Incidence and Clinical Characteristics of Childhood Glaucoma: A Population-Based Study

Objective  To describe the incidence and clinical characteristics of childhood glaucoma in a defined population of the United States.

Methods  The medical records of all pediatric patients younger than 20 years living in Olmstead County, Minnesota, from January 1, 1965, through December 31, 2004, who met diagnostic criteria for glaucoma or glaucoma suspect were reviewed.

Results  Thirty children were diagnosed as having glaucoma during the 40-year study period. The incidence of childhood glaucoma was 2.29 (95% confidence interval, 1.47-3.12) per 100 000 residents younger than 20 years, with the following types and incidences: 19 acquired (1.46/100 000; 0.80-2.12), 6 secondary (0.45/100 000; 0.08-0.82), and 5 primary glaucoma (0.38/100 000; 0.05-0.72). The birth prevalence of primary congenital glaucoma during the 40-year period was 1 per 68 254 residents younger than 20 years or 1.46 per 100 000 (95% confidence interval, 0.03-8.16). Twenty-four individuals with glaucoma suspect were also identified, yielding an incidence of 1.9 per 100 000 residents younger than 20 years (95% confidence interval, 1.14-2.66).

Conclusion  The incidence of childhood glaucoma in this population was 2.29 per 100 000 residents younger than 20 years or 1 per 43 575 residents younger than 20 years. Acquired and secondary forms of glaucoma were the most common, whereas congenital and juvenile glaucoma were rare.

Childhood glaucoma is an uncommon pediatric condition often associated with significant visual loss.^1,2 It consists of a heterogeneous group of diseases leading to optic neuropathy and visual field changes that can be categorized into primary, secondary, and acquired subtypes.³ Primary glaucoma in children is generally divided into primary congenital glaucoma (from birth to early childhood) and juvenile primary open-angle glaucoma (4 years to early adulthood).^3,4 Primary congenital glaucoma (PCG) has previously been reported as the most common type of glaucoma seen in childhood.^4-6 Secondary glaucoma is usually defined as including syndromic disorders or other medical conditions present at birth, such as aniridia, Axenfeld-Rieger syndrome, retinopathy of prematurity, Rubinstein-Taybi syndrome, Sturge-Weber syndrome, and persistent hyperplastic primary vitreous and congenital rubella.^3,4 Acquired glaucoma, often classified as secondary glaucoma, is the result of other processes not present at birth, such as inflammation, drugs, trauma, and surgery.³

The incidence of childhood glaucoma within the United States, including PCG, to our knowledge, has not been reported. The purpose of this study is to describe the incidence and clinical characteristics of childhood glaucoma diagnosed during a 40-year period among patients younger than 20 years who were residents of Olmsted County, Minnesota.

The medical records of all patients younger than 20 years who were residents of Olmsted County when diagnosed as having glaucoma from January 1, 1965, through December 31, 2004, were retrospectively reviewed. Although the term childhood typically excludes patients older than 18 years, we extended the age limit to younger than 20 years because juvenile glaucoma is an entity that has been diagnosed in patients up to the fourth decade of life.^7,8 Institutional review board approval was obtained for this study. Individuals who met diagnostic criteria for glaucoma or glaucoma suspect were identified using the resources of the Rochester Epidemiology Project, a medical record linkage system designed to capture data on any patient-physician encounter in Olmsted County.⁹

The racial distribution of Olmsted County residents in 1990 was 95.7% white, 3.0% Asian American, 0.7% African American, and 0.3% Native American; the remaining 0.3% were of other race. The population is relatively isolated from other urban areas, and virtually all medical care is provided to its residents by Mayo Clinic, Olmsted Medical Group, and their affiliated hospitals. Patients not residing in Olmsted County at the time of their diagnosis were excluded from the study.

Glaucoma was defined generally in this study as an optic neuropathy and/or a visual field change, frequently associated with elevated intraocular pressure (IOP). Clinical signs, such as elevated IOP (>21 mm Hg), optic nerve cupping, asymmetric or progressive disc cupping, visual field defects, or anterior segment changes, such as enlarged corneas or Haab striae, alone or in combination, were used to make the diagnosis of glaucoma. The primary congenital glaucoma group included patients who displayed ocular signs of enlargement combined with 1 or more classic findings, such as photophobia, epiphora, Haab striae, corneal clouding, or optic nerve cupping, and ranging in age from birth to early childhood. The term juvenile glaucoma referred to primary open-angle glaucoma diagnosed when the patient was between 4 and 20 years of age. Secondary glaucoma referred to a systemic or ocular condition present at birth, whereas acquired glaucoma referred to secondary processes occurring after birth. No retrospective diagnosis of glaucoma was made, and all patients included had been diagnosed as having glaucoma by the ophthalmologists who participated in their care.

