BACKGROUND/OBJECTIVES: This study aimed to use plasma metabolites to identify clusters of breast cancer survivors and to compare their dietary characteristics and health-related factors across the clusters using unsupervised machine learning. SUBJECTS/METHODS: A total of 419 breast cancer survivors were included in this crosssectional study. We considered 30 plasma metabolites, quantified by high-throughput nuclear magnetic resonance metabolomics. Clusters were obtained based on metabolites using 4 different unsupervised clustering methods: k-means (KM), partitioning around medoids (PAM), self-organizing maps (SOM), and hierarchical agglomerative clustering (HAC). The t-test, χ2 test, and Fisher’s exact test were used to compare sociodemographic, lifestyle, clinical, and dietary characteristics across the clusters. P-values were adjusted through a false discovery rate (FDR). RESULTS: Two clusters were identified using the 4 methods. Participants in cluster 2 had lower concentrations of apolipoprotein A1 and large high-density lipoprotein (HDL) particles and smaller HDL particle sizes, but higher concentrations of chylomicrons and extremely large very-low-density-lipoprotein (VLDL) particles and glycoprotein acetyls, a higher ratio of monounsaturated fatty acids to total fatty acids, and larger VLDL particle sizes compared with cluster 1. Body mass index was significantly higher in cluster 2 compared with cluster 1 (FDR adjusted-PKM < 0.001; PPAM = 0.001; PSOM < 0.001; and PHAC = 0.043). CONCLUSION The breast cancer survivors clustered on the basis of plasma metabolites had distinct characteristics. Further prospective studies are needed to investigate the associations between metabolites, obesity, dietary factors, and breast cancer prognosis.

BACKGROUND/OBJECTIVES: Isoflavones are estrogen-like compounds found in plants and their health effects remain equivocal. We investigated dietary isoflavone intake and its associated factors in Korean breast cancer survivors, with a comparison to cancer-free women. SUBJECTS/METHODS: The usual dietary intake of breast cancer survivors (n = 981, mean age 52 yrs) in 9 hospitals between 2012 and 2019 was assessed using 3-day food records or food frequency questionnaires (FFQs). They were age-matched to 2,943 cancer-free women who completed FFQs as part of a nationwide study conducted between 2012 and 2016. We used the flavonoid database of common Korean foods and the Phenol-Explorer database to estimate isoflavone intake. The contribution of each food or food group to the total isoflavone intake was calculated. The adjusted least-squares means of dietary isoflavone intake according to lifestyle and clinical factors were calculated using generalized linear models. RESULTS: Breast cancer survivors had a higher mean dietary isoflavone intake (23.59 mg/day) than cancer-free women (17.81 mg/day). Major food sources, including tofu, soybeans, and doenjang, contributed to over 70% of the isoflavone intake in both groups. When we estimated dietary isoflavone intake according to lifestyle characteristics, isoflavone intake increased with higher scores of adherence to the American Cancer Society dietary guidelines but decreased with increasing body mass index in both groups. Among cancer-free women, dietary isoflavone intake was higher among those who had never smoked and among dietary supplement users. Among breast cancer survivors, dietary isoflavone intakes did not vary with clinical characteristics, including time since surgery and estrogen receptor status. CONCLUSION Breast cancer survivors were more likely to consume isoflavones than agematched cancer-free women. Dietary isoflavone intake was associated with healthy lifestyle characteristics in women both with and without breast cancer. Further research is needed to understand the role of the higher isoflavone intake among breast cancer survivors compared to cancer-free women on their prognosis.

BACKGROUND/OBJECTIVES: This study aimed to use plasma metabolites to identify clusters of breast cancer survivors and to compare their dietary characteristics and health-related factors across the clusters using unsupervised machine learning. SUBJECTS/METHODS: A total of 419 breast cancer survivors were included in this crosssectional study. We considered 30 plasma metabolites, quantified by high-throughput nuclear magnetic resonance metabolomics. Clusters were obtained based on metabolites using 4 different unsupervised clustering methods: k-means (KM), partitioning around medoids (PAM), self-organizing maps (SOM), and hierarchical agglomerative clustering (HAC). The t-test, χ2 test, and Fisher’s exact test were used to compare sociodemographic, lifestyle, clinical, and dietary characteristics across the clusters. P-values were adjusted through a false discovery rate (FDR). RESULTS: Two clusters were identified using the 4 methods. Participants in cluster 2 had lower concentrations of apolipoprotein A1 and large high-density lipoprotein (HDL) particles and smaller HDL particle sizes, but higher concentrations of chylomicrons and extremely large very-low-density-lipoprotein (VLDL) particles and glycoprotein acetyls, a higher ratio of monounsaturated fatty acids to total fatty acids, and larger VLDL particle sizes compared with cluster 1. Body mass index was significantly higher in cluster 2 compared with cluster 1 (FDR adjusted-PKM < 0.001; PPAM = 0.001; PSOM < 0.001; and PHAC = 0.043). CONCLUSION The breast cancer survivors clustered on the basis of plasma metabolites had distinct characteristics. Further prospective studies are needed to investigate the associations between metabolites, obesity, dietary factors, and breast cancer prognosis.

