Beta carotene. Beta carotene was associated with aslight increase in cancer incidence (RR, 1.06; 95% CI, 1.00-1.12; P, 0%; P=.O3). This became more apparent when stratified by smoking status. Beta carotene was found to be associated with an increased incidence of cancer among smokers (RR, 1.10; 95% CI, 1.03-1.18; P, 0%; P=.O1) but not among nonsmokers (RR, 1.00; 95% CI, 0.92-1.10; 7^,0%; P=.94), as illustrated in Eigure 3. Results of the test of interaction assessing the effect of beta carotene among smokers and nonsmokers on cancer incidence were significant (P=.O1). Eor other antioxidants (vitamin E, selenium), no such interaction with cancer incidence was observed. We also conducted a formal test of interaction between beta carotene and vitamin E on cancer incidence, which was found to be nonsignificant.
Einally, beta carotene supplementation showed a trend toward increased cancer mortality (RR, 1.10; 95% CI, 0.94- 1.29; P, 13.1%; P=.24); however, 1 trial assessed a combintion of beta carotene, selenium, and vitamin E.^ Excluding this trial (as selenium could have a chemopreventive effect; see next paragraph), beta carotene showed a trend toward increased cancer mortality (RR, 1.16; 95% CI, 0.98-1.37; P, 71.6%), and this was marginally significant (P=.O8) (Figure 3). However, the latter result had substantial heterogeneity, possibly because of differences in the study populations.
particularly by smoking status. In fact, a higher RR was observed in the 2 trials that comprised smokers and former smokers*' (Eigure 3). This adds further weight to the observation that beta carotene might be more harmful among smokers, but unfortunately, we could not confirm this because mortality results were not stratified by smoking status (as the trials did not report the relevant data).