Selected Sun-Sensitizing Medications and Incident Cataract

Abstract

Context

Sunlight exposure has been found to be associated with the development of age-related cataract. Use of some medications has been related to the development of sensitization to sun exposure.

Objective

To examine the relationship of use of sun-sensitizing medication on cumulative incidence of age-related cataract.

Design

Population-based longitudinal cohort study.

Participants

Adults in the Midwestern community of Beaver Dam, Wisconsin, who were seen at baseline and three follow-up visits each five years apart.

Methods

Sun exposure was estimated from residential history that permitted calculation of Wisconsin sun years (WISY) at the baseline examination. Medication history was reported at each examination. Cataract presence was determined by standardized lens photographs that were taken at each examination and graded according to standard protocols.

Main Outcome Measure

Cumulative incidence of age-related cataract.

Results

There were no significant effects of WISY exposure or use of sun-sensitizing medications on cumulative incidence of any type of age-related cataract while controlling for age and gender. However, there was a significant interaction term combining WISY and use of any sun-sensitizing medication on cortical cataract (P=.04) such that risk of cataract is significantly higher for persons in both risk groups: taking sun-sensitizing medication and higher sun exposure. Further controlling for diabetes status, history of heavy drinking and hat or sunglasses use did not alter the relationships.

Conclusions

These data suggest that use of sun-sensitizing medications interacts with sun exposure to influence the risk of cortical cataract, a common age-related cataract. If confirmed, this finding may have important implications for medication use.

Several risk factors, aside from age, have been identified for common types of cataracts and they vary by type of cataract. For example, cigarette smoking has commonly been found to be related to nuclear cataract,^1-5 diabetes mellitus to cortical and posterior subcapsular cataracts,^6,7 and hypertension (or use of thiazide diuretics) to posterior subcapsular cataract.⁸ Ritter and colleagues⁹ reported that heavy drinking was associated with all three types of cataract. Sunlight exposure and ultraviolet light B (UVB) exposure have been found to be associated with cortical cataract.^10,11 Several systemic medications are reported as being sun-sensitizing. The description of signs and symptoms attributable to sun sensitization by a medication include the development of itching or rash in areas of the skin exposed to the sun. Since cortical cataract has been previously associated with sun or UVB exposure and because of the common ectodermal derivation of the lens and skin, we sought to examine the possible modifying effect of sun-sensitizing medication on the association of incident cortical cataract to estimated UVB exposure.

Methods

Population

Methods used to identify and describe the population have appeared in previous reports.^12-16 In brief, a private census of the population of Beaver Dam, Wisconsin, was performed from September 15, 1987, to May 4, 1988, to identify all residents in the city or township of Beaver Dam who were 43 to 84 years of age. Of the 5,924 eligible individuals, 4,926 participated in the baseline examination between March 1, 1988, and September 14, 1990.^12,13 Ninety-nine percent of the population was white. Of these, 3,684 (81.1%) participated in the 5-year follow-up examination in 1993 to 1995.¹⁴ Of the 3,334 surviving participants in the baseline and second examination, 2,764 (82.9%) participated in the 10-year follow-up examination.¹⁵ Of the 2,480 surviving participants who were examined at the baseline, 5- and 10-year follow-up examinations, 2,119 (85.4%) participated in the 15-year follow-up examination.¹⁶ Comparisons between participants and non-participants at the time of the baseline, 5-, 10-, and 15-year follow-up examinations have appeared elsewhere.^13-16

Examination, Interview and Grading Protocols

The same protocols with few additions or deletions were used at each examination phase. A standardized interview was administered. These were approved by the institutional human subjects committee of the University of Wisconsin and conformed to the tenets of the Declaration of Helsinki. Participants were asked about residential history including time spent in the tropics and southern United States. The residential history was used to construct a measure of the average annual ambient UVB exposure for each individual using an adaptation of the technique employed by the Maryland Watermen Study.^11,17 Each year spent in another region was weighted by the ratio of the total ambient UVB light present in that area to the level for one year in Wisconsin (Wisconsin Sun Year (WISY)).¹⁷ One WISY is equivalent to the total ambient UVB irradiance of a horizontal surface in Wisconsin over one year. The cumulative ambient UVB exposure was computed and divided by age to compute the average annual ambient UVB exposure for an individual. Because most participants had spent most of their lives in Wisconsin, this variable was categorized into two groups (<1.01 WISY and ≥1.01 WISY).¹⁰ We assume that a person's exposure is relatively constant over time and so we use WISY computed from baseline data in the time varying models.

During the medical interview, the examiner reviewed the medications that the participants had been asked to bring to the examination. Every prescription and over the counter medication was recorded by name. When the participant neglected to bring them, the examiner called them at home after the interview to obtain the information. Medications were later linked to a local database containing lists of the active ingredients.

Participants were asked whether there had been a time in their lives when they had regularly drunk four or more alcoholic beverages daily on a regular basis.