BACKGROUND/OBJECTIVES: Isoflavones are estrogen-like compounds found in plants and their health effects remain equivocal. We investigated dietary isoflavone intake and its associated factors in Korean breast cancer survivors, with a comparison to cancer-free women. SUBJECTS/METHODS: The usual dietary intake of breast cancer survivors (n = 981, mean age 52 yrs) in 9 hospitals between 2012 and 2019 was assessed using 3-day food records or food frequency questionnaires (FFQs). They were age-matched to 2,943 cancer-free women who completed FFQs as part of a nationwide study conducted between 2012 and 2016. We used the flavonoid database of common Korean foods and the Phenol-Explorer database to estimate isoflavone intake. The contribution of each food or food group to the total isoflavone intake was calculated. The adjusted least-squares means of dietary isoflavone intake according to lifestyle and clinical factors were calculated using generalized linear models. RESULTS: Breast cancer survivors had a higher mean dietary isoflavone intake (23.59 mg/day) than cancer-free women (17.81 mg/day). Major food sources, including tofu, soybeans, and doenjang, contributed to over 70% of the isoflavone intake in both groups. When we estimated dietary isoflavone intake according to lifestyle characteristics, isoflavone intake increased with higher scores of adherence to the American Cancer Society dietary guidelines but decreased with increasing body mass index in both groups. Among cancer-free women, dietary isoflavone intake was higher among those who had never smoked and among dietary supplement users. Among breast cancer survivors, dietary isoflavone intakes did not vary with clinical characteristics, including time since surgery and estrogen receptor status. CONCLUSION Breast cancer survivors were more likely to consume isoflavones than agematched cancer-free women. Dietary isoflavone intake was associated with healthy lifestyle characteristics in women both with and without breast cancer. Further research is needed to understand the role of the higher isoflavone intake among breast cancer survivors compared to cancer-free women on their prognosis.

BACKGROUND/OBJECTIVES: This study aimed to use plasma metabolites to identify clusters of breast cancer survivors and to compare their dietary characteristics and health-related factors across the clusters using unsupervised machine learning. SUBJECTS/METHODS: A total of 419 breast cancer survivors were included in this crosssectional study. We considered 30 plasma metabolites, quantified by high-throughput nuclear magnetic resonance metabolomics. Clusters were obtained based on metabolites using 4 different unsupervised clustering methods: k-means (KM), partitioning around medoids (PAM), self-organizing maps (SOM), and hierarchical agglomerative clustering (HAC). The t-test, χ2 test, and Fisher’s exact test were used to compare sociodemographic, lifestyle, clinical, and dietary characteristics across the clusters. P-values were adjusted through a false discovery rate (FDR). RESULTS: Two clusters were identified using the 4 methods. Participants in cluster 2 had lower concentrations of apolipoprotein A1 and large high-density lipoprotein (HDL) particles and smaller HDL particle sizes, but higher concentrations of chylomicrons and extremely large very-low-density-lipoprotein (VLDL) particles and glycoprotein acetyls, a higher ratio of monounsaturated fatty acids to total fatty acids, and larger VLDL particle sizes compared with cluster 1. Body mass index was significantly higher in cluster 2 compared with cluster 1 (FDR adjusted-PKM < 0.001; PPAM = 0.001; PSOM < 0.001; and PHAC = 0.043). CONCLUSION The breast cancer survivors clustered on the basis of plasma metabolites had distinct characteristics. Further prospective studies are needed to investigate the associations between metabolites, obesity, dietary factors, and breast cancer prognosis.

