the National Health Insurance Research Database are coded using the International Classification of Diseases 9 and 10, Clinical Modification (ICD9-CM and ICD10-CM). All analyses were conducted within a secure environment at the Health and Welfare Data Science Center under the supervision of the Ministry of Health and Welfare. Ethical approval for the study was granted by the Research Ethics Committee of China Medical University Hospital (CMUH110-REC1-038[CR-4]).
Study Population
Eligible patients were adults diagnosed with T2DM, identified using ICD-9-CM codes 250.x0 and 250.x2 or ICD-10-CM code E11. To improve case ascertainment, patients were required to have ≥3 outpatient claims or ≥1 inpatient claim for T2DM. Patients were included if they were prescribed either DPP4is (Anatomical Therapeutic Chemical codes: A10BH01–A10BH05, A10BH09) or other diabetes medications (listed in Table S1, available at www.ophthalmologyscience.org) for a minimum of 3 months. The DPP4i cohort comprised patients initiating DPP4i therapy, with the index date defined as the date of their first DPP4i prescription.
To establish a comparable cohort and ensure a consistent observation period, non-DPP4i users were matched 1:1 with DPP4i users using propensity score matching based on variables such as age, sex, index year, dwelling region, urbanization level, insurance premium, comorbidities, and Charlson Comorbidity Index scores. The index date for each non-DPP4i user was assigned to match the index year of their corresponding DPP4i user.
Additionally, matching considered the number of diabetes medication classes prescribed within 45 days of the index date. For example, if a patient initiating DPP4i therapy was already on 1 class of diabetes medication, the matched patient in the non-DPP4i cohort was also required to initiate a second non-DPP4i diabetes medication class.
Patients with insulin use prior to the index date were excluded to minimize potential confounding, as they often represent a population with more advanced or poorly controlled diabetes, which could independently affect glaucoma risk. Additional exclusion criteria included incomplete baseline data, age <18 years, a history of type 1 diabetes or ocular hypertension, prior glaucoma surgery, and any use of glaucoma medications, GLP-1 RAs, or sodium-glucose cotransporter 2 inhibitors before the index date. The study enrollment period spanned from January 1, 2009, to December 31, 2020, with follow-up until December 31, 2021. A flowchart illustrating patient selection is presented in Figure 1.
Figure 1.
Flowchart of study participants. From the NHIRD, 582 710 patients with type 2 diabetes were included, comprising 291 355 DPP4i users and 291 355 non-DPP4i users matched by propensity scores. DPP4i = dipeptidyl peptidase-4 inhibitor; GLP-1 RA = glucagon-like peptide-1 receptor agonist; NHIRD = National Health Insurance Research Database; T2DM = type 2 diabetes mellitus.
Outcomes and Covariates
The primary outcomes were the incidence of POAG and NTG. Primary open-angle glaucoma was identified using ICD-9-CM codes 365.10 to 365.12, 365.15, and ICD-10-CM codes H40.10 to H40.12, H40.15. Normal tension glaucoma was defined using ICD-9-CM code 365.12 and ICD-10-CM code H40.12. To enhance diagnostic accuracy, a diagnosis of POAG or NTG required ≥2 outpatient claims or 1 inpatient claim with the corresponding codes. Follow-up continued until the first occurrence of POAG, NTG, death, or December 31, 2021.
Baseline characteristics included demographic factors (age, sex, and region of dwelling), socioeconomic status (insurance premium levels and urbanization level), and clinical factors (comorbidities such as hypertension, dyslipidemia, coronary artery disease, stroke, chronic kidney disease, retinopathy, cataract, uveitis, myopia, diabetic retinopathy, diabetic kidney disease, and diabetic neuropathy [DN]), as shown in Table 2. The use of first-generation diabetes medication—including metformin, sulfonylureas, thiazolidinediones, α-glucosidase inhibitors, and glinides—were considered as covariates in the analyses.
Table 2.
Baseline Characteristics in T2DM between DPP4i and Non-DPP4i User
| Variable | Non-DPP4i User |
DPP4i User |
SMD | ||
|---|---|---|---|---|---|
| (N = 291 355) |
(N = 291 355) |
||||
| n | % | n | % | ||
| Sex | 0.003 | ||||
| Female | 128 994 | 44.27 | 129 425 | 44.42 | |
| Male | 162 361 | 55.73 | 161 930 | 55.58 | |
| Age (yrs) | |||||
| 18–39 | 90 025 | 30.90 | 89 907 | 30.86 | 0.001 |
| 40–64 | 85 271 | 29.27 | 85 912 | 29.49 | 0.005 |
| ≥65 | 116 059 | 39.83 | 115 536 | 39.65 | 0.004 |
| Mean ± SD | 60.58 | 12.67 | 60.49 | 12.59 | 0.007 |
| Region of dwelling | |||||
| Northern | 129 826 | 44.56 | 128 010 | 43.94 | 0.013 |
| Central | 68 593 | 23.54 | 68 171 | 23.40 | 0.003 |
| Southern | 84 313 | 28.94 | 86 794 | 29.79 | 0.019 |
| Eastern | 8623 | 2.96 | 8380 | 2.88 | 0.005 |
| Degree of urbanization | |||||
| Lowest | 27 597 | 9.47 | 27 228 | 9.35 | 0.004 |
| Median | 112 405 | 38.58 | 112 807 | 38.72 | 0.003 |
| Highest | 151 353 | 51.95 | 151 320 | 51.94 | <0.001 |
| Insurance premium (NTD per month) | |||||
| <18 000 | 74 172 | 25.46 | 72 597 | 24.92 | 0.012 |
| 18 000–34 999 | 141 257 | 48.48 | 142 764 | 49.00 | 0.010 |
| ≥35 000 | 75 926 | 26.06 | 75 994 | 26.08 | 0.001 |
| Comorbidities | |||||
| Coronary artery disease | 62 012 | 21.28 | 60 150 | 20.64 | 0.016 |
| Stroke | 36 564 | 12.55 | 35 013 | 12.02 | 0.016 |
| Peripheral artery disease | 4378 | 1.50 | 4227 | 1.45 | 0.004 |
| Atrial fibrillation | 26 688 | 9.16 | 25 624 | 8.79 | 0.013 |
| Heart failure | 12 738 | 4.37 | 12 212 | 4.19 | 0.009 |
| Hypertension | 195 969 | 67.26 | 196 614 | 67.48 | 0.005 |
| Dyslipidemia | 185 583 | 63.70 | 186 489 | 64.01 | 0.006 |
| CKD | 17 340 | 5.95 | 17 514 | 6.01 | 0.003 |
| COPD | 31 052 | 10.66 | 30 107 | 10.33 | 0.011 |
| Depression | 18 751 | 6.44 | 18 262 | 6.27 | 0.007 |
| Schizophrenia | 3156 | 1.08 | 3224 | 1.11 | 0.002 |
| Rheumatoid diseases | 5387 | 1.85 | 5308 | 1.82 | 0.002 |
| Osteoporosis | 14 973 | 5.14 | 14 674 | 5.04 | 0.005 |
| Alcoholism | 5780 | 1.98 | 5571 | 1.91 | 0.005 |
| Nicotine dependence | 8357 | 2.87 | 8134 | 2.79 | 0.005 |
| Cancer | 14 732 | 5.06 | 13 830 | 4.75 | 0.014 |
| Retinopathy | 23 733 | 8.15 | 25 518 | 8.76 | 0.022 |
| Myopia | 2690 | 0.92 | 2606 | 0.89 | 0.003 |
| Cataract | 63 530 | 21.81 | 63 296 | 21.72 | 0.002 |
| Uveitis | 1814 | 0.62 | 1686 | 0.58 | 0.006 |
| DR | 10 352 | 3.55 | 12 706 | 4.36 | 0.041 |
| DKD | 21 045 | 7.22 | 22 943 | 7.87 | 0.025 |
| DN | 13 802 | 4.74 | 15 494 | 5.32 | 0.027 |
| CCI score | |||||
| 0 | 225 423 | 77.37 | 228 458 | 78.41 | 0.025 |
| 1 | 31 235 | 10.72 | 30 077 | 10.32 | 0.013 |
| ≥2 | 34 697 | 11.91 | 32 820 | 11.26 | 0.020 |
| Number of DM medication | |||||
| Mean ± SD | 2.11 | 0.76 | 2.1 | 0.77 | 0.009 |
CCI = Charlson Comorbidity Index; CKD = chronic kidney disease; COPD = chronic obstructive pulmonary disease; DKD = diabetic kidney disease; DM = diabetes mellitus; DN = diabetic neuropathy; DPP4i = dipeptidyl peptidase-4 inhibitor; DR = diabetic retinopathy; NTD = New Taiwan Dollar; SD = standard deviation; SMD = standardized mean differences; T2DM = type 2 diabetes mellitus.
