supplementation with 1,000 mg of elemental calcium combined with 400 IU of vitamin D3 had no effect on breast cancer incidence. However, through a stratified analysis, the data demonstrated an increased risk of breast cancer for women who were already consuming 600 IU of vitamin D per day at baseline, to which a supplement of 400 IU/day was added (Pinteraction = 0.003). Serum 25OHD measures were analyzed by quintile and the highest quintile was 67.6 nmol/L and above.
Pancreatic cancer Some, but not all, observational studies suggest that higher serum 25OHD levels are associated with an increased risk of pancreatic cancer. Beginning with negative studies, Skinner et al. (2006) examined two large cohort populations—the HPFS and the NHS—for associations between pancreatic cancer incidence and vitamin D intake from diet and supplements. Another study of the HPFS cohort examined associations between serum 25OHD level and total cancer mortality or digestive (including pancreatic) cancer incidence (Giovannucci et al., 2006b). Both studies found reduced risk for pancreatic cancer incidence: in one instance associated with higher vitamin D intake (≥ 600 IU/day) (Skinner et al., 2006) and in the other based on a predicted serum 25OHD level (as described by the authors) for which RR was calculated for incremental increases in serum 25OHD level of 25 nmol/L (Giovannucci et al., 2006b).
In contrast, an initial study from Stolzenberg-Solomon et al. (2006), using a nested case–control protocol to evaluate associations between vitamin D nutriture and incidence of pancreatic cancer in subjects from the ATBC Study, found a positive association between higher serum 25OHD levels (highest quintile at 83.2 nmol/L) and risk for pancreatic cancer. In a subsequent nested case–control analysis of a cohort from the Prostate, Lung, Colorectal, and Ovarian (PLCO) Cancer Screening Trial, the same investigators found no association between higher serum 25OHD level and increased pancreatic cancer risk as an outcome (Stolzenberg-Solomon, 2009). The difference between the study populations included living at a northern latitude, positive smoking history, and gender (male) in the ATBC Study compared with a mixed gender U.S. population that was controlled for smoking history in the PLCO Cancer Screening Trial.
To address the dissimilarity in results from individual large cohort studies, Stolzenberg-Solomon et al. (2010) conducted a pooled nested case–control analysis of participants from several cohorts (the ATBC Study, CLUE, the Cancer Prevention Study II Nutrition Cohort, the New York University Women’s Health Study, the PLCO Cancer Screening Trial, and the Shanghai Women’s and Men’s Health Studies) to determine associations between serum 25OHD levels pre-diagnosis and risk for incident pancreatic cancer. This large-scale pooled analysis (n = 2,285) found a statistically significant two-fold increased risk for pancreatic cancer in participants with