Abstract
Objective
This study aimed to understand the impact of sickle cell disease (SCD) or sickle cell trait (SCT) on diabetic retinopathy and retinopathy treatment burden.
Design
Cross-sectional study.
Subjects and Controls
Utilizing the American Academy of Ophthalmology IRIS® Registry (Intelligent Research in Sight), 3742 patients with comorbid diabetes mellitus (DM) and either SCD or SCT were included in the analytic sample. A race-stratified group of 3742 patients with DM and no known SCD was included as controls.
Methods
Data analysis was performed using R (R Project 4.2.0). Descriptive statistics summarized demographic data among patients with comorbid SCD or SCT and among controls. Chi-square tests compared clinical outcomes (proliferative diabetic retinopathy [PDR], diabetic macular edema [DME]) and ocular procedure frequency across patients with DM and SCD or SCT and across controls. Multivariate logistic regression models examined the likelihood of clinical outcomes and ocular procedures across patients with and without comorbid SCD or SCT while adjusting for insurance, smoking status, and demographic factors.
Main Outcome Measures
Frequency of clinical outcomes and ocular procedures.
Results
All measured clinical outcomes were more frequently documented among patients with DM and either SCD or SCT as compared with those with DM alone. Of the 7 ocular procedures studied, 5 were more common among the DM with SCT or SCD group as compared with controls (P < 0.001). On regression analysis, DM patients with comorbid SCD or SCT had 7.36 and 4.22 times greater odds of PDR and DME, respectively (confidence interval for PDR: 6.36–8.54, confidence interval for DME: 3.68–4.85).
Conclusions
This analysis suggests that DM patients with comorbid SCD or SCT have an increased likelihood of DM-related microvascular pathology and a higher ocular treatment burden. Prospective studies are needed to further characterize the relationship between DM and SCT or SCD.
Financial Disclosure(s)
Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.
Keywords: Diabetic retinopathy, Sickle cell trait, Sickle cell disease, Diabetes mellitus, IRIS® Registry
Diabetes mellitus (DM) is the leading cause of new cases of blindness in adults aged 18 to 64 years.1 Diabetes mellitus is estimated to affect 11.6% of the US population (38.4 million people) and is the leading cause of new cases of blindness in the United States.1 The Wisconsin Epidemiologic Study of Diabetic Retinopathy Cohort found that after 20 years, 99% of patients with type 1 DM and 60% of patients with type 2 DM showed some degree of the end-organ manifestation of retinopathy.2 As of 2021, the National Diabetes Surveillance System of the Centers for Disease Control and Prevention reports that the age-adjusted percentage of non-Hispanic Black adults who are diagnosed with diabetes is 12.7%.1 Among this group of non-Hispanic Black adults with diabetes, 18.4% of individuals experience visual impairment.1
Sickle cell disease (SCD) and sickle cell trait (SCT), the disease carrier state, affect 1 out of 365 Black American newborns and 1 out of 13 Black American newborns, respectively.3 Sickle cell trait is considered a benign condition, as individuals with SCT are generally asymptomatic with little systemic morbidity. However, SCT may rarely manifest as potentially sight-threatening retinopathy.4, 5, 6 Further, although it is known that both SCD and DM can each independently cause ischemic damage to the retinal microvasculature, resulting in retinopathy, there is little information available about how the 2 disease processes affect the retina when present concurrently, particularly among the affected US patients.7 The available literature on comorbid SCD or SCT and DM reports conflicting results regarding these conditions' combined effect on retinopathy. For example, previous work has found SCT to worsen oxidative stress and contribute to abnormal blood rheology and vascular dysfunction in DM.6,8, 9, 10, 11 Another study reported a possible protective effect of SCT in diabetic retinopathy (DR), with the lack of SCT conferring a higher risk of proliferative diabetic retinopathy (PDR).12 In 2016, a single-center retrospective study from Saudi Arabia showed a protective effect of SCT on the development and progression of DR among 100 eyes of 100 patients.13
The aim of this study is to better clarify existing conflicting literature to analyze the pathologic effects of DM and comorbid sickle cell (SC) with either SCD or SCT on retinopathy in US patients. This topic is particularly relevant to the care of Black American patients who account for approximately 825 000 patients with DR and have a higher rate of age-adjusted blindness when compared with White patients and in whom SCT is most common.14,15 Through utilization of the American Academy of Ophthalmology IRIS® Registry (Intelligent Research in Sight), the world's largest comprehensive eye disease clinical registry of selected deidentified electronic health record data for approximately 24% of the US population, this study evaluates the difference in rates of sight-threatening retinopathy and treatment burden between a large cohort of patients with comorbid DM and SC compared with matched controls.