Glaucoma suspect was defined according to the American Academy of Ophthalmology guidelines¹⁰ as open anterior-chamber angles by gonioscopy and 1 or more of the following clinical findings: (1) appearance of the optic disc or retinal nerve fiber layer suggestive of glaucomatous damage, (2) a visual field suggestive of glaucomatous damage, and/or (3) consistently elevated IOP associated with normal appearance of the optic disc and retinal nerve fiber layer and with normal visual field test results. We included patients with an IOP greater than 21 mm Hg for 6 months or longer as meeting the third American Academy of Ophthalmology clinical criteria. Patients who did not meet the glaucoma suspect diagnosis guideline, as established by the American Academy of Ophthalmology, were excluded. In addition, patients diagnosed as merely having large physiologic cups or whose examination appeared to reveal isolated optic nerve cupping were also excluded. Other patients with normal examination results initially diagnosed with glaucoma suspect based on a positive family history of glaucoma were also excluded. Except for patients with a 6-month history of elevated IOP, those who were no longer considered to have glaucoma suspect at their last follow-up examination were also excluded from the study. Only those patients who had at least 2 ophthalmologic examinations with a diagnosis of glaucoma suspect were included in this study.

Continuous data are presented as a mean, and categorical data are presented as counts and percentages. To determine the incidence of childhood glaucoma in Olmsted County, annual age- and sex-specific incidence rates were constructed using the age- and sex-specific population figures for this county from the US Census. Because PCG is often diagnosed within the first year of life, a birth prevalence for PCG was also calculated from the number of births occurring from January 1, 1965, through December 31, 2004, using the annual birth incidence for this county. The 95% confidence intervals (CIs) were calculated using assumptions based on the Poisson distribution.

Thirty patients younger than 20 years were diagnosed as having glaucoma in Olmsted County during the 40-year study period. This number corresponds to an age- and sex-adjusted incidence of 2.29 per 100 000 residents younger than 20 years (95% CI, 1.47-3.12). Twenty-one (70%) of the children had their glaucoma diagnosed in the second half of the study (Table 1). Fifteen (50%) of the children had bilateral disease, affecting a total of 45 eyes. The incidence of PCG and other forms of childhood glaucoma diagnosed in this population are given in Table 2, whereas the clinical and demographic characteristics are given in Table 3. There were 16 boys and 14 girls, whose conditions were diagnosed at a mean age of 10.4 years (range, 15 days to 19.7 years), of the following races: 18 white (60%), 4 black (13%), 2 Asian (7%), 1 American Indian (3%), and 5 (17%) of unknown race. A family history of glaucoma was present in 6 (20%) study patients, whereas the mean presenting IOP and mean cup-disc ratio at the first visit were 30.8 mm Hg (range, 14-56 mm Hg) and 0.5 (range, 0.1-1) in the glaucomatous eye, respectively.

Nineteen (63%) of the 30 children with glaucoma were diagnosed as having acquired glaucoma at a mean age of 11.1 years (range, 3.0-18.8 years), and the incidence of glaucoma was 1.46 per 100 000 residents or 1 in 68 470 patients younger than 20 years (95% CI, 0.80-2.12). This group consisted of 12 boys and 7 girls, among whom 14 developed glaucoma secondary to trauma or surgery, 4 from uveitis, and 1 after topical corticosteroid use (Table 3). Seven (37%) of the 19 patients had bilateral glaucoma, and 3 (16%) had a family history of glaucoma. The mean IOP and cup-disc ratio at presentation were 34.8 mm Hg (range, 20-56 mm Hg) and 0.4 (range, 0.1-1) in the affected eye, respectively.

Six of the 30 children with glaucoma (20%) were diagnosed as having secondary glaucoma, including 2 with Sturge-Weber syndrome, 2 with neovascular glaucoma (1 patient with Coates disease and 1 with retinopathy of prematurity), and 1 each with Rubinstein-Taybi syndrome and cutis marmorata telangiectatica congenita, yielding an age- and sex-adjusted incidence of 0.45 per 100 000 residents or 1 per 221 877 patients younger than 20 years (95% CI, 0.08-0.82). The mean age at diagnosis for the 6 children was 6.4 years (range, 2 months to 18.4 years), with a mean initial IOP of 19.7 mm Hg (range, 14-27.5 mm Hg) in the glaucomatous eye. Three of 6 patients (50%) with secondary glaucoma had bilateral disease.