Statistical Analysis
Baseline characteristics were summarized using means and standard deviations for continuous variables and counts with percentages for categorical variables. Standardized mean differences were calculated to assess the balance between the DPP4i and non-DPP4i cohorts after propensity score matching, with standardized mean differences <0.1 indicating minimal imbalance. Cox proportional hazards models were employed to estimate hazard ratios and 95% confidence intervals (CIs) for the development of POAG and NTG. Both unadjusted and adjusted models were constructed, with adjustments made for potential confounders such as age, sex, comorbidities, and concurrent medication use.
Cumulative incidences of POAG and NTG were estimated using the Kaplan–Meier method, and differences between the DPP4i and non-DPP4i cohorts were assessed using the log-rank test. Statistical significance was set at a 2-sided P value of <0.05. All analyses were performed using SAS version 9.4 (SAS Inc) and R software version 4.4.1 (R Foundation for Statistical Computing).
Results
Baseline Characteristics
After propensity score matching, a total of 291 355 individuals were included in both the DPP4i users and non-DPP4i users cohorts, resulting in a combined study population of 582 710 patients with T2DM. Baseline characteristics were well-balanced across demographic, socioeconomic, and clinical factors, with standardized mean differences below 0.1 for all variables, demonstrating comparability between groups (Table 2). The mean age in both cohorts was approximately 60 years, with a male predominance (about 56% in both). Comorbidities such as hypertension (67.26% in non-DPP4i group vs. 67.48% in DPP4i group), dyslipidemia (63.70% in non-DPP4i vs. 64.01% in DPP4i), cataract (21.81% in non-DPP4i vs. 21.72% in DPP4i), diabetic retinopathy (3.55% in non-DPP4i vs. 4.36% in DPP4i), diabetic kidney disease (7.22% in non-DPP4i vs. 7.87% in DPP4i), and DN (4.74% in non-DPP4i vs. 5.32% in DPP4i), were similarly distributed between the cohorts.
DPP4i Use and Risk of POAG
During a mean follow-up period of 6.04 ± 3.25 years for DPP4i users and 5.59 ± 3.39 years for non-DPP4i users, a total of 3159 cases of POAG were identified in the DPP4i users cohort and 6128 cases in the non-DPP4i users cohort. Dipeptidyl peptidase-4 inhibitor use was associated with a significantly lower risk of developing POAG compared with non-DPP4i use (adjusted hazard ratio [aHR], 0.53; 95% CI, 0.50–0.56; P < 0.001) (Table 3). Male sex was associated with a slightly higher risk of POAG compared with female sex (aHR, 1.14; 95% CI, 1.09–1.19; P < 0.001). Advanced age and certain comorbidities, including coronary artery disease, depression, cancer, retinopathy, myopia, cataract, and DN, were linked to an increased risk of POAG (Table 3). The cumulative incidence of POAG was consistently lower in DPP4i users compared with non-DPP4i users over the study period, as illustrated by the Kaplan–Meier survival curves (Fig 2). The log-rank test indicated a significant difference between the cohorts (P < 0.001).
Table 3.
Hazard Ratios of POAG among T2DM with DPP4i and Non-DPP4i User
| Variable | POAG |
Crude |
Adjusted∗ |
||||
|---|---|---|---|---|---|---|---|
| N | PY | IR | HR (95% CI) | P Value | HR (95% CI) | P Value | |
| Non-DPP4i user | 6128 | 1 628 190.67 | 3.76 | 1.00 (reference) | - | 1.00 (reference) | - |
| DPP4i user | 3159 | 1 759 130.36 | 1.8 | 0.48 (0.46, 0.50)§ | <0.001 | 0.53 (0.50, 0.56)§ | <0.001 |
| Sex | |||||||
| Female | 4039 | 1 542 610.63 | 2.62 | 1.00 (reference) | - | 1.00 (reference) | - |
| Male | 5248 | 1 844 710.41 | 2.84 | 1.09 (1.04, 1.13)§ | <0.001 | 1.14 (1.09, 1.19)§ | <0.001 |
| Age | |||||||
| 18–39 | 2680 | 1 097 550.35 | 2.44 | 1.00 (reference) | - | 1.00 (reference) | - |
| 40–64 | 2842 | 1 029 490.34 | 2.76 | 1.13 (1.07, 1.19)§ | <0.001 | 1.08 (1.03, 1.14)‡ | 0.004 |
| ≥65 | 3765 | 1 260 280.34 | 2.99 | 1.23 (1.17, 1.29)§ | <0.001 | 1.12 (1.05, 1.18)§ | <0.001 |
| Region of dwelling | |||||||
| Northern | 4604 | 1 532 851.05 | 3 | 1.00 (reference) | - | 1.00 (reference) | - |
| Central | 2080 | 784 404.68 | 2.65 | 0.88 (0.84, 0.93)§ | <0.001 | 0.90 (0.85, 0.95)§ | <0.001 |
| Southern | 2370 | 971 888.85 | 2.44 | 0.81 (0.77, 0.85)§ | <0.001 | 0.82 (0.78, 0.87)§ | <0.001 |
| Eastern | 233 | 98 176.45 | 2.37 | 0.79 (0.69, 0.90)§ | <0.001 | 0.81 (0.71, 0.92)‡ | 0.001 |
| Degree of urbanization | |||||||
| Lowest | 719 | 315 518.46 | 2.28 | 1.00 (reference) | - | 1.00 (reference) | - |
| Median | 3462 | 1 302 505.88 | 2.66 | 1.17 (1.08, 1.26)§ | <0.001 | 1.17 (1.08, 1.27)§ | <0.001 |
| Highest | 5106 | 1 769 296.70 | 2.89 | 1.27 (1.17, 1.37)§ | <0.001 | 1.23 (1.14, 1.33)§ | <0.001 |
| Insurance premium (NTD per month) | |||||||
| <18 000 | 2501 | 850 601.33 | 2.94 | 1.00 (reference) | - | 1.00 (reference) | - |
| 18 000–34999 | 4262 | 1 645 882.03 | 2.59 | 0.88 (0.84, 0.92)§ | <0.001 | 0.93 (0.88, 0.98)‡ | 0.005 |
| ≥35 000 | 2524 | 890 837.68 | 2.83 | 0.96 (0.91, 1.02) | 0.181 | 0.98 (0.92, 1.03) | 0.430 |
| Comorbidities | |||||||
| Coronary artery disease | |||||||