Methods
The IRIS Registry database was used to first identify patients with documented type 1 or type 2 DM between 2013 and 2021. These patients with DM were then stratified based on their SC diagnosis status, identifying individuals with a diagnosis of either SCD or SCT and those without any SC diagnosis. Among those without any SC diagnosis, a race-stratified random sample was selected (Fig 1).
Figure 1.
Inclusion flowchart of the sample identified within the Intelligent Research in Sight (IRIS) database.
From these patient groups within the IRIS Registry, data extracted for analysis included demographic information, insurance status, census region, smoking status, diabetic ocular complications (PDR, diabetic macular edema [DME]), and ocular procedures (focal laser surgery, pan-retinal photocoagulation, vitrectomy, intravitreal injection, nonretinal laser surgery [capsulotomy, LASIK, and iridotomy procedures, among others], and membrane peel) that were performed between the dates of January 1, 2013, and December 31, 2021.
Data analysis was performed using R16 (R Project 4.2.0). Descriptive statistics were used to summarize demographic data among patients with DM and SCD or SCT (DM-SC) and among those with DM but without SC (DM-non-SC). Chi-square tests were used to compare categorical variables related to diabetic ocular outcomes, diabetic disease indicators, and ocular procedures among DM-SC and DM-non-SC groups. Multivariate logistic regression models were implemented to examine the likelihood of diabetic complications, systemic comorbidities, and ocular procedures across DM-SC and DM-non-SC groups. All multivariate logistic regression analyses were adjusted for demographic factors, insurance status, census region, and smoking status. Descriptive statistics were used to summarize visual acuity (VA) data obtained closest to the index date (date of initial DM diagnosis) per eye in the DM-SC and DM-non-SC patient groups using American Snellen fraction subcategories of ≥20/40, 20/40 to 20/60, <20/60, count finger, hand motion , light perception, and no light perception. T tests were used to compare the logarithm of the minimum angle of resolution VA means per eye of the DM-SC and DM-non-SC patient groups. Sensitivity analyses were performed by applying a propensity score weighting approach (weight was truncated at the 99th percentile) to balance the baseline characteristics of the 2 groups, and the previously described analysis was performed.
This study was reviewed by the Johns Hopkins School of Medicine institutional review board (IRB00367889) and was conducted in accordance with the Declaration of Helsinki. Given the deidentified nature of the patient data, we did not obtain written consent for enrollment in this study.
Results
Within the IRIS Registry, 12 856 426 patients were identified with documented type 1 or type 2 DM. Of this cohort, 3742 patients had an additional diagnosis of SCD or SCT. Demographic information of DM-SC patients was compared to a race-stratified random sample of 3742 DM-non-SC patients (Table 1). The mean age of patients with DM-SC was 53.5 years, while the mean age of DM-non-SC patients was greater at 63.4 years. The DM-SC group was of predominantly female (64.8%) and Black or African American (72.2%) patients, while the sample of DM-non-SC was more evenly distributed across sex (52.0% females) and race. The percentage of individuals with Medicaid insurance was greater within the DM-SC group (12.2%) than within the DM-non-SC group (6.0%). As compared with the DM-non-SC group, the DM-SC group included a greater percentage of individuals living in the Southern region of the United States (56.3% vs. 36.8%) and fewer individuals living in the Western United States (4.7% vs. 18.1%). Among DM-SC patients, 64% were lifelong nonsmokers, as compared with 56.3% of DM-non-SC patients.
Table 1.
