Reference
Eliassen AH, Liao X, Rosner B, Tamimi RM, Tworoger SS, Hankinson SE. Plasma carotenoids and risk of breast cancer over 20 y of follow-up. Am J Clin Nutr. 2015;101(6):1197-1205. Epub 2015 Apr 15.
Design
A nested case-control study of plasma carotenoids and breast cancer risk
Participants
Between 1989 and 1990, 32,826 of the 121,701 women enrolled in the Nurses’ Health Study donated blood samples. Between 2000 and 2002, 18,743 of these women contributed a second blood sample. Between the first blood collection and June 2010, 2,188 breast cancer cases were diagnosed in these participants, and 579 cases were diagnosed after the second collection. These cases were matched with control participants.
Outcome Measures
Relative risks (RRs) were calculated for breast cancer occurrence, breast cancer recurrence, and death.
Key Findings
Higher concentrations of α-carotene, β-carotene, lycopene, and total carotenoids were associated with 18% to 28% statistically significant lower risks of breast cancer occurrence. Upper vs lower quintiles of β-carotene had a 28% lower relative risk (RR:0.72; 95% confidence interval [CI]:0.59-0.88; P-trend<0.001). These associations were seen for total carotenoids measured 10 or more years prior to diagnosis (upper vs lower quintile, RR:0.69; 95% CI:0.50-0.95; P-trend=0.01). A significant association was also seen between total carotenoids measured less than 10 years before diagnosis (RR:0.79; 95% CI:0.64-0.98; P-trend=0.04, P-interaction=0.11). Carotenoid concentrations were also strongly inversely associated with breast cancer recurrence and death (upper and lower quintiles RR:0.32; 95% CI:0.21-0.51; P-trend<0.001) compared with not recurrent and not lethal disease (P-heterogeneity<0.001).
These inverse associations of β-carotene and total carotenoids with breast cancer risk were seen only in lean women. Comparing upper and lower quintiles of β-carotene in women with a body mass index less than 25, higher levels were associated with a 38% lower risk for breast cancer (RR:62; 95% CI:0.47-0.83; P-trend=0.001) and for total carotenoids, a 36% lower risk (RR:0.64; 95% CI:0.48-0.84; P-trend=0.001). No significant association was seen in overweight or obese women.
Smoking also may change the action of carotenoids. While there was a strong inverse association between α-carotene and breast cancer in nonsmokers (RR:0.74; 95% CI:0.60-0.92; P-trend=0.01), no significant associations were present in current smokers (RR:1.23; 95% CI:0.54-2.80). Risk for breast cancer actually trended higher in smokers.
In all, α-carotene, β-carotene, β-cryptoxanthin, and total carotenoids were strongly inversely associated with risk of breast tumors that recurred or were ultimately lethal. Risks comparing the top quintile with bottom quintile were 46% to 68% lower for α-carotene (RR:0.54; 95% CI:0.35-0.83; P-trend=0.01); β-carotene (RR:0.32; 95% CI:0.21-0.51; P-trend=0.001); and total carotenoids (RR:0.48; 95% CI:0.31-0.73; P-trend=0.001). These inverse associations did not vary by estrogen receptor status of the participants.
Practice Implications
This study adds weight to our already strong understanding that most people should eat more vegetables—or at least lean, nonsmoking women should. Doing so increases total serum carotenoid levels, and this is associated with lower risk of getting breast cancer, having breast cancer recur, and dying from recurrent breast cancer. Why this association is not present in overweight smokers is unclear. Whether or not taking supplements that supply these nutrients is helpful or harmful also remains unclear.
The 6 major carotenoids—α-carotene, β-carotene, β-cryptoxanthin, lutein, zeaxanthin, and lycopene—together account for about 90% of the total carotenoids found in the blood serum, and it is these carotenoids that are thought to have an anticancer effect. As carotenoids inhibit growth of breast cancer cells in vitro,1 the authors of this current study questioned whether carotenoid serum levels might be predictive of risk for breast cancer. These results certainly suggest they are. Hypothesized mechanisms of action include conversion of the carotenoids to retinol, which regulates cell growth, differentiation and apoptosis, and the antioxidant capacity of these substances, which allows them to scavenge up reactive oxygen species.
High carotenoids in this study did not appear to offer protection to overweight women. Thus for these women, their efforts might be better rewarded if directed toward weight loss. Obviously, increasing consumption of fruits and vegetables may lead to weight loss.
There are 2 considerations that we should remember when considering the results of the present study. First, there is research linking supplemental β-carotene to increased incidence of lung cancer from the Alpha Tocopherol and Beta Carotene (ATBC) cancer prevention trial published in 1996.2 Second, the more recent Life After Cancer Epidemiological (LACE) cohort data suggested women reporting higher carotenoid supplementation use had greater risk of breast cancer recurrence. Again, the implication is that supplementation may be associated with increased risk.
