Objective
We assessed the evidence supporting a reduction in risk for ovarian cancer occurrence or mortality with greater vitamin D exposures.
Study Design
This review followed standard guidelines for systematic literature reviews. The diverse study designs precluded a quantitative metaanalysis. Therefore studies are summarized via tables and abstracted information.
Results
Approximately half of the ecologic and case-control studies reported reductions in incidence or mortality with increasing geographic latitude, solar radiation levels, or dietary/supplement consumption of vitamin D, whereas the other half reported null associations. The cohort studies reported no overall risk reduction with increasing dietary/supplement consumption of vitamin D or with plasma levels of vitamin D prior to diagnosis, although vitamin D intakes were relatively low in all studies.
Conclusion
There is no consistent or strong evidence to support the claim made in numerous review articles that vitamin D exposures reduce the risk for ovarian cancer occurrence or mortality.
Vitamin D (representing both D 2 and D 3 ) is more properly viewed as a hormone rather than a vitamin. Human beings obtain vitamin D via 2 sources: ultraviolet (UV)-B radiation in sunlight that converts 7-dehydrocholesterol to vitamin D 3 in skin, and ingestion of vitamin D (D 2 or D 3 ) in food and supplements. Some fish and fish liver oils contain vitamin D 3 , although small amounts are also found in beef liver, cheese, and egg yolks. Fortified foods, such as milk, margarine, yogurt, cereal, and orange juice provide most of the vitamin D in the US diet. Once in the circulation, vitamin D (either D 2 or D 3 ) is converted to 25-hydroxyvitamin D (25-(OH)D) in the liver and then further metabolized to 1,25-dihydroxyvitamin D (1,25-(OH) 2 D) in the kidney and other tissues in the body. There are biological reasons to suspect that the active form, 1,25-(OH) 2 D, may be related to ovarian cancer incidence and mortality. For example, the vitamin D nuclear receptor, which mediates the effect of 1,25-(OH) 2 D 3 , is found in human ovarian tumor specimens and cell lines. Additionally, 1,25-(OH) 2 D 3 inhibits cell proliferation in ovarian cancer cell lines and induces apoptosis.
If indeed vitamin D can reduce ovarian cancer occurrence and mortality, then it represents a promising modifiable risk factor for ovarian cancer. However, a recent report from the International Agency for Research on Cancer on vitamin D and cancer did not evaluate ovarian cancer because of sparse information from primary research relative to that for other cancers such as colorectal, prostate, and breast cancers. Despite this, numerous review articles state that vitamin D reduces ovarian cancer occurrence or mortality. Therefore, we conducted a systematic review to assess the strength of the evidence with regard to the claim that higher (vs lower) vitamin D-related exposures are associated with a reduction in the risk for ovarian cancer occurrence or mortality.
Materials and Methods
Search strategy
This review followed the guidelines for systematic reviews of observational studies. Ten electronic databases including AltHealthWatch, AMED, Biological Abstracts, BIOSIS, CINAHL, Cochrane Library, EMBASE, MEDLINE, Proquest Dissertations, and the University of York’s Health Technology Assessment database, plus select international government World Wide Web sites, were searched up to July 2008 to identify relevant articles published in English or French. Databases and World Wide Web sites were searched using the following terms, both as text words and, as appropriate, Medical Subject Headings or equivalent subject heading/thesaurus terms: (ovary or ovaries or ovarian) and (cancer* or carcinoma* or carcinogen* or neoplasm* or tumor* or tumour*) and (vitamin D or 1,25-dihydroxyvitamin D or 25-hydroxyvitamin D or latitude or solar or sun or sunlight or sunshine or uv or uvb or uv-b or ultraviolet or calcium or dairy or margarine* or milk or cod liver oil* or fish oil* or halibut or herring or mackerel or oyster* or salmon or sardine* or swordfish or tuna).
The complete search strategy is available on request. Searches were supplemented by scanning bibliographies of all articles undergoing full-text review and review articles, and excluded articles flagged for a reference check. In total, the searches identified 2061 unique citations ( Figure ).
Screening of abstracts for eligibility
Abstracts/titles identified from the search were screened for eligibility by 2 reviewers (L.S.C. and H.K.N.). The 2 reviewers agreed on inclusion and exclusion for 96.8% of the abstracts/titles (kappa = 0.86). The remaining discrepancies were discussed and consensus reached in every instance. Abstracts were chosen for full-text review if they reported original data including: (1) vitamin D exposure or proxies (ie, vitamin D measured directly; dietary exposure [including supplements]; vitamin D in biological samples; solar irradiation measures; geographic location [eg, latitude]; and/or relevant occupational groups [eg, indoor vs outdoor workers]); and (2) ovarian cancer incidence, mortality, and/or survival. To ensure we captured all original data, we retained abstracts and titles that met these criteria even if they appeared to be review articles. We also retained articles whose relevance was uncertain due to lack of published abstracts.
A diagnosis of nonmelanoma skin cancer (NMSC) (as a proxy for sun exposure) was not 1 of the inclusion criteria. NMSCs are most likely underreported because they can be treated by physicians or outpatient clinics without histologic confirmation. Additionally, the majority of NMSCs are basal cell carcinomas for which the pattern of intermittent exposure and sunburn may be more important than chronic or cumulative exposure. Thus, a priori, we considered NMSC a poor proxy for cumulative sun exposure, which may in turn be a proxy for cumulative vitamin D exposure.