Primary childhood glaucoma was diagnosed in 4 children with juvenile glaucoma and 1 child with PCG. This finding corresponds to incidences of 0.38 (95% CI, 0.00-0.20), 0.32 (0.08-0.82), and 0.07 (0.00-0.20) per 100 000 patients younger than 20 years for primary childhood glaucoma, juvenile glaucoma, and PCG, respectively. In addition, because PCG frequently occurs during the first year of life, its birth prevalence was also calculated as 1 per 68 254 births or 1.46 per 100 000 births. Three of the 4 children with juvenile glaucoma were girls, 2 (50%) had a positive family history of glaucoma, and all had glaucoma diagnosed between 13 and 19 years of age with bilateral ocular involvement. The mean IOP for juvenile glaucoma at presentation was 27.6 mm Hg (range, 23.5-31.0 mm Hg), with a mean cup-disc ratio of 0.85 (range, 0.7-1.0) in the glaucomatous eye.

Twenty-four additional children were diagnosed as having glaucoma suspect, generating an incidence of 1.90 per 100 000 residents younger than 20 years (95% CI, 1.14-2.66) (Table 4). Twelve of the 24 children (50%) were diagnosed as having glaucoma suspect on the basis of elevated IOP greater than 21 mm Hg for at least 6 months. Five patients (21%) were considered to have glaucoma suspect based on cup-disc asymmetry and 1 based on varying disc cupping, whose ocular health was stable at final follow-up with a normal visual field and IOP. Six additional patients with glaucoma suspect were identified based on other factors, including 1 patient with pigmentary dispersion syndrome and others with different combinations of unusual disc appearance, suspicious visual fields, cupping, and elevated IOP. The mean age at diagnosis for the 24 patients was 13.7 years (range, 6.7-19.7 years), with a mean initial IOP of 20.1 mm Hg (range, 8.0-26.0 mm Hg) in the affected eye. All patients diagnosed as having glaucoma suspect included in the study were still considered to have this condition at the final follow-up by their ophthalmologist and had not developed glaucoma.

Childhood glaucoma was diagnosed in 30 children or 1 per 43 575 residents younger than 20 years during the 40-year study period. Among this heterogeneous group, 19 were diagnosed as having acquired glaucoma, 6 as having secondary glaucoma, 4 as having primary juvenile glaucoma, and 1 as having PCG. The birth prevalence of PCG in this population was 1 per 68 254 residents younger than 20 years or 1.46 per 100 000 (95% CI, 0.03-8.16). The clinical characteristics of these patients, by glaucoma type, were consistent with prior reports.

The incidence of PCG in this population is lower than that in all prior reports. The highest reported prevalence has been among individuals of Slovakian Roma (1 per 1250)¹¹ and Saudi Arabian (1 per 2500) racial descent.¹² Although the incidence of PCG in Western countries has been estimated at 1 per 10 000 population^3,13,14 to 1 per 12 500 population⁵ in previous reports, a recent population-based study conducted in the United Kingdom reported the incidence of diagnosis of PCG in Great Britain to be 5.41 per 100 000 population (1 per 18 500 population) and 3.31 per 100 000 population (1 per 30 200 population) in the Republic of Ireland.⁴ In addition, a birth prevalence of 2.85 per 100 000 population (1 per 38 000 population) was recently reported in Spain,¹⁵ which was similar to an incidence of 1 in 30 000 births found in an Australian study.¹⁶

This study, to our knowledge, is the first population-based report on the incidence of childhood glaucoma and PCG in the United States. The difference in incidence between this and other studies may be attributable to a number of reasons. First, some of the reported incidences are not population based and are likely to have been artificially elevated.¹⁴ Second, the diagnostic classification and inclusion criteria for childhood glaucomas have been inconsistent among studies.^5,16-19 Finally, race and consanguinity may affect the incidence of PCG, as shown in the British Infantile and Childhood Glaucoma Eye Study,⁴ which reported a 9-fold increased incidence of PCG in children of Pakistani descent compared with white children.⁴ Olmsted County, during the years of this study, was 96% white, a demographic with lower rates of PCG and other forms of glaucoma. Although only 4% of the county's population was of a race other than white, one-fourth of the patients diagnosed as having glaucoma were of other races. Although this disproportionate incidence of glaucoma in children of races other than white is similar to the findings of the British Infantile and Childhood Glaucoma Eye Study, none of the patients of other races in this study had PCG. Instead, the types of glaucoma seen in our racial populations were traumatic or surgical (n = 4), juvenile (n = 2), and uveitic (n = 1). The low incidence of PCG in Olmsted County may reflect a lower occurrence of consanguineous practices in the county.⁵

