Epidemiology and Associated Ocular Conditions of Visual Impairment in the United States: A Decade of Data from TriNetX US Network

Abstract

Purpose

To characterize the demographic profile, within-cohort prevalence of associated ocular conditions, and trends in the incidence and prevalence of visual impairment in the United States using a large, real-world clinical database.

Methods

We conducted a retrospective cohort study utilizing the TriNetX US Collaborative Network. Patients with visual impairment and associated ocular conditions potentially contributing to visual impairment were identified using ICD-10 codes between January 1, 2014 and January 1, 2026. Demographic characteristics, within-cohort prevalence of associated ocular conditions, and annual incidence and prevalence within the network were analyzed.

Results

A total of 805,428 patients with visual impairment were identified. The mean current age was 55.3 ± 26.0 years, with slightly more females (52.93%). Most patients were White (61.68%), followed by Black or African American (20.92%) and Hispanic or Latino (11.54%). Cataract (18.5%), glaucoma (13.9%), age-related macular degeneration (9.7%), and diabetic retinopathy (6.6%) were the most frequently observed associated ocular conditions in our cohort. Between 2014 and 2025, the annual incidence of visual impairment increased significantly from 124.56 to 212.69 per 100,000 persons (P < 0.001). Prevalence increased from 471.56 to 1433.70 per 100,000 persons (P < 0.001).

Conclusion

Visual impairment affected a broad demographic spectrum, with cataract, glaucoma, age-related macular degeneration, and diabetic retinopathy being the leading associated ocular conditions. Network-based incidence and prevalence increased over the decade, reflecting both true epidemiologic trends and network expansion. These findings highlight the importance of targeted screening, early detection, and timely treatment, particularly among high-risk and underserved populations.

Introduction

Visual impairment poses a growing public health challenge in the United States, with 1.02 million people experiencing blindness and 3.22 million people living with visual impairment in 2015; these numbers are expected to double by 2050, reaching approximately 2.01 million people with blindness and 6.95 million people with visual impairment.¹ Visual impairment has substantial impacts on individuals and society, including reduced quality of life, mobility limitations, and economic productivity losses. Visual impairment also has increased risks for social isolation, depressive symptoms, cognitive impairment and dementia, falls, and mortality.^2,3 In 2017, the total economic burden associated with vision loss in the United States was estimated at $134.2 billion, including both direct and indirect costs.⁴

Understanding the epidemiology of visual impairment allows policy makers and healthcare providers to develop strategies and allocate resources to prevent and treat visual impairment.^5,6 Characterizing the demographic profile and associated ocular conditions may help identify target populations and conditions for screening and prevention, while assessing temporal trends in incidence and prevalence can characterize changes in the burden of visual impairment over time. Although the burden of visual impairment is well recognized, there is a lack of epidemiological studies based on up-to-date, large-scale clinical data from the United States. Prior population-based studies, including the Beaver Dam Eye Study and the Baltimore Eye Survey, while landmark contributions, were conducted in specific geographic regions and included limited diversity among age or ethnicity, thereby limiting their generalizability to the entire US population.^7,8 In addition, these studies were conducted several decades ago, further reducing their applicability to the contemporary, increasingly diverse US population.

Large-scale electronic health record (EHR) networks provide a valuable platform to study visual impairment epidemiology using real-world clinical data from diverse patient populations nationwide. The TriNetX US Collaborative Network is a federated research platform comprising over 174 million de-identified EHRs from more than 80 healthcare organizations in the United States. This platform enables comprehensive analyses using detailed demographic, diagnostic, laboratory, procedural, medication, genomic, and longitudinal data, providing a powerful resource for epidemiological research.⁹

This study aimed to characterize the demographic profile, analyze within-cohort prevalence of associated ocular conditions, and describe trends in the incidence and prevalence of visual impairment in the United States between 2014 and 2025, using TriNetX US Collaborative Network.