Full-text review of articles
All the authors developed the full-text screening tool and all participated in a test screening on a random sample of 10 articles to ensure the reproducibility of the full-text screening tool. Agreement was 100%. One reviewer (L.S.C.) then conducted a full-text article screen on 271 articles. A total of 251 articles were excluded based on a priori criteria ( Figure ). Articles that employed the results of previously conducted ecologic studies to estimate potential reductions in ovarian cancer incidence or mortality were excluded because the goal of this review was to assess the primary evidence for an association between vitamin D–related measures and ovarian cancer incidence and mortality. Relevant information was abstracted from the remaining 20 articles. These articles reflected ecologic, case-control, and cohort studies. No randomized controlled trials were identified. One ecologic study used the same mortality data and UV-B irradiance information as in a previous study, with a slightly different analysis; thus only the later analysis was included in the review.
Data abstraction
One reviewer (L.S.C.) completed all data abstractions. The following information was abstracted from all studies: study design; geographic location; time period; source and number of participants; race/ethnic composition; age composition; measure of ovarian cancer (incidence, mortality, survival); ovarian cancer behavior (invasive vs borderline); measure or proxy for vitamin D (food, supplements, food plus supplements, latitude, sunlight, UV-B); time frame of measurement; and outcomes (odds ratios [OR], relative risks [RR], correlations).
Analysis
The various study designs and analytical methods represented in the reviewed studies precluded a quantitative metaanalysis. Therefore, studies are summarized via tables and written descriptions.
Results
Ecologic studies
Ten ecologic studies reported original results on the relationship between latitude or similar measures of UV-B irradiance and ovarian cancer mortality or incidence ( Table 1 ). A common question addressed was whether ovarian cancer mortality increased with increasing latitude from the equator. Two studies noted that ovarian cancer mortality increased with latitude using 17 countries or within Italy alone, whereas 2 others within the United States and Spain did not. Another question was whether mortality decreased with increasing UV-B exposure or average sunlight/solar irradiance. Two studies reported that increasing UV-B exposure in the United States or increased sunlight in the United States was significantly associated with decreasing ovarian cancer mortality, whereas 2 other studies, 1 of solar irradiance in Japan and 1 of UV-B exposure in the United States, found no evidence of an inverse association. For incidence, 2 studies found no evidence that ovarian cancer incidence decreased with increasing UV-B exposure in the United States or across the world, whereas another worldwide analysis reported that incidence increased with increasing latitude and decreased with increasing UV-B exposure.
| Citation | Location | Ovarian cancer measure and calendar time | Vitamin D proxy measure | Increasing latitude or decreasing UV-B/sunlight associated with increasing mortality or incidence? |
|---|---|---|---|---|
| Pinto | United States | Mortality, 1950-1978 | Latitude and latitude/longitude index by county | No |
| Decarli and La Vecchia | Italy | Mortality, 1969-1978 | Latitude, set value for 20 Italian administrative regions | Yes, but level of significance not reported |
| Lefkowitz and Garland | United States | Mortality, 1979-1988 | Average sunlight (cal/cm 2 /d) for main city per county, average over 10 y (calendar years not specified) | Yes, P value < .001 |
| Mizoue | Japan | Mortality, 2000 | Average annual hours of solar radiation for central city in each prefecture (kW h/d), 1961-1990, Japan Meteorological Agency | No |
| Grant | Multiple countries |
| Latitude, set value per country, 17 countries | Yes, P value = .01 |
| Grant | Spain | Mortality, 1978-1992 | Latitude set value per province | No |
| Boscoe and Schymura | United States | Incidence, 1998-2002 Mortality, 1993-2002 | UV-B from NASA TOMS, 1996-2003, monthly average per county | Authors reported “no evidence” |
| Garland et al | Multiple countries | Incidence, 2002 |
| Yes, P value < .01 |
| Waltz and Chodick | Multiple countries | Incidence, 1993-1997 | UV-B data from Robertson-Berger meters | No |
| Grant and Garland | United States | Mortality, 1950-1969 and 1970-1994 |
| Yes, P value < .001 |
Thus, there have been inconsistent findings in ecologic studies of incidence or mortality. Half of the studies found that ovarian cancer mortality or incidence increases with increasing latitude or decreasing solar irradiance, while the other half did not. There was much variation in analytic methods and adjustment for other ecologic variables ranging from lifestyle variables such as population estimates of average number of alcoholic drinks in the last 30 days to atmospheric variables such as stratospheric ozone thickness, but the pattern of results did not align with the type of analysis or the extent of adjustment.
Case-control studies
A total of 6 case-control studies met the inclusion criteria; 1 evaluated ovarian cancer mortality and the other 5 evaluated incidence ( Table 2 ). The mortality study reported that being exposed to high vs low annual solar radiation (based on death certificates and state of residence at death) was associated with a lower risk of ovarian cancer death (odds ratio [OR], 0.84; 95% confidence interval [CI], 0.81–0.88), although the risk for women whose usual occupation listed on the death certificate involved outdoor work was similar to that of indoor workers (OR, 0.94; 95% CI, 0.75–1.17 and OR, 1.12; 95% CI, 0.88–1.41, respectively). The ecologic assignment of low, medium, and high annual mean daily solar radiation in the state of residence at death makes it difficult to directly compare this study to the other case-control studies with individual level information (summarized below).
