Cataract and Ovarian Carcinoma: Is the Vitamin D Hypothesis Alive?

This article is featured in Highlights of This Issue, p. 2469

Sunlight exposure is a recognized risk factor for developing cataract (1–3), a principal cause of blindness worldwide. According to the latest World Health Organization survey, age-related cataract is responsible for 48% of world blindness, representing about 18 million people (4). Solar radiation also contributes to nearly 1 million cases of basal and squamous cell skin cancers in the United States each year (5). Despite of these risks, public health recommendations about appropriate lower limits for sunlight exposure must consider the human requirement for UV radiation to catalyze vitamin D synthesis in the skin (6). Recent evidence suggests that colorectal cancer risk may be reduced through increased levels of circulating vitamin D (7, 8), but a role for vitamin D in the etiology of ovarian cancer remains controversial. Ecological studies have reported inverse associations between sunlight exposure and ovarian cancer incidence (9, 10) and mortality (11, 12). However, dietary studies of vitamin D exposure and ovarian carcinoma risk have been inconsistent (13). A recent meta-analysis of results from 10 longitudinal studies suggests a weak, nonsignificant inverse association between circulating 25(OH)D and ovarian cancer risk that was strongest among overweight women (14). Although evidence for a protective association of vitamin D with ovarian cancer is modest, a single blood measurement, while integrating dietary and nondietary sources of vitamin D, may not represent long-term vitamin D status.

We conducted a population-based study of ovarian carcinoma with a primary focus on dietary mechanisms in the etiology of ovarian cancer (15). Because of the complexity of ascertaining lifetime sunlight exposure, we asked study participants about their history of physician-diagnosed skin cancer and cataract as indicators of high long-term exposure to UV radiation.

Cases were women with histologically confirmed primary incident invasive ovarian carcinoma, 18 years or older, who were identified through the rapid-reporting systems of the Hawaii Tumor Registry and the Los Angeles County Cancer Surveillance Program between 1993 and 2008 (15). In total, 709 women meeting these criteria were recruited, with a participation rate of 78%. Controls were women without a history of ovarian cancer or bilateral oophorectomy, randomly selected from participants in an annual survey of representative households in Hawaii and by random digit dialing in Los Angeles. Interviewed controls (N = 1,001) were frequency matched to cases on ethnicity, 5-year age group, and place of residence (participation rate was 80%). The study was approved by the Institutional Review Boards of the University of Hawaii and the University of Southern California, Los Angeles, CA, and written informed consent was obtained from all participants.

Statistical analysis was conducted using SAS (release 9.2, SAS Institute Inc.). Multivariate unconditional logistic regression models were used to estimate ORs and 95% confidence intervals (CI) for the association of age, education, and other exposures with the risk of cataract among controls (Table 1) and to assess the association of cataract with the risk of ovarian carcinoma among all subjects (see Table 2 for adjustment variables). Heterogeneity of effects by study site (Hawaii, Southern California), body mass index (treated as both a continuous and a categorical variable: underweight or normal vs. overweight; obese vs. nonobese), and other covariates were examined using Wald test of the cataract–covariate interaction term. Heterogeneity of ORs by histology, stage, and grade was examined using polytomous logistic regression; the estimated ORs were compared using the Wald test. In addition, a sensitivity analysis was conducted excluding women of a younger age. All P values were 2-tailed.

Controls with a history of cataract were older (mean age = 70.0 ± 9.2 years vs. 53.3 ± 13.0 years; P < 0.0001), more likely to be diagnosed with type 2 diabetes mellitus (OR = 1.9; 95% CI, 1.3–3.4; P = 0.04), and at greater risk of skin cancer (OR = 2.1; 95% CI, 1.1–4.0; P = 0.03) than controls without a history of cataract (Table 1).

The prevalence of cataract among cases (14%) was significantly lower than among controls (17%), with an OR of 0.6 for ovarian carcinoma (95% CI, 0.4–0.8; P = 0.002) that was independent of known ovarian carcinoma risk factors (Table 2). Adding available potential risk factors for developing a cataract to the model, such as smoking, alcohol consumption, thyroid disease, and type 2 diabetes, did not substantially alter this association. No heterogeneity of the cataract–ovarian cancer association was observed by tumor histologic subtype (P = 0.92), stage (P = 0.16), or grade (P = 0.59; Table 2), and a history of cataract was associated with 50% lower risk of advanced high-grade serous carcinoma (95% CI, 0.3–0.9; P = 0.02). No heterogeneity was observed by study site (P = 0.87), body mass index (P = 0.80), or other known ovarian cancer risk factors. The association of cataract with ovarian carcinoma risk was consistent among women ≥50 years old (OR = 0.6; 95% CI, 0.4–0.8; P = 0.001); ≥60 years old (OR = 0.6; 95% CI, 0.4–0.9; P = 0.003); and ≥70 years old (OR = 0.5; 95% CI, 0.3–0.8; P = 0.002). A nonsignificant inverse association of skin carcinomas with ovarian cancer (OR = 0.7; 95% CI, 0.4–1.1; P = 0.09) was also observed.

Although cataract etiology is not fully understood, oxidative stress associated with UV light is considered central to the pathogenesis of this disease (1–3). Ecological studies have shown strong correlations of cataract development with lifetime sun exposure (3), a factor hypothesized to affect several human cancers, including ovarian malignancy (9–12). In support of an association of vitamin D with ovarian cancer, we have compelling results from our genetic studies providing evidence that several vitamin D receptor genetic variants may influence ovarian cancer risk (16, 17), perhaps through reduced receptor functionality. Furthermore, laboratory investigations showing that vitamin D and its synthetic analogues inhibit growth and induce apoptosis in ovarian cells in culture and in animal models of ovarian cancer (18–20) strengthen the hypothesis that vitamin D may reduce ovarian cancer risk. Although no sun exposure information was collected in our study, we found a direct association of cataract with skin cancer for which sun exposure is an established risk factor. Although an inverse association between carcinomas of the skin and ovary was observed, it was not significant due to the small number of women with skin cancer (N = 114). Moreover, skin cancer and ovarian cancer might have other common mechanisms (e.g., genetic and immunologic) that would obscure this potential association.

This investigation has several strengths including its relatively large size, population-based design, standard histopathologic review of all cases, and full information on a variety of health-related characteristics that allowed us to account for confounding effects. Our study had several limitations. Information on cataract history was self-reported and might lead to misclassification bias. Compared with cataract prevalence data in Hawaii (21) for all females (19%), history of cataract may have been modestly underreported (17%) by our control study participants. However, the misclassification would likely have affected both cases and controls equally and thus would tend to attenuate the association. No data on cataract type and grade were available in our study.

These findings add indirect evidence to the hypothesis that lifetime vitamin D exposure is inversely associated with risk of ovarian carcinoma and suggest some potential new avenues for research. Aside from sunlight exposure, a variety of drugs, such as corticosteroids (22) and phenothiazines (23), have been found to be associated with an increased risk of cataract; and statin drugs may reduce the risk for cataract (24). Additional studies are needed to further investigate the potential behavioral (e.g., UV radiation exposure) and biologic (e.g., medication history) factors that might influence associations of cataract with ovarian cancer.

No potential conflicts of interest were disclosed.

This work was supported by the NIH (grants R01 CA58598, N01-CN-55424, and N01-PC-67001 to M.T. Goodman).