| No | 7306 | 2 707 965.44 | 2.7 | 1.00 (reference) | - | 1.00 (reference) | - |
| Yes | 1981 | 679 355.59 | 2.92 | 1.08 (1.03, 1.14)‡ | 0.002 | 1.09 (1.04, 1.15)‡ | 0.001 |
| Stroke | |||||||
| No | 8275 | 3 014 968.45 | 2.74 | 1.00 (reference) | - | 1.00 (reference) | - |
| Yes | 1012 | 372 352.58 | 2.72 | 0.99 (0.93, 1.06) | 0.801 | 0.98 (0.91, 1.05) | 0.509 |
| Peripheral artery disease | |||||||
| No | 9177 | 3 341 777.20 | 2.75 | 1.00 (reference) | - | 1.00 (reference) | - |
| Yes | 110 | 45 543.84 | 2.42 | 0.88 (0.73, 1.06) | 0.184 | 0.83 (0.68, 1.00)† | 0.048 |
| Atrial fibrillation | |||||||
| No | 8548 | 3 118 458.50 | 2.74 | 1.00 (reference) | - | 1.00 (reference) | - |
| Yes | 739 | 268 862.53 | 2.75 | 1.00 (0.93, 1.08) | 0.922 | 1.00 (0.93, 1.08) | 0.955 |
| Heart failure | |||||||
| No | 9002 | 3 264 911.80 | 2.76 | 1.00 (reference) | - | 1.00 (reference) | - |
| Yes | 285 | 122 409.23 | 2.33 | 0.85 (0.75, 0.95)‡ | 0.006 | 0.86 (0.76, 0.97)† | 0.016 |
| Hypertension | |||||||
| No | 3126 | 1 142 074.35 | 2.74 | 1.00 (reference) | - | 1.00 (reference) | - |
| Yes | 6161 | 2 245 246.69 | 2.74 | 1.00 (0.96, 1.05) | 0.927 | 0.98 (0.94, 1.03) | 0.488 |
| Dyslipidemia | |||||||
| No | 3557 | 1 269 263.53 | 2.8 | 1.00 (reference) | - | 1.00 (reference) | - |
| Yes | 5730 | 2 118 057.50 | 2.71 | 0.96 (0.92, 1.01) | 0.087 | 0.94 (0.90, 0.98)∗∗ | 0.003 |
| CKD | |||||||
| No | 8899 | 3 229 270.76 | 2.76 | 1.00 (reference) | - | 1.00 (reference) | - |
| Yes | 388 | 158 050.28 | 2.45 | 0.89 (0.81, 0.99)† | 0.030 | 0.88 (0.79, 0.98)† | 0.019 |
| COPD | |||||||
| No | 8456 | 3 078 773.39 | 2.75 | 1.00 (reference) | - | 1.00 (reference) | - |
| Yes | 831 | 308 547.64 | 2.69 | 0.98 (0.91, 1.05) | 0.617 | 0.96 (0.90, 1.04) | 0.331 |
| Depression | |||||||
| No | 8725 | 3 196 017.31 | 2.73 | 1.00 (reference) | - | 1.00 (reference) | - |
| Yes | 562 | 191 303.72 | 2.94 | 1.08 (0.99, 1.17) | 0.090 | 1.11 (1.01, 1.21)† | 0.025 |
| Schizophrenia | |||||||
| No | 9192 | 3 352 577.92 | 2.74 | 1.00 (reference) | - | 1.00 (reference) | - |
| Yes | 95 | 34 743.11 | 2.73 | 1.00 (0.82, 1.22) | 1.000 | 1.05 (0.85, 1.29) | 0.651 |
| Rheumatoid diseases | |||||||
| No | 9136 | 3 333 248.84 | 2.74 | 1.00 (reference) | - | 1.00 (reference) | - |
| Yes | 151 | 54 072.19 | 2.79 | 1.02 (0.87, 1.20) | 0.811 | 1.04 (0.89, 1.23) | 0.607 |
| Osteoporosis | |||||||
| No | 8878 | 3 229 792.55 | 2.75 | 1.00 (reference) | - | 1.00 (reference) | - |
| Yes | 409 | 157 528.49 | 2.6 | 0.95 (0.86, 1.04) | 0.276 | 0.92 (0.83, 1.02) | 0.097 |
| Alcoholism | |||||||
| No | 9173 | 3 332 001.00 | 2.75 | 1.00 (reference) | - | 1.00 (reference) | - |
| Yes | 114 | 55 320.04 | 2.06 | 0.75 (0.62, 0.90)‡ | 0.002 | 0.79 (0.66, 0.95)† | 0.014 |
| Nicotine dependence | |||||||
| No | 9139 | 3 316 582.26 | 2.76 | 1.00 (reference) | - | 1.00 (reference) | - |
| Yes | 148 | 70 738.77 | 2.09 | 0.76 (0.65, 0.90)‡ | 0.001 | 0.78 (0.67, 0.92)‡ | 0.003 |
| Cancer | |||||||
| No | 8906 | 3 246 943.29 | 2.74 | 1.00 (reference) | - | 1.00 (reference) | - |
| Yes | 381 | 140 377.75 | 2.71 | 0.99 (0.90, 1.10) | 0.910 | 1.16 (1.00, 1.33)† | 0.044 |
| Retinopathy | |||||||
| No | 8083 | 3 106 238.86 | 2.6 | 1.00 (reference) | - | 1.00 (reference) | - |
| Yes | 1204 | 281 082.17 | 4.28 | 1.65 (1.55, 1.75)§ | <0.001 | 1.35 (1.24, 1.47)§ | <0.001 |
| Myopia | |||||||
| No | 9162 | 3 360 536.52 | 2.73 | 1.00 (reference) | - | 1.00 (reference) | - |
| Yes | 125 | 26 784.51 | 4.67 | 1.71 (1.44, 2.04)§ | <0.001 | 1.61 (1.35, 1.92)§ | <0.001 |
| Cataract | |||||||
| No | 6713 | 2 684 783.88 | 2.5 | 1.00 (reference) | - | 1.00 (reference) | - |
| Yes | 2574 | 702 537.16 | 3.66 | 1.47 (1.40, 1.53)§ | <0.001 | 1.39 (1.32, 1.47)§ | <0.001 |
| Uveitis | |||||||
| No | 9219 | 3 369 797.95 | 2.74 | 1.00 (reference) | - | 1.00 (reference) | - |
| Yes | 68 | 17 523.08 | 3.88 | 1.42 (1.12, 1.80)‡ | 0.004 | 1.19 (0.94, 1.51) | 0.157 |
| DR | |||||||
| No | 8662 | 3 247 469.72 | 2.67 | 1.00 (reference) | - | 1.00 (reference) | - |
| Yes | 625 | 139 851.31 | 4.47 | 1.67 (1.54, 1.82)§ | <0.001 | 1.09 (0.97, 1.21) | 0.155 |
| DKD | |||||||
| No | 8615 | 3 156 898.16 | 2.73 | 1.00 (reference) | - | 1.00 (reference) | - |
| Yes | 672 | 230 422.88 | 2.92 | 1.07 (0.99, 1.16) | 0.094 | 1.02 (0.94, 1.11) | 0.6 |
| DN | |||||||
| No | 8652 | 3 210 264.07 | 2.7 | 1.00 (reference) | - | 1.00 (reference) | - |
| Yes | 635 | 177 056.96 | 3.59 | 1.33 (1.23, 1.44)§ | <0.001 | 1.24 (1.15, 1.35)§ | <0.001 |
| CCI score | |||||||
| 0 | 7637 | 2 731 688.32 | 2.8 | 1.00 (reference) | - | 1.00 (reference) | - |
| 1 | 839 | 327 369.00 | 2.56 | 0.92 (0.85, 0.98)† | 0.016 | 0.90 (0.83, 0.96)‡ | 0.003 |
| ≥2 | 811 | 328 263.71 | 2.47 | 0.89 (0.82, 0.95)‡ | 0.001 | 0.82 (0.74, 0.91)§ | <0.001 |
CCI = Charlson Comorbidity Index; CI = confidence interval; CKD = chronic kidney disease; COPD = chronic obstructive pulmonary disease; DKD = diabetic kidney disease; DN = diabetic neuropathy; DPP4i = dipeptidyl peptidase-4 inhibitor; DR = diabetic retinopathy; HR = hazard ratio; IR = incidence rate per 1000 person-years; NTD = New Taiwan Dollar; POAG = primary open angle glaucoma; PY = person-years; T2DM = type 2 diabetes mellitus.