Methods

Study Design and Data Source

This retrospective cohort study was conducted using the TriNetX US Collaborative Network. All data within TriNetX are de-identified in compliance with the Health Insurance Portability and Accountability Act (HIPAA) Privacy Rule, in accordance with the de-identification standard of protected health information defined in 45 CFR §164.514(a). Because this study used only de-identified patient data from the database, approval from the Institutional Ethics Committee of Willis Knighton Health was not required. This study was conducted in accordance with the tenets of the Declaration of Helsinki.

Cohort Identification

We extracted data on June 7 and June 9, 2026, using the International Classification of Diseases, 10^th Revision (ICD-10) codes (Supplemental Table 1). We identified patients with a diagnosis of visual impairment using the ICD-10 code H54, which encompasses blindness and low vision categories involving one or both eyes and generally corresponds to best-corrected visual acuity (BCVA) worse than 20/70, between January 1, 2014 and January 1, 2026. Associated ocular conditions occurring in patients diagnosed with visual impairment were further identified using ICD-10 codes corresponding to major clinically recognized causes of visual impairment, including glaucoma (H40.1-6,8.9, H42), cataract (H25, H26, H28), age-related macular degeneration (H35.31-H35.32), diabetic retinopathy (E08.31-35,37, E09.31-35,37, E10.31-35,37, E11.31-35,37, E13.31-35,37), retinal detachments and breaks (H33), hereditary retinal dystrophy (H35.5), retinopathy of prematurity (H35.1), other non-diabetic proliferative retinopathy (H35.2), cystoid macular degeneration (H35.35), separation of retinal layers (H35.7), angioid streaks of macula (H35.33), macular cyst, hole, or pseudohole (H35.34), puckering of macula (H35.37), toxic maculopathy (H35.38), degenerative myopia (H44.2), retinal vascular occlusions (H34), retinal hemorrhage (H35.6), optic papillitis (H46.0), retrobulbar neuritis (H46.1), other optic neuritis (H46.8), unspecified optic neuritis (H46.9), nutritional optic neuropathy (H46.2), toxic optic neuropathy (H46.3), ischemic optic neuropathy (H47.01), papilledema (H47.1), optic atrophy (H47.2), and other disorders of optic disc (H47.3), corneal diseases including keratitis and corneal opacities (H16, H17) and keratoconus (H18.6), H20 (iridocyclitis), H30 (chorioretinal inflammation), H44.11 (panuveitis), H44.13 (sympathetic uveitis), and ophthalmic injury (S05). Pediatric causes such as amblyopia and strabismus (H53.0, H50) were also identified when associated with visual impairment. No specific temporal sequence was imposed between the visual impairment diagnosis and the associated ocular diagnosis. Demographic variables, including age at index diagnosis, current age, sex, race, and ethnicity, were collected. The index date was defined as the first recorded diagnosis of visual impairment (H54) within the study period.

Study Outcomes and Statistical Analysis

The outcomes of this study included the annual incidence and prevalence of visual impairment within the network between 2014 and 2025, standardized per 100,000 persons. We derived these estimates using the TriNetX analytics environment. Incidence was defined as the number of newly recorded cases of visual impairment during each calendar year divided by the at-risk population within the TriNetX network for that year. Prevalence was defined as the total number of patients with a recorded diagnosis of visual impairment during each calendar year divided by the total active population within the network. Additional outcomes included demographic profile and the prevalence of associated ocular conditions of visual impairment within the cohort, based on the pre-specified etiologic categories.

We used descriptive statistics to summarize demographic characteristics and distribution of associated ocular conditions. Continuous variables were reported as means with standard deviations, and categorical variables were reported as counts and percentages. Annual incidence and prevalence per 100,000 persons were calculated using the TriNetX analytics platform. The 95% confidence intervals (CI) for annual incidence and prevalence were calculated using the Wilson score interval, and temporal trend analyses were performed using binomial logistic regression; both analyses were performed using R version 4.4.0 (R Foundation for Statistical Computing, Vienna, Austria).