Acquired glaucoma was the most prevalent type of childhood glaucoma diagnosed in Olmsted County, comprising more than 63% of cases and occurring in 1 per 68 470 patients younger than 20 years.The relative predominance of boys in this glaucoma subtype likely reflects the higher rate of traumatic glaucoma among boys (11 boys and 3 girls). The incidence of traumatic hyphema from the same population was 12 per 100 000, with a male to female ratio of 5:1.²⁰ Fourteen of the 19 children (74%) with acquired glaucoma had either traumatic or iatrogenic causes for the development of their glaucoma. The larger proportion of patients whose glaucoma was attributable to traumatic and aphakic causes in this study highlights the importance of protective eyewear and limiting the glaucomatous effects of anterior segment surgery.

Secondary glaucoma was found in 6 children, for an incidence of 0.45 per 100 000 or 1 in 221 877 patients younger than 20 years (95% CI, 0.08-0.82). In the British Infantile and Childhood Glaucoma Eye Study, 52 of 99 children with glaucoma who were younger than 16 years were diagnosed as having secondary glaucoma, a classification that included lens-related conditions and other medical or syndromic diseases. In our population-based study, secondary and acquired glaucoma cases consisted of most glaucoma cases seen. Although the distinction between secondary and acquired glaucomas was not made in prior studies, our population-based report shows that secondary glaucoma manifested itself earlier than acquired glaucoma, possibly because of the association of other disorders present at birth.

Glaucoma suspect was diagnosed in 24 patients, or 1.90 per 100 000 residents younger than 20 years, in this study. We are unaware of any other population-based incidence rate of glaucoma suspect among children. However, the incidence rate reported in this study is likely an underestimation, given the asymptomatic nature of the disease. None of the children with glaucoma suspect in this study were known to develop visual field loss characteristic of glaucoma during their follow-up. Interestingly, patients with cup-disc asymmetry were identified earlier than other patients with glaucoma suspect and were more likely to have a positive family history.

The findings in this study have several limitations. The retrospective design is limited by nonstandardized and incomplete data collection. The pressure measurements and other clinical findings were not all performed by a single examiner or a glaucoma specialist, potentially introducing the uncertainty of examiner error. In addition, some forms of glaucoma are asymptomatic or otherwise unrecognized by the patient or outside observer, particularly juvenile glaucoma and glaucoma suspect. As a result, the incidence of these forms, unlike that of PCG, is likely to be underestimated. Similarly, the apparent increase in the incidence of glaucoma and glaucoma suspect in this cohort during the later years of the study is most likely owing to an improved understanding and recognition of the disease.²¹ Moreover, although a relatively isolated community, some residents of Olmsted County with glaucoma may have sought care outside the region, leading to a further underestimation of disease occurence. Finally, the low incidence of childhood glaucoma in this population makes it difficult to extrapolate the findings of the various subtypes of glaucoma. The generalization of data from this study is further limited by the demographics of Olmsted County, a relatively homogeneous, semiurban, white population.

This study provides population-based incidence rates for childhood glaucoma diagnosed during a 40-year period. Childhood glaucoma was found in 1 per 43 575 patients younger than 20 years. The most common type of glaucoma was acquired glaucoma (traumatic, surgical, uveitic, or drug induced), accounting for 63% of patients with glaucoma, whereas the secondary and primary forms were less common. Primary congenital glaucoma was diagnosed in 1 per 68 254 residents younger than 20 years of age (1.46 per 100 000 residents younger than 20 years of age), a rate that is lower than that reported in the Spanish, British, or Australian general population.

Correspondence: Brian G. Mohney, MD, Department of Ophthalmology, Mayo Clinic, 200 First St SW, Rochester, MN 55905 (mohney@mayo.edu).

Submitted for Publication: January 6, 2009; final revision received October 11, 2009; accepted November 10, 2009.

Author Contributions: Drs Aponte and Mohney had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Funding/Support: This study was supported in part by an unrestricted grant from Research to Prevent Blindness Inc, New York, New York.

Financial Disclosure: None reported.