Adjusted HR: adjusted by sex, age, region of dwelling, degree of urbanization, insurance premium, comorbidities, CCI score, and medication.
P value <0.05.
P < 0.01.
P < 0.001.
Figure 2.
The cumulative incidence curves of POAG in DPP4i and non-DPP4i users. Kaplan–Meier curves showing a significantly lower cumulative incidence of POAG in DPP4i users compared with non-DPP4i users. DPP4i = dipeptidyl peptidase-4 inhibitor; POAG = primary open-angle glaucoma.
POAG Risk across Different Subgroups
Subgroup analysis was conducted to assess the consistency of the association between DPP4i use and POAG risk across various demographic and clinical subgroups (Table 4). Dipeptidyl peptidase-4 inhibitor use was linked to a reduced risk of POAG across various demographic factors, including both men (aHR, 0.54; 95% CI, 0.51–0.58; P < 0.001) and women (aHR, 0.51; 95% CI, 0.47–0.55; P < 0.001). Notably, the risk reduction became more pronounced with increasing age (aged 18–39: aHR, 0.56; 95% CI, 0.51–0.62; aged 40–64: aHR, 0.52; 95% CI, 0.47–0.57; aged ≥65 years: aHR, 0.51; 95% CI, 0.47–0.56). Among patients with comorbid conditions, DPP4i users consistently showed a lower risk of POAG, with aHR of 0.53 (95% CI, 0.47–0.59; P < 0.001) for coronary artery disease, 0.57 (95% CI, 0.48–0.66; P < 0.001) for stroke, 0.45 (95% CI, 0.27–0.75; P = 0.002) for peripheral artery disease, 0.58 (95% CI, 0.48–0.69; P < 0.001) for atrial fibrillation, 0.43 (95% CI, 0.32–0.59; P < 0.001) for heart failure, 0.52 (95% CI, 0.49–0.56; P < 0.001) for hypertension, 0.54 (95% CI, 0.50–0.57; P < 0.001) for dyslipidemia, 0.44 (95% CI, 0.34–0.57; P < 0.001) for chronic kidney disease, 0.52 (95% CI, 0.43–0.62; P < 0.001) for chronic obstructive pulmonary disease, 0.53 (95% CI, 0.42–0.65; P < 0.001) for depression, 0.40 (95% CI, 0.26–0.61; P < 0.001) for rheumatoid diseases, 0.47 (95% CI, 0.36–0.60; P < 0.001) for osteoporosis, 0.54 (95% CI, 0.36–0.81; P = 0.003) for nicotine dependence, 0.51 (95% CI, 0.39–0.66; P < 0.001) for cancer, 0.47 (95% CI, 0.41–0.54; P < 0.001) for retinopathy, 0.50 (95% CI, 0.45–0.55; P < 0.001) for cataract, 0.43 (95% CI, 0.35–0.52; P < 0.001) for diabetic retinopathy, 0.49 (95% CI, 0.41–0.59; P < 0.001) for diabetic kidney disease, and 0.43 (95% CI, 0.35–0.52; P < 0.001) for DN. This association persisted even when compared with users of first-generation diabetes medication (including metformin, sulfonylureas, thiazolidinediones, α-glucosidase inhibitors, and glinide), with aHR ranging from 0.42 to 0.55 across various comparisons (all P < 0.001) (Table 5).
Table 4.
Stratified Analysis of Different Covariates for the Risk of POAG
| Variable | POAG in Non-DPP4i User |
POAG in DPP4i User |
Crude |
Adjusted∗ |
||||||
|---|---|---|---|---|---|---|---|---|---|---|
| n | PY | IR | n | PY | IR | HR (95% CI) | P Value | HR (95% CI) | P Value | |
| Sex | ||||||||||
| Female | 2688 | 743 042.61 | 3.62 | 1351 | 799 568.02 | 1.69 | 0.47 (0.44, 0.50)§ | <0.001 | 0.51 (0.47, 0.55)§ | <0.001 |
| Male | 3440 | 885 148.07 | 3.89 | 1808 | 959 562.34 | 1.88 | 0.48 (0.46, 0.51)§ | <0.001 | 0.54 (0.51, 0.58)§ | <0.001 |
| Age | ||||||||||
| 18–39 | 1765 | 532 619.79 | 3.31 | 915 | 564 930.56 | 1.62 | 0.49 (0.45, 0.53)§ | <0.001 | 0.56 (0.51, 0.62)§ | <0.001 |
| 40–64 | 1861 | 493 652.67 | 3.77 | 981 | 535 837.68 | 1.83 | 0.48 (0.45, 0.52)§ | <0.001 | 0.52 (0.47, 0.57)§ | <0.001 |
| ≥65 | 2502 | 601 918.22 | 4.16 | 1263 | 658 362.12 | 1.92 | 0.46 (0.43, 0.49)§ | <0.001 | 0.51 (0.47, 0.56)§ | <0.001 |
| Region of dwelling | ||||||||||
| Northern | 2988 | 744 609.20 | 4.01 | 1616 | 788 241.85 | 2.05 | 0.51 (0.48, 0.54)§ | <0.001 | 0.57 (0.53, 0.62)§ | <0.001 |
| Central | 1409 | 378 788.79 | 3.72 | 671 | 405 615.89 | 1.65 | 0.44 (0.40, 0.48)§ | <0.001 | 0.48 (0.43, 0.53)§ | <0.001 |
| Southern | 1584 | 457 844.23 | 3.46 | 786 | 514 044.62 | 1.53 | 0.44 (0.40, 0.48)§ | <0.001 | 0.49 (0.44, 0.55)§ | <0.001 |
| Eastern | 147 | 46 948.45 | 3.13 | 86 | 51 228.00 | 1.68 | 0.53 (0.41, 0.70)§ | <0.001 | 0.52 (0.38, 0.71)§∗ | <0.001 |
| Degree of urbanization | ||||||||||
| Lowest | 491 | 152 788.39 | 3.21 | 228 | 162 730.07 | 1.40 | 0.43 (0.37, 0.51)§ | <0.001 | 0.50 (0.41, 0.60)§ | <0.001 |
| Median | 2310 | 621 874.46 | 3.71 | 1152 | 680 631.42 | 1.69 | 0.45 (0.42, 0.49)§ | <0.001 | 0.51 (0.47, 0.56)§ | <0.001 |
| Highest | 3327 | 853 527.83 | 3.9 | 1779 | 915 768.87 | 1.94 | 0.50 (0.47, 0.53)§ | <0.001 | 0.54 (0.51, 0.58)§ | <0.001 |
| Insurance premium (NTD per month) | ||||||||||
| <18 000 | 1639 | 409 038.16 | 4.01 | 862 | 441 563.16 | 1.95 | 0.48 (0.45, 0.53)§ | <0.001 | 0.53 (0.48, 0.58)§ | <0.001 |
| 18 000–34999 | 2869 | 789 948.13 | 3.63 | 1393 | 855 933.90 | 1.63 | 0.45 (0.42, 0.48)§ | <0.001 | 0.50 (0.47, 0.54)§ | <0.001 |