Patient Demographics
| Demographic Information | DM with SC | DM without SC |
|---|---|---|
| N | 3352 | 3527 |
| Age, mean (SD) | 57.50 (15.65) | 59.92 (14.46) |
| Sex at birth (%) | ||
| Female | 2036 (60.7) | 2048 (58.1) |
| Male | 1298 (38.7) | 1467 (41.6) |
| Not reported | 19 (0.6) | 12 (0.3) |
| Race (%) | ||
| Asian | ||
| Black or African American | 1564 (46.7) | 1381 (39.1) |
| Hispanic | 244 (7.3) | 270 (7.6) |
| Native American or Alaska native | 62 (1.8) | 88 (2.5) |
| Native Hawaiian or Pacific Islander | ||
| Unknown | 671 (20.0) | 679 (19.2) |
| White | 812 (24.2) | 1111 (31.5) |
| Insurance (%) | ||
| Private | 931 (27.8) | 974 (27.6) |
| Government | 86 (2.6) | 70 (2.0) |
| Medicaid | 310 (9.3) | 290 (8.2) |
| Medicare fee-for-service | 1355 (40.4) | 1419 (40.2) |
| Medicare managed | 311 (9.3) | 364 (10.3) |
| Military | 40 (1.2) | 40 (1.1) |
| Unknown/missing | 320 (9.5) | 371 (10.5) |
| US region (%) | ||
| Midwest | 492 (14.7) | 564 (16.0) |
| Northeast | 387 (11.5) | 423 (12.0) |
| South | 1736 (51.8) | 1599 (45.3) |
| Unknown | 495 (14.8) | 526 (14.9) |
| West | 243 (7.3) | 415 (11.8) |
| Smoking status (%) | ||
| Never smoker | 2060 (61.4) | 2067 (58.6) |
| Former smoker | 700 (20.9) | 803 (22.8) |
| Unknown | 166 (4.9) | 204 (5.8) |
| Active smoker | 427 (12.7) | 453 (12.8) |
DM = diabetes mellitus; SC = sickle cell disease or trait; SD = standard deviation.
Diabetic eye disease, including PDR and DME, was more frequently documented among the DM-SC group as compared with those with DM alone (Table 2); 37.1% and 31.9% of the DM-SC patients had PDR and DME, respectively, as compared with 8.4% and 11.1% within the DM-non-SC group (P < 0.001 for both conditions). Most ocular procedure code documentation was more common among DM-SC patients as compared with the DM-non-SC group (P < 0.001). Rates of documented nonretinal laser surgery were similar across groups (P > 0.05).
Table 2.
Diabetic Ocular Outcomes, Diabetic Disease Indicators, and Ocular Procedures among Diabetic Patients with and without Sickle Cell
| DM with SC, n (%) | DM without SC, n (%) | P | |
|---|---|---|---|
| Total | 3352.5 | 3527.3 | |
| Proliferative diabetic retinopathy | 1242.9 (37.1) | 295.2 (8.4) | <0.001 |
| Diabetic macular edema | 1070.9 (31.9) | 391.8 (11.1) | <0.001 |
| Panretinal photocoagulation | 695.4 (20.7) | 131.6 (3.7) | <0.001 |
| Intravitreal injection | 679.0 (20.3) | 297.0 (8.4) | <0.001 |
| Vitrectomy | 393.1 (11.7) | 89.9 (2.5) | <0.001 |
| Focal laser | 229.8 (6.9) | 127.1 (3.6) | <0.001 |
| Membrane peel | 160.2 (4.8) | 25.0 (0.7) | <0.001 |
DM = diabetes mellitus; SC = sickle cell disease or trait.
Separate multivariate logistic regression models assessed the likelihood of diabetic complications, procedures, and comorbidities while controlling for age, sex, race and ethnicity, insurance, region, and smoking status (Table 3). As compared with DM-non-SC patients, those with SC had 7.36 and 4.22 greater odds of PDR and DME, respectively (confidence interval for PDR: 6.36–8.54; confidence interval for DME: 3.68–4.85). In separate models, the likelihood of documented ocular procedures of focal laser surgery, intravitreal injection, vitrectomy, membrane peel, and panretinal photocoagulation was greater among the DM-SC group as compared with those with DM alone (P < 0.001).
Table 3.