Results

Cohort Demographics

A total of 805,428 patients were identified with a diagnosis of visual impairment (H54) between January 1, 2014 and January 1, 2026. The mean current age of patients in the cohort was 55.3 ± 26.0 years, and the mean age at index diagnosis was 49.6 ± 26.5 years. Of the total cohort, 426,354 individuals (52.93%) were female, and 573,542 (71.21%) were identified as non-Hispanic or Latino. In terms of race, the majority of patients were White 496,811 (61.68%), followed by Black or African American 168,517 (20.92%). Other racial groups included Asian (29,044; 3.61%), American Indian or Alaska Native (4,951; 0.61%), and Native Hawaiian or Other Pacific Islander (4,717; 0.59%). Race was reported as other in 38,241 (4.75%) patients and unknown in 63,147 (7.84%) patients (Table 1).

Table 1.

Demographic Characteristics of Patients with Visual Impairment in the United States

Category	Subcategory	n= 805,428	% of Cohort
Age	Current Age (Mean ± SD)	55.3 ± 26.0 years
	Current Age, Median (IQR)	62 (45) years
	Age at Index (Mean ± SD)	49.6 ± 26.5 years
	Age at Index, Median (IQR)	56 (45) years
Sex	Female	426,354	52.93%
	Male	378,706	47.02%
	Unknown	368	0.05%
Ethnicity	Not Hispanic or Latino	573,542	71.21%
	Unknown Ethnicity	138,909	17.25%
	Hispanic or Latino	92,977	11.54%
Race	White	496,811	61.68%
	Black or African American	168,517	20.92%
	Unknown Race	63,147	7.84%
	Other Race	38,241	4.75%
	Asian	29,044	3.61%
	American Indian or Alaska Native	4,951	0.61%
	Native Hawaiian or Other Pacific Islander	4,717	0.59%

Prevalence of Associated Ocular Conditions Within the Cohort

Among the associated ocular conditions, cataract was the most common, affecting 149,432 (18.5%) patients, followed by glaucoma in 112,048 (13.9%), age-related macular degeneration (AMD) in 78,073 (9.7%), and diabetic retinopathy (DR) in 53,653 (6.6%). Other associated ocular conditions included amblyopia or strabismus in 48,877 (6.1%) patients, corneal opacities or keratitis in 34,651 (4.3%), retinal vascular occlusions or hemorrhage in 32,253 (4.0%), ophthalmic injury in 25,393 (3.1%), uveitis in 18,584 (2.3%), keratoconus in 3,147 (0.39%), and trachoma in 16 (0.002%). Other non-AMD retinal disorders accounted for a substantial proportion of the cohort including retinal detachments or breaks (3.8%), macular cyst, hole, pseudohole, or puckering (3.4%), hereditary retinal dystrophies (1.6%), and other less common retinal disorders.

In the optic neuropathies group, optic atrophy was observed in 3.9% of the cohort, followed by optic neuritis in 2.9%, papilledema in 1.5%, ischemic optic neuropathy in 1.3%, other disorders of the optic disc in 1.3%, toxic optic neuropathy in 0.048%, and nutritional optic neuropathy in 0.037% (Table 2). Patients may have multiple associated ocular conditions; therefore, the percentages may overlap. Overall, cataract, glaucoma, AMD, and DR were the most common ocular conditions associated with visual impairment.

Table 2.