BACKGROUND/OBJECTIVES: Isoflavones are estrogen-like compounds found in plants and their health effects remain equivocal. We investigated dietary isoflavone intake and its associated factors in Korean breast cancer survivors, with a comparison to cancer-free women. SUBJECTS/METHODS: The usual dietary intake of breast cancer survivors (n = 981, mean age 52 yrs) in 9 hospitals between 2012 and 2019 was assessed using 3-day food records or food frequency questionnaires (FFQs). They were age-matched to 2,943 cancer-free women who completed FFQs as part of a nationwide study conducted between 2012 and 2016. We used the flavonoid database of common Korean foods and the Phenol-Explorer database to estimate isoflavone intake. The contribution of each food or food group to the total isoflavone intake was calculated. The adjusted least-squares means of dietary isoflavone intake according to lifestyle and clinical factors were calculated using generalized linear models. RESULTS: Breast cancer survivors had a higher mean dietary isoflavone intake (23.59 mg/day) than cancer-free women (17.81 mg/day). Major food sources, including tofu, soybeans, and doenjang, contributed to over 70% of the isoflavone intake in both groups. When we estimated dietary isoflavone intake according to lifestyle characteristics, isoflavone intake increased with higher scores of adherence to the American Cancer Society dietary guidelines but decreased with increasing body mass index in both groups. Among cancer-free women, dietary isoflavone intake was higher among those who had never smoked and among dietary supplement users. Among breast cancer survivors, dietary isoflavone intakes did not vary with clinical characteristics, including time since surgery and estrogen receptor status. CONCLUSION Breast cancer survivors were more likely to consume isoflavones than agematched cancer-free women. Dietary isoflavone intake was associated with healthy lifestyle characteristics in women both with and without breast cancer. Further research is needed to understand the role of the higher isoflavone intake among breast cancer survivors compared to cancer-free women on their prognosis.

BACKGROUND/OBJECTIVES: This study aimed to use plasma metabolites to identify clusters of breast cancer survivors and to compare their dietary characteristics and health-related factors across the clusters using unsupervised machine learning. SUBJECTS/METHODS: A total of 419 breast cancer survivors were included in this crosssectional study. We considered 30 plasma metabolites, quantified by high-throughput nuclear magnetic resonance metabolomics. Clusters were obtained based on metabolites using 4 different unsupervised clustering methods: k-means (KM), partitioning around medoids (PAM), self-organizing maps (SOM), and hierarchical agglomerative clustering (HAC). The t-test, χ2 test, and Fisher’s exact test were used to compare sociodemographic, lifestyle, clinical, and dietary characteristics across the clusters. P-values were adjusted through a false discovery rate (FDR). RESULTS: Two clusters were identified using the 4 methods. Participants in cluster 2 had lower concentrations of apolipoprotein A1 and large high-density lipoprotein (HDL) particles and smaller HDL particle sizes, but higher concentrations of chylomicrons and extremely large very-low-density-lipoprotein (VLDL) particles and glycoprotein acetyls, a higher ratio of monounsaturated fatty acids to total fatty acids, and larger VLDL particle sizes compared with cluster 1. Body mass index was significantly higher in cluster 2 compared with cluster 1 (FDR adjusted-PKM < 0.001; PPAM = 0.001; PSOM < 0.001; and PHAC = 0.043). CONCLUSION The breast cancer survivors clustered on the basis of plasma metabolites had distinct characteristics. Further prospective studies are needed to investigate the associations between metabolites, obesity, dietary factors, and breast cancer prognosis.

BACKGROUND/OBJECTIVES: Isoflavones are estrogen-like compounds found in plants and their health effects remain equivocal. We investigated dietary isoflavone intake and its associated factors in Korean breast cancer survivors, with a comparison to cancer-free women. SUBJECTS/METHODS: The usual dietary intake of breast cancer survivors (n = 981, mean age 52 yrs) in 9 hospitals between 2012 and 2019 was assessed using 3-day food records or food frequency questionnaires (FFQs). They were age-matched to 2,943 cancer-free women who completed FFQs as part of a nationwide study conducted between 2012 and 2016. We used the flavonoid database of common Korean foods and the Phenol-Explorer database to estimate isoflavone intake. The contribution of each food or food group to the total isoflavone intake was calculated. The adjusted least-squares means of dietary isoflavone intake according to lifestyle and clinical factors were calculated using generalized linear models. RESULTS: Breast cancer survivors had a higher mean dietary isoflavone intake (23.59 mg/day) than cancer-free women (17.81 mg/day). Major food sources, including tofu, soybeans, and doenjang, contributed to over 70% of the isoflavone intake in both groups. When we estimated dietary isoflavone intake according to lifestyle characteristics, isoflavone intake increased with higher scores of adherence to the American Cancer Society dietary guidelines but decreased with increasing body mass index in both groups. Among cancer-free women, dietary isoflavone intake was higher among those who had never smoked and among dietary supplement users. Among breast cancer survivors, dietary isoflavone intakes did not vary with clinical characteristics, including time since surgery and estrogen receptor status. CONCLUSION Breast cancer survivors were more likely to consume isoflavones than agematched cancer-free women. Dietary isoflavone intake was associated with healthy lifestyle characteristics in women both with and without breast cancer. Further research is needed to understand the role of the higher isoflavone intake among breast cancer survivors compared to cancer-free women on their prognosis.