The 6 major carotenoids—α-carotene, β-carotene, β-cryptoxanthin, lutein, zeaxanthin, and lycopene—together account for about 90% of the total carotenoids found in the blood serum, and it is these carotenoids that are thought to have an anticancer effect.
Past research suggests that the carotenoids have a different action in smokers than in nonsmokers. While this study does not confirm a positive association between smoking and carotenoid levels, even the nonsignificant trend should give us pause in light of earlier research on smoking and β-carotene supplementation.
In years past, epidemiological studies reported inverse associations between β-carotene plasma levels and lung cancer risk similar to what this study reports with breast cancer.3 Those early reports were followed by intervention trials that, contrary to expectations, reported increased tumor risk in smokers taking β-carotene supplements. Three of these studies deserve mention.
In 1994 the ATBC Cancer Prevention Study Group reported the results of their randomized, double-blind, placebo-controlled primary-prevention trial with a total of 29,133 male smokers who were given 50 mg α-tocopherol, 20 mg β-carotene, α-tocopherol plus β-carotene, or a placebo daily for 5 to 8 years. They found that β-carotene supplementation was associated with an 18% increase in risk of lung cancer.4
The Beta-Carotene and Retinol Efficacy Trial (CARET) was also a surprising failure. In this study, 18,314 study participants received 30 mg β-carotene plus 25,000 IU of retinol or placebo daily for an average of 4 years. All participants were at high risk for lung cancer, either because of heavy smoking or asbestos exposure. The trial was stopped earlier than planned in 1996 because participants taking the supplements had a 28% increase in lung cancer incidence and a 17% increase in death compared to the placebo group. The study participants remained at higher risk than the placebo group even 5 years after stopping supplementation.5
The Antioxidant Polyp Prevention trial also reported less than desirable effects for β-carotene supplementation in smokers. This was a multicenter, double-blind, placebo-controlled clinical trial of antioxidants for the prevention of colorectal adenomas. In all, 864 participants with history of adenomatous polyp removal took β-carotene (25 mg or placebo) and/or vitamins C and E. Among participants who neither smoked nor drank alcohol, β-carotene was associated with a 44% decrease in polyp recurrence (RR:0.56; 95% CI:0.35-0.89). In smokers though, taking β-carotene was associated with a nonsignificant 36% increase in risk (RR:1.36; 95% CI:0.70-2.62). For study participants who both smoked and drank alcohol, taking β-carotene was associated with a significant increase in risk; the risk of adenoma recurrence more than doubled (RR:2.07; 95% CI:1.39-3.08; P for difference from nonsmoker/nondrinker RR<.001).6
Animal studies, particularly those with ferrets, have shown clear negative effects of smoking combined with β-carotene supplementation.7 One proposed mechanism for this effect is through an increase in conversion of procarcinogens (such as those in cigarettes) to carcinogens via upregulation of phase I enzymes in the liver by β-carotene.8 In a much less cited paper published in 1996, the ATBC and CARET data were revisited, and both studies showed that participants with the highest baseline levels of circulating β-carotene had the lowest risk of developing lung cancer.9 The implication is that supplementing with high dose β-carotene may not equate to carotenoids in the diet.
Nevertheless, these earlier studies raise the question whether dietary carotenoids or β-carotene has similar unwanted action for smokers when it comes to breast cancer risk. A 2004 article by Nkondjock and Ghadirian tracked dietary intake of carotenoids and essential fatty acids, seeking associations with risk of breast cancer. This population-based, case-control study examined 414 cases of breast cancer. In premenopausal women who had ever smoked, those in the upper quartile of β-carotene intake had nearly 2.5 times higher risk of breast cancer than those in the lower quartiles of intake.10
In this current paper by Eliassen and colleagues, the association between smoking and breast cancer risk did not reach statistical significance. This lack of significance may have been because relatively few women in this cohort, made up entirely of nurses, were smokers.