The Associated Ocular Conditions of Visual Impairment in the United States

Associated Conditions	n= 807,231	% of Cohort
Cataract	149,432	18.5%
Glaucoma	112,048	13.9%
Age-related macular degeneration	78,073	9.7%
Diabetic Retinopathy	53,653	6.6%
Amblyopia or strabismus	48,877	6.1%
Corneal opacities/keratitis	34,651	4.3%
Retinal vascular occlusions or hemorrhage	32,253	4.0%
Ophthalmic injury	25,393	3.1%
Uveitis	18,584	2.3%
Retinal detachment or breaks	30,876	3.8%
Macular cyst, hole, pseudohole, or macular pucker	27,708	3.4%
Hereditary retinal dystrophy	12,782	1.6%
Other non-diabetic proliferative retinopathy, cystoid macular degeneration, or separation of retinal layers	13,371	1.7%
Degenerative myopia	5,366	0.7%
Retinopathy of prematurity	2,805	0.3%
Toxic maculopathy	499	0.1%
Angioid streaks of macula	130	0.016%
Optic atrophy	31,571	3.9%
Optic neuritis	23,489	2.9%
Papilledema	11,752	1.5%
Ischemic optic neuropathy	10,340	1.3%
Other disorders of optic disc	10,606	1.3%
Toxic optic neuropathy	391	0.048%
Nutritional optic neuropathy	302	0.037%
Keratoconus	3,147	0.39%
Trachoma	16	0.002%

Note: The percentages may overlap because patients may have multiple associated ocular conditions.

Network-Based Incidence and Prevalence Trends

Between 2014 and 2025, the average annual incidence of visual impairment in the study population was 154.22 per 100,000 persons. The annual incidence ranged from a low of 124.56 per 100,000 persons in 2014 (95% CI, 123.35–125.78) to a peak of 212.69 per 100,000 persons in 2025 (95% CI, 211.17–214.23). Incidence declined in 2020 to 129.42 per 100,000 persons (95% CI, 128.41–130.45) compared to preceding years, followed by a steady increase from 2021 onward (Figure 1; Supplemental Table 2).

Figure 1.

Network-based annual incidence of visual impairment per 100,000 persons.

The prevalence of visual impairment increased consistently over the study period. Prevalence rose from 471.56 per 100,000 persons in 2014 (95% CI, 469.21–473.93) to 1433.70 per 100,000 persons in 2025 (95% CI, 1429.78–1437.64). The average prevalence during the study period was 859.99 per 100,000 persons. Similar to incidence trends, prevalence demonstrated a continuous upward trajectory, with the most pronounced increases observed after 2021 (Figure 2; Supplemental Table 2).

Figure 2.

Network-based annual prevalence of visual impairment per 100,000 persons.

Binomial logistic regression demonstrated significant overall increase in temporal trends in both annual incidence and prevalence over the study period (both P for trend <0.001).

Discussion

This study is one of the largest epidemiologic studies of visual impairment in the United States, including real-world clinical data from multiple healthcare organizations over a 12-year period. Our findings provide an updated overview of the demographic characteristics, associated ocular conditions, and network-based incidence and prevalence trends of visual impairment.

The demographic characteristics of our cohort may be related to both the underlying population structure and known disparities in visual impairment. The mean age at index diagnosis of 49.6 years suggests that visual impairment affects many individuals of working age, which may impose substantial psychological and financial burden on affected individuals and their families, as well as socioeconomic burden on society. The wide distribution of both current age and age at diagnosis indicates that individuals across a broad age range, not just older adults, are affected by visual impairment. The slight female predominance (52.93%) is consistent with prior epidemiologic studies and likely reflects longer female life expectancy, and potentially higher healthcare utilization.^10–12

The racial composition of the cohort should be interpreted with caution. While White patients constituted the majority of the cohort (61.68%), Black or African American patients represented 20.92% of cases, which was notably higher than their representation in the general United States population (approximately 12.6% per the 2022 US Census).¹³ This overrepresentation may reflect underlying disparities in both disease burden and healthcare utilization, including higher prevalence of glaucoma and diabetic retinopathy among Black Americans, as well as lower utilization of ophthalmic care that may result in more advanced disease at presentation.^14–16 Hispanic or Latino individuals accounted for 11.54% of cases, which was lower than the 19.1% proportion reported in the 2022 US Census.¹³ This discrepancy may partly reflect the 17.25% of individuals with unknown ethnicity in our dataset. These findings highlight the need for targeted screening and intervention strategies aimed at reducing disparities in preventable vision loss among racial and ethnic minority populations.