| ≥35 000 | 1620 | 429 204.38 | 3.77 | 904 | 461 633.30 | 1.96 | 0.52 (0.48, 0.56)§ | <0.001 | 0.58 (0.53, 0.64)§ | <0.001 |
| Comorbidities | ||||||||||
| Coronary artery disease | ||||||||||
| No | 4814 | 1 304 795.78 | 3.69 | 2492 | 1 403 169.67 | 1.78 | 0.48 (0.46, 0.50)§ | <0.001 | 0.53 (0.50, 0.56)§ | <0.001 |
| Yes | 1314 | 323 394.90 | 4.06 | 667 | 355 960.69 | 1.87 | 0.46 (0.42, 0.50)§ | <0.001 | 0.53 (0.47, 0.59)§ | <0.001 |
| Stroke | ||||||||||
| No | 5492 | 1 449 798.52 | 3.79 | 2783 | 1 565 169.94 | 1.78 | 0.47 (0.45, 0.49)§ | <0.001 | 0.52 (0.50, 0.55)§ | <0.001 |
| Yes | 636 | 178 392.16 | 3.57 | 376 | 193 960.42 | 1.94 | 0.54 (0.47, 0.61)§ | <0.001 | 0.57 (0.48, 0.66)§ | <0.001 |
| Peripheral artery disease | ||||||||||
| No | 6054 | 1 605 724.27 | 3.77 | 3123 | 1 736 052.93 | 1.80 | 0.48 (0.46, 0.50)§ | <0.001 | 0.53 (0.50, 0.56)§ | <0.001 |
| Yes | 74 | 22 466.40 | 3.29 | 36 | 23 077.43 | 1.56 | 0.47 (0.32, 0.70)§ | <0.001 | 0.45 (0.27, 0.75)‡ | 0.002 |
| Atrial fibrillation | ||||||||||
| No | 5651 | 1 498 667.50 | 3.77 | 2897 | 1 619 791.00 | 1.79 | 0.47 (0.45, 0.49)§ | <0.001 | 0.52 (0.50, 0.55)§ | <0.001 |
| Yes | 477 | 129 523.18 | 3.68 | 262 | 139 339.36 | 1.88 | 0.51 (0.44, 0.59)§ | <0.001 | 0.58 (0.48, 0.69)§ | <0.001 |
| Heart failure | ||||||||||
| No | 5942 | 1 569 937.63 | 3.78 | 3060 | 1 694 974.17 | 1.81 | 0.48 (0.45, 0.50)§ | <0.001 | 0.53 (0.50, 0.56)§ | <0.001 |
| Yes | 186 | 58 253.04 | 3.19 | 99 | 64 156.19 | 1.54 | 0.48 (0.38, 0.62)§ | <0.001 | 0.43 (0.32, 0.59)§ | <0.001 |
| Hypertension | ||||||||||
| No | 2087 | 560 151.09 | 3.73 | 1039 | 581 923.25 | 1.79 | 0.48 (0.44, 0.51)§ | <0.001 | 0.54 (0.49, 0.59)§ | <0.001 |
| Yes | 4041 | 1 068 039.58 | 3.78 | 2120 | 1 177 207.11 | 1.80 | 0.47 (0.45, 0.50)§ | <0.001 | 0.52 (0.49, 0.56)§ | <0.001 |
| Dyslipidemia | ||||||||||
| No | 2408 | 617 076.68 | 3.9 | 1149 | 652 186.85 | 1.76 | 0.45 (0.42, 0.48)§ | <0.001 | 0.52 (0.47, 0.56)§ | <0.001 |
| Yes | 3720 | 1 011 113.99 | 3.68 | 2010 | 1 106 943.51 | 1.82 | 0.49 (0.46, 0.52)§ | <0.001 | 0.54 (0.50, 0.57)§ | <0.001 |
| CKD | ||||||||||
| No | 5890 | 1 556 670.38 | 3.78 | 3009 | 1 672 600.38 | 1.80 | 0.47 (0.45, 0.50)§ | <0.001 | 0.53 (0.51, 0.56)§ | <0.001 |
| Yes | 238 | 71 520.30 | 3.33 | 150 | 86 529.98 | 1.73 | 0.51 (0.41, 0.62)§ | <0.001 | 0.44 (0.34, 0.57)§ | <0.001 |
| COPD | ||||||||||
| No | 5575 | 1 479 660.70 | 3.77 | 2881 | 1 599 112.69 | 1.80 | 0.48 (0.46, 0.50)§ | <0.001 | 0.53 (0.50, 0.56)§ | <0.001 |
| Yes | 553 | 148 529.98 | 3.72 | 278 | 160 017.67 | 1.74 | 0.46 (0.40, 0.54)§ | <0.001 | 0.52 (0.43, 0.62)§ | <0.001 |
| Depression | ||||||||||
| No | 5769 | 1 535 066.40 | 3.76 | 2956 | 1 660 950.91 | 1.78 | 0.47 (0.45, 0.49)§ | <0.001 | 0.53 (0.50, 0.56)§ | <0.001 |
| Yes | 359 | 93 124.27 | 3.86 | 203 | 98 179.45 | 2.07 | 0.53 (0.45, 0.63)§ | <0.001 | 0.53 (0.42, 0.65)§ | <0.001 |
| Schizophrenia | ||||||||||
| No | 6071 | 1 610 741.76 | 3.77 | 3121 | 1 741 836.16 | 1.79 | 0.47 (0.45, 0.49)§ | <0.001 | 0.53 (0.50, 0.55)§ | <0.001 |
| Yes | 57 | 17 448.91 | 3.27 | 38 | 17 294.20 | 2.20 | 0.66 (0.44, 1.00)† | 0.048 | 0.83 (0.50, 1.40) | 0.49 |
| Rheumatoid diseases | ||||||||||
| No | 6019 | 1 602 198.30 | 3.76 | 3117 | 1 731 050.54 | 1.80 | 0.48 (0.46, 0.50)§ | <0.001 | 0.53 (0.50, 0.56)§ | <0.001 |
| Yes | 109 | 25 992.38 | 4.19 | 42 | 28 079.82 | 1.50 | 0.36 (0.25, 0.51)§ | <0.001 | 0.40 (0.26, 0.61)§ | <0.001 |
| Osteoporosis | ||||||||||
| No | 5849 | 1 551 350.18 | 3.77 | 3029 | 1 678 442.37 | 1.80 | 0.48 (0.46, 0.50)§ | <0.001 | 0.53 (0.50, 0.56)§ | <0.001 |
| Yes | 279 | 76 840.49 | 3.63 | 130 | 80 687.99 | 1.61 | 0.44 (0.36, 0.54)§ | <0.001 | 0.47 (0.36, 0.60)§ | <0.001 |
| Alcoholism | ||||||||||
| No | 6055 | 1 601 125.16 | 3.78 | 3118 | 1 730 875.84 | 1.80 | 0.47 (0.45, 0.50)§ | <0.001 | 0.53 (0.50, 0.56)§ | <0.001 |
| Yes | 73 | 27 065.51 | 2.7 | 41 | 28 254.52 | 1.45 | 0.54 (0.37, 0.79)‡ | 0.001 | 0.64 (0.40, 1.02) | 0.063 |
| Nicotine dependence | ||||||||||
| No | 6033 | 1 593 951.21 | 3.78 | 3106 | 1 722 631.05 | 1.80 | 0.47 (0.45, 0.50)§ | <0.001 | 0.53 (0.50, 0.56)§ | <0.001 |
| Yes | 95 | 34 239.47 | 2.77 | 53 | 36 499.31 | 1.45 | 0.52 (0.37, 0.72)§ | <0.001 | 0.54 (0.36, 0.81)‡ | 0.003 |
| Cancer | ||||||||||
| No | 5871 | 1 560 193.10 | 3.76 | 3035 | 1 686 750.18 | 1.80 | 0.48 (0.46, 0.50)§ | <0.001 | 0.53 (0.50, 0.56)§ | <0.001 |
| Yes | 257 | 67 997.57 | 3.78 | 124 | 72 380.18 | 1.71 | 0.45 (0.36, 0.55)§ | <0.001 | 0.51 (0.39, 0.66)§ | <0.001 |
| Retinopathy | ||||||||||
| No | 5343 | 1 504 462.05 | 3.55 | 2740 | 1 601 776.82 | 1.71 | 0.48 (0.46, 0.50)§ | <0.001 | 0.54 (0.51, 0.57)§ | <0.001 |