Photographs of the lens were taken after pharmacologic dilation. Slit lamp photographs were taken to grade the degree of nuclear sclerosis. Retroillumination photographs were taken to grade presence and severity of cortical and posterior subcapsular cataracts. The protocols for photography and for the grading procedures were based on detailed codified rules.¹⁸ Graders were masked as to subject identity.

Definitions

Scores for cortical and posterior subcapsular cataracts were based on the estimated amount of ‘area’ involved after superimposition of the film on a grading grid. Cortical opacities involving more than a weighted average of 5% of the total lens were considered to be cortical cataract. Posterior subcapsular opacities involving a weighted average of more than 5% of any of the 9 individual grid segments were considered to be posterior subcapsular cataract.¹⁸

Scores for nuclear sclerosis were based on comparisons with standard photographs which resulted in a five-step scale of severity based on opacity of the nucleus; severities greater than standard three were considered nuclear cataract.

Diabetes was defined as a history of diabetes mellitus, treated with either insulin, oral hypoglycemic agents, and/or diet or, if there was no previous medical history of diabetes mellitus or use of hypoglycemic medications, then a casual blood sugar value greater than 11.1 mmol/L and a glycosylated hemoglobin value that was greater than two standard deviations above the mean for a given age-sex group (43-54 years of age, men >9.5% and women >9.6%; 55-64 years of age, men >9.4% and women >10.0%; 65-74 years of age, men >9.6% and women >9.6%; and 75 years of age or older, men >9.5% and women >9.6%) and a random blood sugar value of >200 mg/dL.¹⁹

Heavy drinking was defined as a history of ever having had a period in life when the individual drank 4 or more drinks per day on a regular basis.

Photo-sensitizing medications considered in this report are hydrochlorothiazide, furosemide, glyburide, amitriptyline, paroxetine, sertraline, tetracycline, sulfamethoxazole, trimethoprin, ciprofloxacin, amiodarone, and naproxen.²⁰ In time varying models drug use at the beginning of each time interval is used.

Statistical Analysis

Analyses were based on cataract incidence when it was the first eye to develop that lesion. Individuals contributed data for every interval that photographs for both eyes were gradable and when gradable photographs were also available for all previous examinations. Those who had cataract surgery in one eye were not considered for analysis because it was not possible to know whether the phakic eye was the first to develop a given lesion. Cumulative incidence was calculated accounting for the competing events of death and cataract surgery. Discrete linear logistic regression models were used to determine whether medication (and any interaction terms) added significantly to explaining incident cataract. Medication use was obtained at each examination and time varying models updating medication use at each examination were also performed. Since all the relevant information for updating WISY information was not available at later examinations, we assumed the same relative exposure for each examination in the time varying models. Statistical analyses were performed using SAS version 9.1 (Cary, NC).

Results

There were 4,926 persons at the baseline examination. The percentage of persons using sun-sensitizing medications as a group increased across the 15 years, although the percentages using individual medications varied (Table 1). Of the sun-sensitizing medications used at that examination, preparations containing hydrochlorothiazide were the most commonly used and ciprofloxacin the least commonly used.

Table 1.

Use of Sun-sensitizing Medications at Each Examination Prior to 15-Year Follow-Up.

	Baseline (N=4926)		5 Years (N=3722)		10 Years (N=2962)		15 Years (N=2395)
Active Ingredient	N	%	N	%	N	%	N	%
Amiodarone hydrochloride	4	0.1	6	0.2	4	0.1	19	0.8
Amitriptyline hydrochloride	76	1.5	78	2.1	80	2.7	58	2.4
Ciprofloxacin	4	0.1	4	0.1	14	0.5	17	0.7
Furosemide	198	4.0	185	5.0	218	7.4	227	9.6
Glyburide	80	1.6	112	3.0	92	3.1	71	3.0
Hydrochlorothiazide	809	16.4	479	12.9	467	15.8	545	22.9
Naproxen	67	1.4	70	1.9	203	6.9	231	9.7
Paroxetine and/or sertraline hydrochloride	0	0	21	0.6	90	3.0	94	4.0
Tetracycline hydrochloride	29	0.6	26	0.7	20	0.7	7	0.3
Sulfamethoxazole	25	0.5	22	0.6	15	0.5	21	0.9
Any sun-sensitizing medication total	1188	24.1	879	23.6	1019	34.4	1063	44.8

Of the 4,068 persons seen at baseline and at least one follow-up examination or who had died after the baseline examination (and thus could contribute some information for incidence), 156 had cataract surgery in at least one eye at baseline. An additional 214 did not have any gradable photographs at either baseline or the first follow-up visit and are excluded. Among the remaining 3,698 participants, the numbers of participants in each analysis varied by cataract. Individuals with a specific lens opacity at baseline or missing information for a specific opacity are excluded from analysis for that opacity, but were included for the other opacities. For cortical cataract, an additional 531 persons were excluded because of prevalent cataract at baseline and 169 were excluded due to missing information in at least one eye for cortical cataract at later visits. Parallel exclusions were made when analyzing the other cataract types.