Cataract, affecting 18.5% of patients, was the most common associated ocular disease in our cohort. This finding aligns with global studies identifying cataract as the most common cause of reversible blindness among older adults.^17,18 Although cataract surgery is highly effective in restoring vision and widely performed in developed countries, the high proportion of cataract in our cohort underscores ongoing barriers to accessing ophthalmic care.¹⁹ In a recent US study, non-White patients, including Black, Hispanic, and Asian patients and those of lower socioeconomic status were reported to have poorer visual acuity secondary to cataracts, highlighting the need to remove barriers to access to cataract surgery, particularly among racial minorities and low-income population.²⁰

Glaucoma, a leading cause of irreversible blindness worldwide, was identified in 13.9% of cases in our study. Glaucoma can cause insidious vision loss, which typically progresses without symptoms until late stages.²¹ While there is no cure for glaucoma, lowering intraocular pressure can effectively slow disease progression.²¹ Given its insidious nature, screening programs are of critical importance to detect glaucoma, especially in individuals with risk factors, such as a family history of glaucoma, African ancestry, and elevated intraocular pressure.²² Long-term management of glaucoma requires sustained adherence to glaucoma medications and regular follow-up; however, suboptimal adherence remains a challenge in achieving effective disease control.²³

Age-related macular degeneration, another leading cause of irreversible vision loss, was present in 9.7% of cases in our study. This finding is consistent with previous studies reporting AMD as a major cause of irreversible vision loss in older adults in developed countries, and the prevalence of AMD is projected to increase with population aging.^24,25 Patients with early-stage or intermediate-stage AMD are typically asymptomatic or mildly symptomatic, whereas late-stage AMD, including wet AMD and geographic atrophy, can cause severe vision loss. Thus, regular ophthalmic screening is essential for early detection and timely intervention to prevent progression to late-stage disease.

Diabetic retinopathy was found in 6.6% of cases in our study. In 2021, an estimated 9.60 million people in the United States had DR, including 1.84 million with vision-threatening DR—representing 26.43% and 5.06% of people with diabetes, respectively—and the numbers of affected individuals are projected to increase with the growth of the diabetic population.¹⁵ Unlike AMD, DR commonly affects younger, working-age adults, imposing more significant socioeconomic burdens on the patients and society.²⁶ The prevalence of DR is higher in rural communities, partially due to lack of ophthalmic care providers and long travel distances.²⁷ The integration of mobile retinal imaging with artificial intelligence, together with ocular telehealth programs, offers a promising strategy to improve early detection and reduce vision loss from DR in underserved and rural populations.²⁸

Other non-AMD retinal disorders including retinal detachments or breaks, macular cyst, hole, pseudohole, or puckering, hereditary retinal dystrophies, degenerative myopia, and various macular pathologies also represented an important component of visual impairment in our cohort. Despite advances in diagnostic and therapeutic strategies for retinal disorders, including retinal imaging, intravitreal injection, vitreoretinal surgery, and emerging genetic therapies, many of these conditions still result in permanent vision loss,^29–32 underscoring the need for long-term ophthalmic care and visual rehabilitation. Optic neuropathies also accounted for a substantial proportion of the cohort. Optic atrophy, which represents end-stage axonal loss from various etiologies, was the most common diagnosis in this group. The diverse and complex etiologies of optic neuropathies often make diagnosis challenging and may require multidisciplinary evaluation and management.³³ The remaining associated ocular conditions, including amblyopia or strabismus, corneal disease, retinal vascular occlusions or hemorrhage, ophthalmic injury, uveitis, and keratoconus, were less common individually but collectively represented a meaningful component of the visual impairment burden. The near absence of trachoma in our cohort confirms the successful elimination of this infectious cause of blindness in the United States.³⁴