| Yes | 785 | 123 728.63 | 6.34 | 419 | 157 353.54 | 2.66 | 0.42 (0.37, 0.48)§ | <0.001 | 0.47 (0.41, 0.54)§ | <0.001 |
| Myopia | ||||||||||
| No | 6058 | 1 615 745.16 | 3.75 | 3104 | 1 744 791.36 | 1.78 | 0.47 (0.45, 0.49)§ | <0.001 | 0.53 (0.50, 0.55)§ | <0.001 |
| Yes | 70 | 12 445.52 | 5.62 | 55 | 14 339.00 | 3.84 | 0.65 (0.46, 0.93)† | 0.019 | 0.74 (0.49, 1.12) | 0.154 |
| Cataract | ||||||||||
| No | 4412 | 1 294 052.01 | 3.41 | 2301 | 1 390 731.87 | 1.65 | 0.48 (0.46, 0.51)§ | <0.001 | 0.54 (0.51, 0.57)§ | <0.001 |
| Yes | 1716 | 334 138.67 | 5.14 | 858 | 368 398.49 | 2.33 | 0.45 (0.42, 0.49)§ | <0.001 | 0.50 (0.45, 0.55)§ | <0.001 |
| Uveitis | ||||||||||
| No | 6081 | 1 619 666.37 | 3.75 | 3138 | 1 750 131.58 | 1.79 | 0.48 (0.46, 0.50)§ | <0.001 | 0.53 (0.50, 0.56)§ | <0.001 |
| Yes | 47 | 8524.30 | 5.51 | 21 | 8998.78 | 2.33 | 0.43 (0.26, 0.72)‡ | 0.001 | 0.70 (0.38, 1.29) | 0.253 |
| DR | ||||||||||
| No | 5728 | 1 572 520.62 | 3.64 | 2934 | 1 674 949.10 | 1.75 | 0.48 (0.46, 0.50)§ | <0.001 | 0.54 (0.51, 0.57)§ | <0.001 |
| Yes | 400 | 55 670.05 | 7.19 | 225 | 84 181.26 | 2.67 | 0.38 (0.32, 0.44)§ | <0.001 | 0.43 (0.35, 0.52)§ | <0.001 |
| DKD | ||||||||||
| No | 5716 | 1 528 295.61 | 3.74 | 2899 | 1 628 602.55 | 1.78 | 0.47 (0.45, 0.50)§ | <0.001 | 0.53 (0.50, 0.56)§ | <0.001 |
| Yes | 412 | 99 895.07 | 4.12 | 260 | 130 527.81 | 1.99 | 0.48 (0.41, 0.56)§ | <0.001 | 0.49 (0.41, 0.59)§ | <0.001 |
| DN | ||||||||||
| No | 5713 | 1 553 963.27 | 3.68 | 2939 | 1 656 300.80 | 1.77 | 0.48 (0.46, 0.50)§ | <0.001 | 0.54 (0.51, 0.57)§ | <0.001 |
| Yes | 415 | 74 227.40 | 5.59 | 220 | 102 829.56 | 2.14 | 0.38 (0.32, 0.45)§ | <0.001 | 0.43 (0.35, 0.52)§ | <0.001 |
| CCI score | ||||||||||
| 0 | 5061 | 1 316 456.12 | 3.84 | 2576 | 1 415 232.20 | 1.82 | 0.47 (0.45, 0.50)§ | <0.001 | 0.53 (0.50, 0.56)§ | <0.001 |
| 1 | 542 | 155 780.52 | 3.48 | 297 | 171 588.48 | 1.73 | 0.49 (0.43, 0.57)§ | <0.001 | 0.55 (0.47, 0.65)§ | <0.001 |
| ≥2 | 525 | 155 954.04 | 3.37 | 286 | 172 309.67 | 1.66 | 0.49 (0.42, 0.56)§ | <0.001 | 0.54 (0.45, 0.64)§ | <0.001 |
CCI = Charlson Comorbidity Index; CI = confidence interval; CKD = chronic kidney disease; COPD = chronic obstructive pulmonary disease; DKD = diabetic kidney disease; DN = diabetic neuropathy; DPP4i = dipeptidyl peptidase-4 inhibitor; DR = diabetic retinopathy; HR = hazard ratio; IR = incidence rate per 1000 person-years; NTD = New Taiwan Dollar; POAG = primary open angle glaucoma; PY = person-years; T2DM = type 2 diabetes mellitus.
Adjusted HR: adjusted by sex, age, region of dwelling, degree of urbanization, insurance premium, comorbidities, CCI score, and medication.
P value <0.05.
P < 0.01.
P < 0.001.
Table 5.
Hazard Ratios of POAG among T2DM with Different DM Drug User
| Variable | POAG |
Crude |
Adjusted∗ |
||||
|---|---|---|---|---|---|---|---|
| n | PY | IR | HR (95% CI) | P Value | HR (95% CI) | P Value | |
| Metformin user | 5788 | 1 564 830.05 | 3.70 | 1.00 (reference) | - | 1.00 (reference) | - |
| DPP4i user | 3159 | 1 759 130.36 | 1.80 | 0.48 (0.46, 0.50)§ | <0.001 | 0.55 (0.53, 0.58)§ | <0.001 |
| SU user | 5048 | 1 249 983.76 | 4.04 | 1.00 (reference) | - | 1.00 (reference) | - |
| DPP4i user | 3159 | 1 759 130.36 | 1.80 | 0.44 (0.42, 0.46)§ | <0.001 | 0.45 (0.43, 0.47)§ | <0.001 |
| TZD user | 1518 | 333 211.13 | 4.56 | 1.00 (reference) | - | 1.00 (reference) | - |
| DPP4i user | 3159 | 1 759 130.36 | 1.80 | 0.39 (0.37, 0.42)§ | <0.001 | 0.42 (0.39, 0.45)§ | <0.001 |
| AGI user | 1321 | 284 857.21 | 4.64 | 1.00 (reference) | - | 1.00 (reference) | - |
| DPP4i user | 3159 | 1 759 130.36 | 1.80 | 0.39 (0.36, 0.41)§ | <0.001 | 0.43 (0.40, 0.46)§ | <0.001 |
| Glinide user | 187 | 55 475.45 | 3.37 | 1.00 (reference) | - | 1.00 (reference) | - |
| DPP4i user | 3159 | 1 759 130.36 | 1.80 | 0.50 (0.43, 0.58)§ | <0.001 | 0.53 (0.46, 0.62)§ | <0.001 |
AGI = α-glucosidase inhibitor; CI = confidence interval; DM = diabetes mellitus; DPP4i = dipeptidyl peptidase-4 inhibitor; HR = hazard ratio; IR = incidence rate per 1000 person-years; POAG = primary open angle glaucoma; PY = person-years; SU = sulfonylurea; T2DM = type 2 diabetes mellitus; TZD = thiazolidinedione.
Adjusted HR: adjusted by sex, age, region of dwelling, degree of urbanization, insurance premium, comorbidities, CCI score, and medication.
P < 0.001.
DPP4i Use and Risk of NTG
Similarly, DPP4i users demonstrated a lower risk of NTG, with an incidence rate of 0.41 per 1000 person-years compared with 0.74 in non-DPP4i users. The aHR for NTG among DPP4i users was 0.55 (95% CI, 0.50–0.62; P < 0.001) (Table 6). Figure 3 illustrates the cumulative incidence curve of NTG, with a consistently lower trend for DPP4i users throughout the study period (log-rank P < 0.001).
Table 6.