Multivariate Logistic Regression Analysis of Ocular and Systemic Factors among DM Patients with SC vs. without SC
| Model∗ | Covariate | OR (95% CI) |
|---|---|---|
| 1 | Hypertension | 4.24 (3.77–4.78) |
| 2 | Proliferative diabetic retinopathy | 7.36 (6.36–8.54) |
| 3 | Diabetic macular edema | 4.22 (3.68–4.85) |
| 4 | Panretinal photocoagulation | 6.75 (5.59–8.22) |
| 5 | Intravitreal injection | 2.92 (2.53–3.37) |
| 6 | Hyperlipidemia | 2.06 (1.73–2.47) |
| 7 | Nephropathy | 12.03 (8.89–16.67) |
| 8 | Vitrectomy | 5.24 (4.19–6.62) |
| 9 | Focal laser | 2.01 (1.62–2.51) |
| 10 | Membrane peel | 6.89 (4.65–10.61) |
CI = confidence interval; DM = diabetes mellitus; OR = odds ratio; SC = sickle cell disease or trait.
All models controlled for age, sex, race, insurance, region, and smoking status.
For both the right and left eyes, a comparatively higher percentage of patients with VA within the <20/60, count finger, hand motion, light perception, and no light perception subcategories was noted among the DM-SC group, with comparatively higher percentages of patients with VA within the ≥20/40 and 20/40 to 20/60 subcategories being noted among the DM-non-SC group (Table 4). T tests of the mean logarithm of the minimum angle of resolution VA for each eye within the DM-SC group (right eye 0.34, standard deviation [SD] 0.59; left eye 0.35, SD 0.59) and DM-non-SC group (right eye 0.27, SD 0.42; left eye 0.27, SD 0.43) indicated generally worse vision of both eyes in DM-SC patients compared with DM-non-SC patients (P < 0.001).
Table 4.
Visual Acuity among DM Patients with SC vs. without SC
| DM with SC | DM without SC | P Value | |
|---|---|---|---|
| Total | 3742 | 3742 | |
| OD, mean (SD) | 0.34 (0.59) | 0.27 (0.42) | <0.001 |
| OS, mean (SD) | 0.35 (0.59) | 0.27 (0.43) | <0.001 |
| OD | <0.001 | ||
| ≥20/40 | 2161 (63.1) | 2214 (64.8) | |
| 20/40–20/60 | 664 (19.4) | 699 (20.4) | |
| ≤20/200 | 601 (16.1) | 506 (13.5) | |
| OS | <0.001 | ||
| ≥20/40 | 2124 (61.7) | 2223 (64.5) | |
| 20/40–20/60 | 669 (19.4) | 736 (21.4) | |
| ≤20/200 | 652 (17.4) | 486 (13.0) |
DM = diabetes mellitus; OD = right eye, OS = left eye; SC = sickle cell disease or trait; SD = standard deviation.
Discussion
The present findings contribute to our understanding of the relationship between SC and DM, demonstrating the additional burden of disease and intervention that DM patients may experience with a diagnosis of SCD or SCT carrier status. Previous literature examining DM-associated ocular pathology and SC has reported mixed results, with some studies suggesting higher8 or comparable9,12 rates of diabetic ocular pathology among those with SC and other studies suggesting a protective effect of SC against diabetic ocular pathology.10,13 Notably, Abadie et al8 reported on early retinopathy in 4 subjects with SCD and recently diagnosed DM, suggesting that SC may represent an additional risk factor for DR progression. Many prior reports were limited by sample size, and all were single-site investigations that examined either SCD or SCT rather than investigating both conditions.8, 9, 10,12,13 Recent work from a cohort of 176 876 individuals with DM in the United Kingdom, however, has demonstrated an increased risk of mortality for DM-SC individuals as compared with the DM-non-SC group.17 Our finding of increased DM-associated ocular pathology among DM-SC individuals provides additional context for the scope by which SC may mediate the DM disease process. Widespread microvascular and inflammatory effects of SC may serve as a mechanism for the observed increases in ocular pathology and procedural intervention among DM-SC patients. Specifically, the activation of proinflammatory or proangiogenic cytokines and signaling molecules, such as hypoxia-inducible factor 1 and hypoxia-inducible factor 2, with resultant neovascularization and endothelial disease, has been well-documented in individuals with SC.18, 19, 20 Further investigation is needed to understand the role of comorbid DM on these pathways and to better clarify the associations between DM and SCT as compared with SCD. With the current understanding of the interactions between SC and DM within the eye, ophthalmologists may choose to more closely monitor DM-SC patients for DR progression or a need for procedural intervention.