There were 2,998 persons who contributed data for the 15-year cumulative incidence of cortical cataract. Those not included in any of the analyses were older, female, more likely to have had cardiovascular disease, and more likely to have diabetes at the baseline examination. Those not included were slightly more likely to have been taking sun-sensitizing medications.

The cumulative incidence of each cataract type by WISY and use of any sun-sensitizing medications is given in Table 2. Cumulative incidence of each cataract type is higher for those with high WISY and for those taking sun-sensitizing medications, but the differences are not significant. However, in multivariate models adjusting for age and gender, the interaction term for WISY with sun-sensitizing medication is significant for cortical cataract. There was no difference in the estimated odds whether or not diabetes (or history of heavy drinking) was included in the models. When stratified by WISY, the association of sun-sensitizing medication with cumulative incidence of cortical cataract was significant among those with high WISY (Figure 1). No other associations were significant. Use of hats with brims or use of sunglasses for at least half the time spent outside did not affect this relationship (data not shown). However, we had this information for only about half of our population and therefore could not assess the impact of this on the entire population.

Table 2.

Incidence of Cataract Type with Any Sun-sensitizing Medications

		Low WISY		High WISY		Cumulative WISY
Cataract Type	Sun Meds	N	Incidence %	N	Incidence %	N	Incidence %
Cortical	No	1784	21.7	587	23.1	2379	22.0
	Yes	469	24.6	149	31.8	619	26.3
		P for interaction = 0.042* or 0.043†
Nuclear
	No	1745	27.2	596	33.2	2348	28.8
	Yes	464	32.6	140	36.8	605	33.5
		P for interaction = 0.436* or 0.433†
Posterior subcapsular
	No	1972	7.7	652	9.5	2633	8.1
	Yes	562	9.1	174	10.2	736	9.4
		P for interaction = 0.783* or 0.903†

Abbreviations: WISY=Wisconsin sun years

^*P-value from age and gender adjusted model with sun years, sun-sensitizing medications, and interaction term

^†P-value from age, gender, diabetes, heavy drinking adjusted model with sun years, sun-sensitizing medications, and interaction term

Figure 1.

Hazards of Specific Cataract Types for Users of Sun-Sensitizing Medication by Wisconsin (WI) Sun Years. Hazard ratio (and 95% Confidence Interval) for incident cataract in those taking any sun-sensitizing medication. PSC = posterior subcapsular cataract.

Similar analyses were done for each individual medication. For those medications where we had adequate numbers, the Odds Ratios (OR) were in the same direction as for the entire category of sun-sensitizing medications (data not shown).

Time varying models incorporating medication use at each examination after baseline failed to find significance for sun exposure, medication use or for the interaction term (data not shown).

Discussion

We found evidence of an interaction between UVB exposure and use of sun-sensitizing medications on 15-year cumulative incidence of cortical cataract in the Beaver Dam Eye Study cohort. The medications (active ingredients) are a broad range of chemical compounds and the specific mechanism for the interaction is not clear. The effect seems to be specific to cortical cataract, the cataract type that has been found in some cross-sectional studies to be associated with UVB exposure. We previously reported such an effect in men in a cross-sectional analysis of the baseline data.¹⁰ The current analyses may indicate that the effect of exposure to sunlight may be modified by use of sun-sensitizing medications.

This study has several limitations. We measured UVB exposure accounting for average ambient UVB light in Wisconsin.¹⁰ Because the range of sunlight exposure was not great, we dichotomized this to <1.01 and ≥1.01 WISY. Thus, we cannot assess whether there might be an effect across the range of UVB exposure. Also, because our measure was based on residential history, it is a measure of potential exposure and not a direct estimate of ocular or lens exposure. Use of sunglasses and brimmed hats may alter direct UVB exposure on the lens, but adjustment for this did not alter the associations.

In addition, while we have data on medication use as reported at each examination, we do not know dose or duration of use. Thus, our measure of medication exposure is dichotomous and only accounts for current use or nonuse of any medication at that one time point.

We have included those sun-sensitizing medications that appeared on a current list compiled by the Department of Administration, Bureau of State Risk Management and the Wisconsin State Employees Union.²⁰ We used this source as there was no codified listing available from the Center for Disease Control, the Food and Drug Administration, the Slone drug data base (Slone Epidemiology Center, Boston University), or from other local or national sources. We have not included all medications that have been reported to be associated with sun sensitivity in a compendium such as the Physician's Desk Reference.²¹ We have no way to assess the possibility that entries in that publication were coincidental findings or that the sun sensitivity report was made in error or if it was appropriate. Also, ‘sun sensitivity’ is not always clearly defined.

The finding we report needs to be evaluated in other populations, especially in view of the increasing frequency of use of medications that we included as sun-sensitizing (Table 1). If our findings are confirmed, it would be important to examine whether the effect is greater in those with higher levels of ambient sunlight (UVB) exposure and if dose and/or duration of use of the medication is also important. Because cortical cataract is a common lens opacity in adults, present in about 16% of the Beaver Dam population at the baseline examination,²² this finding may have important public health impact.