Relationship between Carotenoids, Retinol, and Estradiol Levels in Older Women

BACKGROUND In vitro evidence suggests anti-estrogenic properties for retinol and carotenoids, supporting a chemo-preventive role of these phytochemicals in estrogen-dependent cancers. During aging there are significant reductions in retinol and carotenoid concentrations, whereas estradiol levels decline during menopause and progressively increase from the age of 65. We aimed to investigate the hypothesis of a potential relationship between circulating levels of retinol, carotenoids, and estradiol (E2) in a cohort of late post-menopausal women. METHODS We examined 512 women ≥ 65 years from the InCHIANTI study. Retinol, α-caroten, β-caroten, β-criptoxantin, lutein, zeaxanthin, and lycopene levels were assayed at enrollment (1998-2000) by High-Performance Liquid Chromatography. Estradiol and testosterone (T) levels were assessed by Radioimmunometry (RIA) and testosterone-to-estradiol ratio (T/E2), as a proxy of aromatase activity, was also calculated. General linear models adjusted for age (Model 1) and further adjusted for other confounders including Body Mass Index (BMI) BMI, smoking, intake of energy, lipids, and vitamin A; C-Reactive Protein, insulin, total cholesterol, liver function, and testosterone (Model 2) were used to investigate the relationship between retinol, carotenoids, and E2 levels. To address the independent relationship between carotenoids and E2 levels, factors significantly associated with E2 in Model 2 were also included in a fully adjusted Model 3. RESULTS After adjustment for age, α-carotene (β ± SE = -0.01 ± 0.004, p = 0.02) and β-carotene (β ± SE = -0.07 ± 0.02, p = 0.0007) were significantly and inversely associated with E2 levels. α-Carotene was also significantly and positively associated with T/E2 ratio (β ± SE = 0.07 ± 0.03, p = 0.01). After adjustment for other confounders (Model 2), the inverse relationship between α-carotene (β ± SE = -1.59 ± 0.61, p = 0.01), β-carotene (β ± SE = -0.29 ± 0.08, p = 0.0009), and E2 persisted whereas the relationship between α-carotene and T/E2 ratio was attenuated (β ± SE = 0.22 ± 0.12, p = 0.07). In a fully adjusted model (Model 3), only β-carotene (β ± SE = -0.05 ± 0.02, p = 0.03) was significantly and inversely associated with E2 levels independent of α-carotene. No association was found between retinol, total non-pro-vitamin A carotenoids, lutein, zeaxanthin, and lycopene, and E2 levels. CONCLUSIONS In older women, β-carotene levels are independently and inversely associated with E2.