Strengths and Limitations
To our knowledge, the present study represents the largest retrospective cohort applied to examine the relationship between SC and DM and the prevalence of ophthalmic procedures. By utilizing the IRIS Registry, the sample examined in this study is broadly representative of the patients who attend ophthalmology practices within the United States. The sample is not, however, a complete representation of population-level data; it is not possible to calculate the prevalence of ocular complications or microvascular outcomes from this dataset. Additional limitations of this retrospective database are related to data completeness and data availability. Due to the constraints of the dataset available, we were unable to differentiate between SCT and SCD within the present analysis. Further studies will be needed to better understand the effects of DM within the subgroups of SCT and SCD. Investigation of SCD individuals with and without DM may be particularly useful to understand the relationship between procedural interventions across DM status, as SCD individuals with proliferative SC retinopathy may have received photocoagulation and other interventions due to SCD rather than diabetic disease. A further limitation of the IRIS Registry is the reliance of the data on specific coding and diagnosis by participating ophthalmologists. As PDR and proliferative SC retinopathy share many phenotypic features, it is possible treating ophthalmologists may not have been able to distinguish PDR from proliferative SC retinopathy. Diabetes-related parameters were unavailable within the IRIS database and were therefore not controlled for within the present study. Additionally, differentiation between type 1 and type 2 DM within the analysis was not possible due to dataset constraints. Further investigation of differences between the effect of type 1 and type 2 diabetes among patients with and without SCT or SCD would be useful to better understand potential cumulative processes, comorbid disease progression, and demographic factor influence given differences in pathogenesis, disease management, and ethnic prevalence between type 1 and type 2 DM. The number of patients coded with non-PDR was limited in this cohort, preventing meaningful analysis. In addition, the race-stratified random sample applied to this analysis was based on the distribution of race within the DM-non-SC group of the IRIS Registry. We did not racially match or age-match the groups compared within this analysis. The percentage of African American patients within the DM-SC group of the IRIS Registry was 72.2%, lower than the expected 90% based on the current Centers for Disease Control and Prevention data.3 This is a limitation of the IRIS Registry data analysis and likely due to racial self-report data, with an unusually high 6.8% of the DM-SC group being documented as White and 16.1% of the group being of unknown race.21 Further study with racial demographic distributions more reflective of the US SC population is warranted. While our analysis indicates worse VA trends among DM-SC patients compared with DM-non-SC patients, it is difficult to draw strong conclusions due to the limited sample size of study groups and VA subcategories, random sampling of the study design, and potential documentation of imperfections within the IRIS Registry database. We are unable to find clear associations with systemic comorbidities among the groups. Patients with incomplete data entries were excluded from the study sample before analysis. It is unclear whether the outcomes of the population that was excluded due to incomplete data may have varied meaningfully from the population that was included in the final sample.
Further, individuals carrying SCT are commonly unaware of their SCT status.22,23 Therefore, as there was no opportunity to uniformly conduct SCT testing in this retrospective study, an unknown number of individuals within the DM-non-SC group may have had undiagnosed SCT. This effect may have led to an underestimation of the study results. Finally, as the IRIS Registry may also be subject to errors in accurately documenting nonocular diseases such as diabetes, there may also have been an unknown number of individuals with incorrect DM status.
Individuals with DM and SC represent a vulnerable population with a high burden of potential for health care morbidity. The results of this analysis suggest that DM-SC individuals may have worse VA trends, an increased likelihood of both DM-related retinal disease and eye disease requiring procedural intervention. Identification of these trends has implications for the care teams of DM-SC individuals. Ophthalmologists may aim to collaborate with primary care teams to promote access to DM care, retinopathy surveillance screening, and ophthalmic services. These interdisciplinary teams may focus on obtaining close glycemic control, initiating diet and exercise interventions, addressing barriers to health care or medication access, and scheduling regular eye examinations to monitor for ocular complications. Given the signal that comorbid SCT may be associated with an increased need for procedural interventions for DR, ophthalmologists should be aware of SCT status in patients diagnosed with DM. For example, in the US population, Black patients are more likely to present with more severe diabetic eye complications such as DME.21 As this population is more likely to carry SCT than non-Black patients, when SCT status is unknown, ophthalmologists could consider testing patients with DM for the presence of comorbid SCT.