Hazard Ratios of NTG among T2DM with DPP4i and Non-DPP4i User
| Variable | NTG |
Crude |
Adjusted∗ |
||||
|---|---|---|---|---|---|---|---|
| n | PY | IR | HR (95% CI) | P Value | HR (95% CI) | P Value | |
| Non-DPP4i user | 1220 | 1 655 251.00 | 0.74 | 1.00 (reference) | - | 1.00 (reference) | - |
| DPP4i user | 734 | 1 768 746.13 | 0.41 | 0.56 (0.51, 0.61)§ | <0.001 | 0.55 (0.50, 0.62)§ | <0.001 |
CI = confidence interval; DPP4i = dipeptidyl peptidase-4 inhibitor; HR = hazard ratio; IR = incidence rate per 1000 person-years; NTG = normal tension glaucoma; PY = person-years; T2DM = type 2 diabetes mellitus.
Adjusted HR: adjusted by sex, age, region of dwelling, degree of urbanization, insurance premium, comorbidities, CCI score, and medication.
P < 0.001.
Figure 3.
The cumulative incidence curves of NTG in DPP4i and non-DPP4i users. Kaplan–Meier curves showing a consistently lower cumulative incidence of NTG in DPP4i users compared with non-DPP4i users. DPP4i = dipeptidyl peptidase-4 inhibitor; NTG = normal tension glaucoma.
Discussion
In this large nationwide retrospective cohort study, we found that DPP4i use was associated with a significantly lower risk of developing POAG and NTG in patients with T2DM. This association remained consistent across various subgroups and was evident in cumulative incidence analyses. These findings suggest that DPP4i may confer benefits beyond glycemic control, possibly influencing optic nerve health through multiple mechanisms.
Incretins Action and Beyond
The beneficial effects associated with DPP4i on neurodegenerative diseases are likely multifactorial, encompassing both incretin-dependent and incretin-independent pathways. Previous studies have primarily focused on incretin-related mechanisms. For example, in a rat model of cisplatin-induced neurotoxicity, DPP4i (vildagliptin) significantly increased levels of adenosine monophosphate-activated protein kinase, protein kinase B, and cyclic adenosine monophosphate response element-binding protein—key components of the incretin signaling pathway—thereby alleviating neuroinflammation and cognitive decline.15 Similarly, Li et al11 reported that DPP4i (sitagliptin) elevated CNS incretin levels in a rodent model of Parkinson disease, reducing inflammatory markers in the brain.
Li et al further demonstrated that DPP4i, in combination with GLP-1 and glucose-dependent insulinotropic polypeptide, provided superior neuroprotective effects in both rat and human neuronal cultures compared with glucose-dependent insulinotropic polypeptide or GLP-1 alone, suggesting potential synergistic mechanisms beyond incretin signaling.11 These findings underscore the complex interplay of pathways in modulating inflammation as well as promoting neuroregeneration, which may contribute to the observed association between DPP4i use and glaucoma.17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27
Antioxidant Effects and Ocular Connective Tissue Remodeling
Hyperglycemia in diabetic retinas induces excessive mitochondrial production of ROS, leading to the formation of methylglyoxal (MG) and advanced glycation end products (AGEs).18 In turn, MG and AGEs contribute to mitochondrial dysfunction and oxidative stress, both of which are implicated in the development and progression of glaucoma.19 Moreover, MG and AGEs increase the stiffness of collagen-rich tissues through protein cross-linking, affecting structures such as the lamina cribrosa and peripapillary sclera. This reduced elasticity compromises the tissue's load-bearing capacity, impairs neurovascular function, and heightens the vulnerability of retinal ganglion cells.20,21 Notably, Hopkins et al21 argue that tissue stiffness, rather than absolute intraocular pressure or translaminar pressure, drives glaucomatous cupping.
Dipeptidyl peptidase-4 inhibitors exhibit both direct and indirect antioxidant effects by reducing ROS production and disrupting the feedback loop through inhibition of AGEs and their receptor interactions.22 In diabetic animal models, DPP4is (linagliptin) significantly decreased ROS and MG levels. The treatment also preserved cell numbers in the ganglion cell layer, maintaining levels comparable to nondiabetic controls.12 These antioxidative properties, combined with their ability to remodel ocular connective tissue, may explain the associations observed in our study. This suggests that DPP4i may mitigate some of the oxidative toxicity as well as structural changes associated with glaucoma progression.
Mitigating Glutamate-Induced Excitotoxicity
Excessive glutamate, the primary excitatory neurotransmitter in the CNS, can cause excitotoxicity, contributing to the pathogenesis of various neurodegenerative conditions, including Alzheimer disease, Parkinson disease, and glaucoma.23, 24, 25 Glutamate is typically cleared by astrocytic transporters such as glutamate-aspartate transporter (GLAST), which converts it to nontoxic glutamine via glutamine synthetase.23,25 However, diabetic retinas show reduced GLAST expression and elevated glutamate levels, exacerbating excitotoxic damage.26,27
Dipeptidyl peptidase-4 inhibitors have demonstrated efficacy in reducing glutamate concentrations and preserving GLAST levels in diabetic models. For instance, topical DPP4i (saxagliptin and sitagliptin) treatment significantly reduced retinal glutamate levels to those comparable with controls, and prevented GLAST downregulation in diabetic rat model, highlighting its potential neuroprotective effects independent of glycemic control.26 These mechanisms may contribute to the reduced risk of glaucoma observed among DPP4i users in our study.
The strengths of this study include its large sample size, robust matching methods, and comprehensive consideration of potential confounders. However, several limitations should be acknowledged. Our study relied on administrative claims data, which lacked detailed clinical metrics such as intraocular pressure, family history, genetic predisposition, and corneal thickness. Additionally, researchers using health insurance databases often encounter coding bias. This could potentially influence the identification of cases. Future prospective studies are warranted to explore the mechanisms underlying DPP4i's effects on glaucoma risk.
In conclusion, our nationwide cohort study provides evidence that DPP4i use is associated with a lower risk of developing POAG and NTG in adults with T2DM. These findings highlight the potential benefits of DPP4i for glaucoma prevention strategies in clinical practice.
Acknowledgments
The authors thank the Health Data Science Center, China Medical University Hospital, for their administrative, technical, and funding support. The authors also acknowledge Taichung Veterans General Hospital for its funding support.
Manuscript no. XOPS-D-24-00588.
Footnotes
Supplemental material available atwww.ophthalmologyscience.org.
Disclosure(s):
All authors have completed and submitted the ICMJE disclosures form.
The author has made the following disclosures:
C.-H.W.: Consultant – Novo Nordisk Inc.; Support for attending meetings and/or travel – Gilead Sciences, Inc. (Food and Beverage for a dinner lecture); Participation on a Data Safety Monitoring Board or Advisory Board – Novo Nordisk Inc.
This study is partially supported by the Taiwan Ministry of Health and Welfare Clinical Trial Center (MOHW113-TDU-B-212-114009). The funders had no role in the study design, data collection and analysis, publication decision, or manuscript preparation. No additional external funding was received for this study.
HUMAN SUBJECTS: No human subjects were included in this study. Ethical approval for the study was granted by the Research Ethics Committee of China Medical University Hospital (CMUH110-REC1-038[CR-4]).
No animal subjects were used in this study.
