Indoor Tanning and Risk of Melanoma: a Case-Control Study in a Highly Exposed Population – Response

Grant and colleagues state that our observed multivariate-adjusted odds ratio of 1.74 (95% confidence intervals, 1.42-2.14; ref. 1) for past versus never use of indoor tanning is “puzzling(ly)” high compared with some of the studies summarized in the IARC meta-analysis (2), primarily because we included individuals at high risk for melanoma. To back this claim, Grant and colleagues cite his re-analysis (3) of the IARC meta-analysis and use our published data to estimate how much of the melanoma risk associated with indoor tanning could be explained by these risk factors.

The weakly positive association for indoor tanning in relation to melanoma reported in the IARC meta-analysis, in contrast with our results, is not surprising when we consider major differences between past research on this topic and our study. To give just a few examples, we note the less detailed exposure assessment in the studies reviewed by IARC, the fact that prior studies did not restrict their analyses to cases with melanomas occurring by the age of 59 years, and they typically had smaller numbers of participants with substantial exposure. Although we did not publish the distribution of indoor tanning use across the various risk factors that would allow the reader to fully assess their potential for confounding, the crude odds ratio of 1.62 (see Table 5 of our report) was, in fact, stronger after adjustment for the very risk factors that Grant and colleagues suggest would “…explain nearly all the risk found” with exposure to indoor tanning.

To unequivocally address the concerns raised by Grant and colleagues, we present a stratified analysis according to the characteristics known to increase the risk of melanoma (see Table 1). Regardless of the characteristic used to stratify, the risk of melanoma in relation to indoor tanning use is elevated at about the same magnitude or greater as in our published report, the confidence intervals exclude the null value in almost all comparisons, and the P values for interaction are nonsignificant, except for family history. Although not reported here, we also consistently found a dose-response for melanoma risk in relation to the total number of indoor tanning sessions stratified by these same risk factors. Comparing persons in the highest category of use (e.g., 100+ sessions) to never users, the risk of melanoma was increased 2.2 to 4.9 times across all categories of risk defined in Table 1. Confidence intervals excluded the null value, and the P values for trend were highly significant in all but one instance (among red-haired persons).

In their letter, Grant and colleagues conclude that “those with preexisting high-risk factors… should be careful in using indoor tanning…”. Our data clearly indicate that both persons with and persons without these factors should expect that use of indoor tanning, especially frequent use, will increase their risk of developing malignant melanoma.

See the original Letter to the Editor, p. 2685.