This study identifies multiple opportunities for future research. Future cross-sectional studies are needed to identify interactions between DM status within the subgroups of SCT and SCD. Prospective studies may utilize hemoglobin testing to precisely identify samples of DM-SCT, DM-SCD, and DM-non-SC groups and monitor both ocular outcomes and DM progression over time. Thus, further research efforts may build on the findings of this retrospective study to expand our understanding of the relationship between SC and DM in relation to ocular and systemic microvascular pathology.
Manuscript no. XOPS-D-24-00458.
Footnotes
Disclosure(s):
All authors have completed and submitted the ICMJE disclosures form.
The author(s) have made the following disclosure(s):
J.A.M.: Research support – Research to Prevent Blindness and American Academy of Ophthalmology Grant Award for IRIS Registry Research; X.L.: Research support – Grant no. NIH P30EY001765 (Grant that support Johns Hopkins Wilmer biostatistic center which provide statistical support to research in Wilmer Eye institute); A.W.S.: Research support – NIH P30 Core grant: P30EY001765, the Wilmer Eye Institute Biostatistics Center Core grant P30EY001765, by the 2022 Research to Prevent Blindness/American Academy of Ophthalmology Award for IRIS Registry Research, Genentech Roche (not relevant to this work); Consulting fees: Regeneron, Genentech Roche, EyePoint, Iveric Bio an Astellas Company, Apellis (not relevant to this work); Honoraria: Dutch Ophthalmic, Allergan Abbvie (not relevant to this work); Travel expenses: RPB AAO IRIS Registry Grant, Roche Genentech, Regeneron (not relevant to this work); Food and Beverage: Alcon, Carl Zeiss Meditech (not relevant to this work).
Supported by the NIH P30 Core Grant P30EY001765 and by the 2023 Research to Prevent Blindness/American Academy of Ophthalmology Award for IRIS® (Intelligent Research in Sight) Registry. Dr Scott receives research support from an endowment as the Frederick M. Leader Family Professor of Ophthalmology.
Support for Open Access publication was provided by the NIH and an AAO grant.
HUMAN SUBJECTS: Human subjects were included in this study. This study was reviewed by the Johns Hopkins School of Medicine institutional review board (IRB00367889) and was conducted in accordance with the Declaration of Helsinki. This study utilized patient information from the American Academy of Ophthalmology's IRIS (Intelligent Research in Sight) Registry. Given the deidentified nature of the patient data, we did not obtain written consent for enrollment in this study.
No animal subjects were used in this study.
Author Contributions:
Conception and design: Nampomba, Scott Obtained funding: Scott
Data collection: Ambrosino, McDonald, Li, Nampomba, Scott
Analysis and interpretation: Ambrosino, McDonald, Li, Scott
Obtained funding: Scott
Overall responsibility: Ambrosino, McDonald, Li, Nampomba, Scott
Presented in part at the American Academy of Ophthalmology IRIS Registry Symposium at the 2023 Academy Meeting, San Francisco, California.