The second reason to remain cautious regarding the interpretation of Eliassen’s results is the 2011 study by Greenlee et al that examined breast cancer recurrence rates in the 2,264 women in the LACE cohort. Greenlee’s study sought to answer the long repeated concern by oncologists that any antioxidant supplement use during chemotherapy and radiation therapy will decrease the benefit of treatment increasing recurrence rates and decreasing long-term survival.11
Women in the LACE cohort had been diagnosed with early stage primary breast cancer between 1997 and 2000. They enrolled in the study on average 2 years after their initial diagnosis. Data on antioxidant supplementation use since diagnosis and other factors were collected and compared with recurrence and mortality. About 81% of the women in the cohort took antioxidant supplements. In the LACE cohort, frequent use of vitamin C and vitamin E was associated with a decreased risk of breast cancer recurrence. Those taking vitamin C had a 27% lower risk of recurrence, and those taking vitamin E had a 29% lower risk (vitamin C, hazard ratio [HR]:0.73; 95% CI:0.55-0.97; vitamin E, HR:0.71; 95% CI:0.54-0.94). Vitamin E use was associated with a 24% decrease in mortality from any cause (HR:0.76; 95% CI:0.58-1.00). On the other hand, frequent use of carotenoid supplements appeared to double the risk of death from breast cancer (HR:2.07; 95% CI:1.21-3.56) and increase risk of mortality from all causes by 75% (HR:1.75; 95% CI:1.13-2.71).11
Can we equate these contradictory results? The difference may be in dietary carotenoids vs carotenoid-containing nutritional supplements. According to Eliassen (personal communication, May 4, 2015), their “study focused on blood levels of carotenoids measured prior to breast cancer diagnosis and did not focus on supplement use.” She notes that some women in her study used supplements, and their results were similar to those of the whole group. When the research team excluded data from supplement users and homed in on carotenoid levels in the diet, they found “that women with higher blood levels of carotenoids were at lower risk of developing breast cancer and particularly lower risk for developing breast cancers that recurred or were lethal.”
Elaissen further pointed out that
there are a number of differences between our study and the LACE study that could have contributed to these seemingly different findings, including supplement use vs blood levels. LACE focused on supplement use after diagnosis, and women reported their use an average of 2 years after diagnosis, when many women would have already completed treatment. Another study that has examined postdiagnosis carotenoids is the WHEL trial, which sought to increase fruit and vegetable intake in women diagnosed with breast cancer.
The Women’s Healthy Eating and Living (WHEL) trial was started in 2002. Epidemiological studies had linked diet and prognosis in breast cancer, and many women were attempting to modify their diets and “eat better.” The WHEL trial tested the hypothesis that a diet high in fruits and vegetables and low in fats would reduce risk of secondary cancer events and increase survival rates. The study randomly assigned 3,088 such women to either an intensive diet intervention, which consisted of “five vegetable servings, 16 ounces of vegetable juice, three fruit servings, 30 grams of fiber, and 15-20% energy from fat” or to a comparison group.12,13
While the WHEL trial was in progress, a much smaller study in China compared diets of 127 breast cancer patients against 632 healthy women and reported an inverse association between dietary β-carotene and occurrence of a primary breast cancer. Women in the upper quartile of carotenoid consumption had a 57% lower risk of breast cancer than the women in the lowest quartile.14 While these results were encouraging, the WHEL trial results were initially discouraging.
The first WHEL results, published in the Journal of the American Medical Association in July 2007, were a huge disappointment. The encouraged dietary pattern had no significant impact on primary outcomes. After a mean follow-up of 7.3 years, the 1,537 women in the intervention group had a cancer recurrence rate of 16.7% compared to the “control” group of 1,551 women who had a recurrence rate of 16.9%. In the intervention group, 10.1% of the women had died and 10.3% in the control group had. These slight differences were not significant.15
A second look at the WHEL participants has produced some interesting and significant findings. In particular, research that examined serum carotenoid levels at enrollment in the study now suggests carotenoids have a protective effect. Rock et al looked for an association between plasma carotenoid concentrations measured at the start of the study and found associations with their outcome measures: “Women in the highest quartile of plasma total carotenoid concentration had significantly reduced risk for a new breast cancer event (HR, 0.57; 95% CI, 0.37 to 0.89), controlled for covariates influencing breast cancer prognosis.”16 In other words, high β-carotene levels were associated with a 43% reduction in risk of recurrence.16
A more recent 2011 analysis of the WHEL cohort data reported that high baseline consumption of vegetables alone was associated with a 31% reduction in recurrence risk and in women who took tamoxifen a 44% reduction. High baseline consumption of cruciferous vegetables alone reduced risk by 35%. For women using tamoxifen who reported above average cruciferous vegetable intake and who were in the upper tertile for total vegetable intake, risk of recurrence was reduced 52%.17
Should the carotenoids get credit for these better outcomes, or are they just a marker for total vegetable consumption? Association does not prove causation.
Elaissen is still hesitant about women taking carotenoid supplements (personal communication, May 4, 2015):
Given the history of β-carotene supplement use and lung cancer, use of carotenoid supplements is not recommended. However, increasing consumption of colorful fruits and vegetables that are high in carotenoids have many health benefits. Our study and others suggest that reduced risk of breast cancer may be one of the many health benefits.
Greenlee’s study on the LACE cohort that reported the troubling positive association between carotenoid intake and breast cancer recurrence looked at supplement use rather than diet or blood levels.11 Could this explain why the positive association observed in her study conflicts with Eliassen’s results? Or is there something about the carotenoids and cancer that we still have to learn?