Author Contributions:
Conception and design: Lu, Tien, Cheng, Hsieh, Cho, Tsai, Weng, Heng-Jun Lin, Hui-Ju Lin, Wang, Chou
Data collection: Tien, Cheng, Hsieh, Cho, Tsai, Weng, Chou
Obtained funding: Chou
Analysis and interpretation: Lu, Heng-Jun Lin, Hui-Ju Lin, Chou
Overall responsibility: Lu, Chou
Supplementary Data
References
- 1.Zhao D., Cho J., Kim M.H., et al. Diabetes, fasting glucose, and the risk of glaucoma: a meta-analysis. Ophthalmology. 2015;122:72–78. doi: 10.1016/j.ophtha.2014.07.051. [DOI] [PubMed] [Google Scholar]
- 2.McMonnies C.W. Glaucoma history and risk factors. J Optom. 2017;10:71–78. doi: 10.1016/j.optom.2016.02.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Ju W.-K., Perkins G.A., Kim K.-Y., et al. Glaucomatous optic neuropathy: mitochondrial dynamics, dysfunction and protection in retinal ganglion cells. Prog Retin Eye Res. 2023;95 doi: 10.1016/j.preteyeres.2022.101136. [DOI] [PubMed] [Google Scholar]
- 4.Rangaswami J., Bhalla V., Boer IHd, et al. Cardiorenal protection with the newer antidiabetic agents in patients with diabetes and chronic kidney disease: a scientific statement from the American heart association. Circulation. 2020;142:e265–e286. doi: 10.1161/CIR.0000000000000920. [DOI] [PubMed] [Google Scholar]
- 5.Gilstrap L.G., Blair R.A., Huskamp H.A., et al. Assessment of second-generation diabetes medication initiation among medicare enrollees from 2007 to 2015. JAMA Netw Open. 2020;3 doi: 10.1001/jamanetworkopen.2020.5411. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Lin H.-C., Stein J.D., Nan B., et al. Association of geroprotective effects of metformin and risk of open-angle glaucoma in persons with diabetes mellitus. JAMA Ophthalmol. 2015;133:915–923. doi: 10.1001/jamaophthalmol.2015.1440. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Sterling J., Hua P., Dunaief J.L., et al. Glucagon-like peptide 1 receptor agonist use is associated with reduced risk for glaucoma. Br J Ophthalmol. 2023;107:215–220. doi: 10.1136/bjophthalmol-2021-319232. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Niazi S., Gnesin F., Thein A.S., et al. Association between glucagon-like peptide-1 receptor agonists and the risk of glaucoma in individuals with type 2 diabetes. Ophthalmology. 2024;136:1056–1063. doi: 10.1016/j.ophtha.2024.03.004. [DOI] [PubMed] [Google Scholar]
- 9.Shao S.C., Su Y.C., Lai E.C., et al. Association between sodium glucose co-transporter 2 inhibitors and incident glaucoma in patients with type 2 diabetes: a multi-institutional cohort study in Taiwan. Diabetes Metab. 2022;48 doi: 10.1016/j.diabet.2022.101318. [DOI] [PubMed] [Google Scholar]
- 10.Drucker D.J., Nauck M.A. The incretin system: glucagon-like peptide-1 receptor agonists and dipeptidyl peptidase-4 inhibitors in type 2 diabetes. Lancet. 2006;368:1696–1705. doi: 10.1016/S0140-6736(06)69705-5. [DOI] [PubMed] [Google Scholar]
- 11.Li Y., Vaughan K.L., Wang Y., et al. Sitagliptin elevates plasma and CSF incretin levels following oral administration to nonhuman primates: relevance for neurodegenerative disorders. GeroScience. 2024;46:4397–4414. doi: 10.1007/s11357-024-01120-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Dietrich N., Kolibabka M., Busch S., et al. The DPP4 inhibitor linagliptin protects from experimental diabetic retinopathy. PLoS One. 2016;11(12) doi: 10.1371/journal.pone.0167853. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Nauck M.A., El-Ouaghlidi A. The therapeutic actions of DPP-IV inhibition are not mediated by glucagon-like peptide-1. Diabetologia. 2005;48:608–611. doi: 10.1007/s00125-005-1704-8. [DOI] [PubMed] [Google Scholar]
- 14.Vella A. Mechanism of action of DPP-4 inhibitors--new insights. J Clin Endocrinol Metab. 2012;97:2626–2628. doi: 10.1210/jc.2012-2396. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Mahmoud A.M.A., Mantawy E.M., Wahdan S.A., et al. Vildagliptin restores cognitive function and mitigates hippocampal neuronal apoptosis in cisplatin-induced chemo-brain: imperative roles of AMPK/Akt/CREB/BDNF signaling cascades. Biomed Pharmacother. 2023;159 doi: 10.1016/j.biopha.2023.114238. [DOI] [PubMed] [Google Scholar]
- 16.Ma M., Hasegawa Y., Koibuchi N., et al. DPP-4 inhibition with linagliptin ameliorates cognitive impairment and brain atrophy induced by transient cerebral ischemia in type 2 diabetic mice. Cardiovasc Diabetol. 2015;14:54. doi: 10.1186/s12933-015-0218-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Eng K., Zebardast N., Boland M.V., et al. Sodium-glucose cotransporter 2 inhibitors for the primary prevention of glaucoma in patients with type 2 diabetes: a target trial emulation. Am J Ophthalmol. 2025;271:286–298. doi: 10.1016/j.ajo.2024.10.029. [DOI] [PubMed] [Google Scholar]
- 18.Giacco F., Brownlee M. Oxidative stress and diabetic complications. Circ Res. 2010;107:1058–1070. doi: 10.1161/CIRCRESAHA.110.223545. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Abu-Amero K.K., Morales J., Bosley T.M. Mitochondrial abnormalities in patients with primary open-angle glaucoma. Invest Ophthalmol Vis Sci. 2006;47:2533–2541. doi: 10.1167/iovs.05-1639. [DOI] [PubMed] [Google Scholar]
- 20.Spoerl E., Boehm A.G., Pillunat L.E. The influence of various substances on the biomechanical behavior of lamina cribrosa and peripapillary sclera. Invest Ophthalmol Vis Sci. 2005;46:1286–1290. doi: 10.1167/iovs.04-0978. [DOI] [PubMed] [Google Scholar]
- 21.Hopkins A.A., Murphy R., Irnaten M., et al. The role of lamina cribrosa tissue stiffness and fibrosis as fundamental biomechanical drivers of pathological glaucoma cupping. Am J Physiology-Cell Physiol. 2020;319:C611–C623. doi: 10.1152/ajpcell.00054.2020. [DOI] [PubMed] [Google Scholar]
- 22.Kröller-Schön S., Knorr M., Hausding M., et al. Glucose-independent improvement of vascular dysfunction in experimental sepsis by dipeptidyl-peptidase 4 inhibition. Cardiovasc Res. 2012;96:140–149. doi: 10.1093/cvr/cvs246. [DOI] [PubMed] [Google Scholar]
- 23.Pajarillo E., Rizor A., Lee J., et al. The role of astrocytic glutamate transporters GLT-1 and GLAST in neurological disorders: potential targets for neurotherapeutics. Neuropharmacology. 2019;161 doi: 10.1016/j.neuropharm.2019.03.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Martin K.R., Levkovitch-Verbin H., Valenta D., et al. Retinal glutamate transporter changes in experimental glaucoma and after optic nerve transection in the rat. Invest Ophthalmol Vis Sci. 2002;43:2236–2243. [PubMed] [Google Scholar]
- 25.Ishikawa M., Yoshitomi T., Zorumski C.F., Izumi Y. Downregulation of glutamine synthetase via GLAST suppression induces retinal axonal swelling in a rat ex vivo hydrostatic pressure model. Invest Ophthalmol Vis Sci. 2011;52:6604–6616. doi: 10.1167/iovs.11-7375. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Hernández C., Bogdanov P., Solà-Adell C., et al. Topical administration of dipeptidyl peptidase IV (DPP-IV) inhibitors prevents retinal neurodegeneration in experimental diabetes. Invest Ophthalmol Vis Sci. 2017;58:5795. doi: 10.1007/s00125-017-4388-y. [DOI] [PubMed] [Google Scholar]
- 27.El-Sahar A.E., Safar M.M., Zaki H.F., et al. Sitagliptin attenuates transient cerebral ischemia/reperfusion injury in diabetic rats: implication of the oxidative-inflammatory-apoptotic pathway. Life Sci. 2015;126:81–86. doi: 10.1016/j.lfs.2015.01.030. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.