References
- 1.National Diabetes Statistics Report | Diabetes | CDC. https://www.cdc.gov/diabetes/data/statistics-report/index.html
- 2.Klein R., Lee K.E., Knudtson M.D., et al. Changes in visual impairment prevalence by period of diagnosis of diabetes: the Wisconsin Epidemiologic Study of Diabetic Retinopathy. Ophthalmology. 2009;116:1937–1942. doi: 10.1016/j.ophtha.2009.03.012. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Data & statistics on sickle cell disease | CDC. https://www.cdc.gov/ncbddd/sicklecell/data.html
- 4.Nagpal K.C., Asdourian G.K., Patrianakos D., et al. Proliferative retinopathy in sickle cell trait: report of seven cases. Arch Intern Med. 1977;137:325–328. [PubMed] [Google Scholar]
- 5.Caranfa J.T., Witkin A.J. Severe proliferative retinopathy in a patient with sickle cell trait. Am J Ophthalmol Case Rep. 2023;30 doi: 10.1016/j.ajoc.2023.101824. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Broadhead G.K., Wiley H.E., Peprah D., et al. Proliferative retinopathy associated with repeated high-altitude exposure in a patient with sickle cell trait. High Alt Med Biol. 2022;23:369–371. doi: 10.1089/ham.2022.0070. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Page M.M., MacKay J.M., Paterson G. Sickle cell trait and diabetic retinopathy. Br J Ophthalmol. 1979;63:837. doi: 10.1136/bjo.63.12.837. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Abadie E., Lombrail P., Passa P. Is sickle cell trait an additional risk factor for diabetic angiopathy? Diabetes Care. 1981;4:659. doi: 10.2337/diacare.4.6.659a. [DOI] [PubMed] [Google Scholar]
- 9.Bleyer A.J., Reddy S.V., Sujata L., et al. Sickle cell trait and development of microvascular complications in diabetes mellitus. Clin J Am Soc Nephrol. 2010;5:1015. doi: 10.2215/CJN.08841209. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Koduri P.R., Patel A.R., Bernstein H.A. Concurrent sickle cell hemoglobin C disease and diabetes mellitus: no added risk of proliferative retinopathy? J Natl Med Assoc. 1994;86:682. [PMC free article] [PubMed] [Google Scholar]
- 11.Jackson H., Bentley C.R., Hingorani M., et al. Sickle retinopathy in patients with sickle trait. Eye (Lond) 1995;9:589–593. doi: 10.1038/eye.1995.145. [DOI] [PubMed] [Google Scholar]
- 12.Radparvar A., Muhtadi R., Jackson N., et al. The combined effects of diabetes mellitus and sickle cell disease on retinopathy and vascular comorbidities. Invest Ophthalmol Vis Sci. 2021;62:3195. [Google Scholar]
- 13.Al Harbi M., Khandekar R., Kozak I., Schatz P. Association between sickle cell trait and the prevalence and severity of diabetic retinopathy. PLoS One. 2016;11 doi: 10.1371/journal.pone.0159215. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Eye health among Black/African American people | National Eye Institute. https://www.nei.nih.gov/learn-about-eye-health/outreach-resources/eye-health-among-african-americans
- 15.Sommer A., Tielsch J.M., Katz J., et al. Racial differences in the cause-specific prevalence of blindness in east Baltimore. N Engl J Med. 1991;325:1412–1417. doi: 10.1056/NEJM199111143252004. [DOI] [PubMed] [Google Scholar]
- 16.R Core Team R: the R project for statistical computing. https://www.r-project.org/
- 17.Olvera-Barrios A., Warwick A.N., Wu Y., et al. Increased all-cause mortality in people with Sickle trait and diabetes after adjusting for diabetic retinopathy (DR) severity. Invest Ophthalmol Vis Sci. 2024;65:5125. [Google Scholar]
- 18.Pathare A., Al Kindi S., Daar S., Dennison D. Cytokines in sickle cell disease. Hematology. 2003;8:329–337. doi: 10.1080/10245330310001604719. [DOI] [PubMed] [Google Scholar]
- 19.Barbu E.A., Mendelsohn L., Samsel L., Thein S.L. Pro-inflammatory cytokines associate with NETosis during sickle cell vaso-occlusive crises. Cytokine. 2020;127 doi: 10.1016/j.cyto.2019.154933. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Zhang J., Qin Y., Martinez M., et al. HIF-1α and HIF-2α redundantly promote retinal neovascularization in patients with ischemic retinal disease. J Clin Invest. 2021;131 doi: 10.1172/JCI139202. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Varma R., Bressler N.M., Doan Q.V., et al. Prevalence of and risk factors for diabetic macular edema in the United States. JAMA Ophthalmol. 2014;132:1334–1340. doi: 10.1001/jamaophthalmol.2014.2854. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Treadwell M.J., McClough L., Vichinsky E. Using qualitative and quantitative strategies to evaluate knowledge and perceptions about sickle cell disease and sickle cell trait. J Natl Med Assoc. 2006;98:704. [PMC free article] [PubMed] [Google Scholar]
- 23.Oyeku S., Faro E., Homer C. Improving care of patients with sickle cell disease and sickle cell trait: the hemoglobinopathy learning collaborative series. J Clin Outcomes Manag. 2014;21:1. [Google Scholar]