Causal relationship between current alcohol intake and incidence of breast cancer in postmenopausal women
Contributors
-
Fjorida Llaha1
- 1 UiT The Arctic University of Norway, Department of Community Medicine, Tromsø, Norway
Description
This DAG represents the causal relationship between current alcohol intake and incidence of breast cancer in postmenopausal Norwegian women. Women were randomly recruited in the Norwegian Women and Health (NOWAC) cohort study from 1996 to 2004. For this analysis we included 32 735 women included who reported that they consumed alcohol and were not prevalent cancer cases.
This study was designed to address the effect of light-moderate drinking since the majority of women in this cohort (~95%) reported no more than 1 drink/day. Except for the outcome, other nodes in the DAG represent self-reported variables collected at baseline only (1996-2004), same time as the current alcohol intake. The outcome information was collected prospectively until December 2020 from the national cancer registries. Nodes in this DAG are presented shortly, for example breastfeeding duration represents breast feeding duration in months for each born child. More details on how each variable was collected and computed are given in the supplementary material of the published paper: https://doi.org/10.2147/CLEP.S531643.
The low-moderate drinking, Norwegian context, and background knowledge from the NOWAC data were taken into account when depicting the DAG for this specific cohort. For example, only education was used as an indicator socioeconomic status and this choice was based on these assumptions: 1) NOWAC studies found that variation in cancer risk was most evident by education levels than by income; 2) other Norwegian studies found that risk of cancer varies meaningfully by income among Norwegian men but not women; and 3) Norwegian studies consistently find that education influences lifestyle choices and (including alcohol intake) and medical behaviors.
Another example is the lack of an arrow from alcohol intake to BMI (body mass index). Empirical evidence shows that heavy drinking affects body weight status, evidence mainly found for men. The assumption in the presented DAG was that this relationship is not applicable to the NOWAC cohort, which consists of women who are generally low to moderate drinkers. These and other assumptions together with the empirical evidence are provided in the supplementary material of the published paper: https://doi.org/10.2147/CLEP.S531643.
But these assumptions may be wrong, and since I plan to continue working on alcohol and health outcomes, any discussion is greatly appreciated!
Visualisation
dag {
bb="0,0,1,1"
"Age at menarche" [pos="0.840,0.789"]
"Breast cancer in family" [pos="0.500,0.201"]
"Breastfeeding duration" [pos="0.087,0.814"]
"Calcium intake" [pos="0.596,0.735"]
"Current alcohol" [exposure,pos="0.194,0.577"]
"Fruits and vegetables" [pos="0.375,0.692"]
"Lifetime alcohol " [pos="0.077,0.426"]
"Medical behaviours" [pos="0.321,0.420"]
"Menopause hormone therapy" [pos="0.679,0.907"]
"Oral contraceptives " [pos="0.561,0.929"]
"PM Breast cancer" [outcome,pos="0.855,0.624"]
"Pack-years smoked " [pos="0.901,0.250"]
"Parity / bearing child & age at first birth" [pos="0.319,0.937"]
"Physical activity" [pos="0.614,0.416"]
Age [pos="0.405,0.041"]
BMI [pos="0.754,0.063"]
Education [pos="0.146,0.129"]
Height [pos="0.904,0.536"]
"Age at menarche" -> "PM Breast cancer"
"Age at menarche" -> BMI
"Breast cancer in family" -> "Medical behaviours"
"Breast cancer in family" -> "PM Breast cancer"
"Breastfeeding duration" -> "Lifetime alcohol "
"Breastfeeding duration" -> "PM Breast cancer"
"Breastfeeding duration" -> "Pack-years smoked "
"Breastfeeding duration" -> BMI
"Calcium intake" -> "Current alcohol"
"Calcium intake" -> "PM Breast cancer"
"Calcium intake" -> BMI
"Current alcohol" -> "PM Breast cancer"
"Fruits and vegetables" -> "Current alcohol"
"Fruits and vegetables" -> "PM Breast cancer"
"Fruits and vegetables" -> BMI
"Lifetime alcohol " -> "Current alcohol"
"Lifetime alcohol " -> "PM Breast cancer"
"Medical behaviours" -> "Calcium intake"
"Medical behaviours" -> "Current alcohol"
"Medical behaviours" -> "Fruits and vegetables"
"Medical behaviours" -> "Menopause hormone therapy"
"Medical behaviours" -> "Physical activity"
"Medical behaviours" -> BMI
"Menopause hormone therapy" -> "PM Breast cancer"
"Menopause hormone therapy" -> "Pack-years smoked "
"Menopause hormone therapy" -> "Physical activity"
"Menopause hormone therapy" -> BMI
"Oral contraceptives " -> "Breastfeeding duration"
"Oral contraceptives " -> "Parity / bearing child & age at first birth"
"Oral contraceptives " -> BMI
"Pack-years smoked " -> "Calcium intake"
"Pack-years smoked " -> "Current alcohol"
"Pack-years smoked " -> "Fruits and vegetables"
"Pack-years smoked " -> "Lifetime alcohol "
"Pack-years smoked " -> "PM Breast cancer"
"Pack-years smoked " -> "Physical activity"
"Pack-years smoked " -> BMI
"Parity / bearing child & age at first birth" -> "Breastfeeding duration"
"Parity / bearing child & age at first birth" -> "Calcium intake"
"Parity / bearing child & age at first birth" -> "Current alcohol"
"Parity / bearing child & age at first birth" -> "Fruits and vegetables"
"Parity / bearing child & age at first birth" -> "Lifetime alcohol "
"Parity / bearing child & age at first birth" -> "PM Breast cancer"
"Parity / bearing child & age at first birth" -> "Pack-years smoked "
"Parity / bearing child & age at first birth" -> "Physical activity"
"Parity / bearing child & age at first birth" -> BMI
"Physical activity" -> "Calcium intake"
"Physical activity" -> "Current alcohol"
"Physical activity" -> "Fruits and vegetables"
"Physical activity" -> "PM Breast cancer"
"Physical activity" -> BMI
Age -> "Calcium intake"
Age -> "Current alcohol"
Age -> "Fruits and vegetables"
Age -> "Lifetime alcohol "
Age -> "Medical behaviours"
Age -> "Menopause hormone therapy"
Age -> "PM Breast cancer"
Age -> "Pack-years smoked "
Age -> "Parity / bearing child & age at first birth"
Age -> "Physical activity"
Age -> BMI
Age -> Education
Age -> Height
BMI -> "PM Breast cancer"
Education -> "Breastfeeding duration"
Education -> "Calcium intake"
Education -> "Current alcohol"
Education -> "Fruits and vegetables"
Education -> "Lifetime alcohol "
Education -> "Medical behaviours"
Education -> "Menopause hormone therapy"
Education -> "Oral contraceptives "
Education -> "PM Breast cancer"
Education -> "Pack-years smoked "
Education -> "Parity / bearing child & age at first birth"
Education -> "Physical activity"
Education -> BMI
Height -> "PM Breast cancer"
}
Discussion
No comments yet. Be the first to comment!