From 2014 to 2025, the annual incidence and prevalence of visual impairment within the network showed an overall upward trend, with the incidence increasing from 124.56 per 100,000 persons (95% CI, 123.35–125.78) to 212.69 per 100,000 persons (95% CI, 211.17–214.23), and prevalence rising from 471.56 per 100,000 persons (95% CI, 469.21–473.93) to 1433.70 per 100,000 persons (95% CI, 1429.78–1437.64). A transient decline in incidence was observed in 2020, likely attributable to reduced healthcare utilization during the COVID-19 pandemic. The rise in incidence and prevalence may be partially attributed to population aging and the increasing diabetic prevalence. Diabetic retinopathy, including vision-threatening diabetic retinopathy, has risen over time alongside the growing diabetic population.^35–37 Moreover, the growing racial and ethnic diversity of the United States may have contributed to the observed trends, as blindness prevalence is higher in certain minority populations.^13,16 Furthermore, expansion of the TriNetX network, with additional healthcare organizations joining over the study period, may have contributed to the rise in incidence and prevalence.

Our findings are generally consistent with prior population-based studies. Congdon et al reported AMD, DR, glaucoma, and cataract as the leading causes of visual impairment among adults in the United States.³⁸ Flaxman et al estimated that approximately 2.17% of the US population had visual acuity loss or blindness (BCVA ≤20/40 in the worse-seeing eye) in 2017.³⁹ The higher prevalence reported in their study may be attributed to differences in case definitions. While their study used a threshold of BCVA ≤20/40 in the worse-seeing eye, our study identified cases using ICD-10 code H54, which generally corresponds to a more severe level of visual impairment (BCVA <20/70).⁴⁰

This study has several limitations that should be acknowledged. First, this study relied on data from the TriNetX network and therefore the results may not fully represent the general population, since cases from healthcare organizations outside the network were not included. Second, the reliance on ICD-10 codes may have introduced classification bias because variability in coding practices across providers and institutions can result in coding inaccuracies. In addition, as with all database-based studies, our findings may be subject to reporting bias. Third, the TriNetX network is continuously expanding, with new healthcare organizations joining over time, which may have influenced temporal trends in incidence and prevalence. In addition, age-specific incidence and prevalence were not analyzed, and age-standardized rates could not be reliably calculated within the TriNetX platform. Therefore, variations in disease burden across age groups could not be assessed, and temporal trends may have been influenced by changes in the age distribution of the TriNetX population. In the absence of temporal sequencing of diagnoses, and with the possibility of concurrent ocular conditions (eg, cataract and glaucoma, or cataract and AMD), causal relationships between visual impairment and associated ocular conditions could not be established. Furthermore, the absence of socioeconomic variables such as income, educational level, and insurance status limited the ability to fully characterize the social determinants of visual impairment.^41,42 Finally, the data were obtained from two separate TriNetX queries. Because TriNetX is a dynamic database that is continuously updated, the total number of patients varies slightly between query runs. However, these minor differences are unlikely to affect the overall trends, proportions, or interpretation of the study findings.

Conclusion

In conclusion, in this large-scale study, visual impairment affected a broad demographic spectrum, with cataract, glaucoma, AMD, and DR being the most common associated ocular conditions. Over the 12-year study period, the incidence and prevalence of visual impairment within the network increased substantially, likely driven by both genuine epidemiologic trends and the expansion of the clinical network. These findings emphasize the importance of targeted screening, early detection, and timely treatment to address preventable and treatable causes of vision loss. They also underscore the disproportionate burden of visual impairment among racial and ethnic minority populations and the need to address disparities in access to ophthalmic care. Large-scale EHR networks offer a powerful and increasingly essential tool for monitoring the evolving burden of visual impairment in the United States.

Funding Statement

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Abbreviations

EHR, electronic health record; HIPAA, Health Insurance Portability and Accountability Act; ICD, International Classification of Diseases; BCVA, best-corrected visual acuity; CI, confidence interval; AMD, age-related macular degeneration; DR, diabetic retinopathy.

Disclosure

The authors report there are no competing interests to declare.