BACKGROUND/OBJECTIVES: This study aimed to use plasma metabolites to identify clusters of breast cancer survivors and to compare their dietary characteristics and health-related factors across the clusters using unsupervised machine learning. SUBJECTS/METHODS: A total of 419 breast cancer survivors were included in this crosssectional study. We considered 30 plasma metabolites, quantified by high-throughput nuclear magnetic resonance metabolomics. Clusters were obtained based on metabolites using 4 different unsupervised clustering methods: k-means (KM), partitioning around medoids (PAM), self-organizing maps (SOM), and hierarchical agglomerative clustering (HAC). The t-test, χ2 test, and Fisher’s exact test were used to compare sociodemographic, lifestyle, clinical, and dietary characteristics across the clusters. P-values were adjusted through a false discovery rate (FDR). RESULTS: Two clusters were identified using the 4 methods. Participants in cluster 2 had lower concentrations of apolipoprotein A1 and large high-density lipoprotein (HDL) particles and smaller HDL particle sizes, but higher concentrations of chylomicrons and extremely large very-low-density-lipoprotein (VLDL) particles and glycoprotein acetyls, a higher ratio of monounsaturated fatty acids to total fatty acids, and larger VLDL particle sizes compared with cluster 1. Body mass index was significantly higher in cluster 2 compared with cluster 1 (FDR adjusted-PKM < 0.001; PPAM = 0.001; PSOM < 0.001; and PHAC = 0.043). CONCLUSION The breast cancer survivors clustered on the basis of plasma metabolites had distinct characteristics. Further prospective studies are needed to investigate the associations between metabolites, obesity, dietary factors, and breast cancer prognosis.

BACKGROUND/OBJECTIVES: Isoflavones are estrogen-like compounds found in plants and their health effects remain equivocal. We investigated dietary isoflavone intake and its associated factors in Korean breast cancer survivors, with a comparison to cancer-free women. SUBJECTS/METHODS: The usual dietary intake of breast cancer survivors (n = 981, mean age 52 yrs) in 9 hospitals between 2012 and 2019 was assessed using 3-day food records or food frequency questionnaires (FFQs). They were age-matched to 2,943 cancer-free women who completed FFQs as part of a nationwide study conducted between 2012 and 2016. We used the flavonoid database of common Korean foods and the Phenol-Explorer database to estimate isoflavone intake. The contribution of each food or food group to the total isoflavone intake was calculated. The adjusted least-squares means of dietary isoflavone intake according to lifestyle and clinical factors were calculated using generalized linear models. RESULTS: Breast cancer survivors had a higher mean dietary isoflavone intake (23.59 mg/day) than cancer-free women (17.81 mg/day). Major food sources, including tofu, soybeans, and doenjang, contributed to over 70% of the isoflavone intake in both groups. When we estimated dietary isoflavone intake according to lifestyle characteristics, isoflavone intake increased with higher scores of adherence to the American Cancer Society dietary guidelines but decreased with increasing body mass index in both groups. Among cancer-free women, dietary isoflavone intake was higher among those who had never smoked and among dietary supplement users. Among breast cancer survivors, dietary isoflavone intakes did not vary with clinical characteristics, including time since surgery and estrogen receptor status. CONCLUSION Breast cancer survivors were more likely to consume isoflavones than agematched cancer-free women. Dietary isoflavone intake was associated with healthy lifestyle characteristics in women both with and without breast cancer. Further research is needed to understand the role of the higher isoflavone intake among breast cancer survivors compared to cancer-free women on their prognosis.