Young Breast Cancer Survivors Study

This is an observational epidemiologic study in which the investigators will collect data from approximately 384 participants at one point in time using online questionnaires (REDCaP and the NCI Diet History Questionnaire III). Recruitment will be conducted via breast cancer survivor social media support groups and advocacy groups. Electronic consent will be granted at the beginning of each online questionnaire. Data collection will occur via self-report in the location chosen by the participant using an online questionnaire. The investigators will use established validated questionnaires used by other previously conducted cohorts to enquire about demographics, occupation, cancer screening history, family history of cancer, comorbidities, cancer and other comorbidity treatment including fertility-related, HRQoL, dietary intake using the online NCI Dietary History Questionnaire III which includes dietary supplement use, residential history, hormonal status, physical activity, tobacco product and alcohol use, experiences with racism, social support, and information on health care utilization.

To accomplish Aim 1, the investigators will use the NCI Diet History Questionnaire III, an online 135-item food frequency questionnaire with 26 dietary supplement questions, reflecting the past one month of intake to estimate food and nutrient intakes and overall dietary patterns.Health related quality of life (HRQoL) is a multidimensional concept that not only includes physical, psychological and social domains, but may also encompass other domains such as cognitive functioning. Cancer patients also exhibit many symptoms (e.g., fatigue, pain, sleep disturbance) that are not measurable directly from laboratory tests. Thus, assessing HRQoL and these symptom burdens among cancer survivors will need to rely on patients' self-reports, measured by validated instruments especially for patients with breast cancer, and includes the Functional Assessment of Cancer Therapy (FACT-B).

Descriptive statistics will be presented as mean (standard deviation) and median (inter quartile range) for continuous variables and as frequency (percentage) for categorical variables. Data visualization tools such as histogram, boxplot, scatterplot and line plot will be employed to assess the trends and outliers in the data. Bivariate association between categorical variables will be tested for statistical significance using Chi-square test or exact test. Differences in continuous variables will be tested for statistical significance using Kruskal-Wallis test. HRQoL data typically are not distributed normally with left-skewed distributions and potential ceiling effects requiring consideration of alternative estimators in multivariate models. A number of different alternative models have been proposed over the standard ordinary least squares approach including beta regression, tobit regression, and two-part modeling. The investigators will assess the distribution of the primary HRQoL measures in building analytic models. With healthy dietary pattern as the primary independent variable and measures of HRQoL serving as primary dependent variable, the investigators expect to find that as diet quality increases, HRQoL increases.

For Aim 2, the investigators will ask study participants to complete questionnaires related to racism, fatalism, and demographics and the investigators will geocode residential histories of participants to measure the role of racism, fatalism, income, education, and proximity to treatment facilities in treatment adherence. The investigators will use geographic information systems to geocode participants' residential histories. Participants will be asked the ZIP code of residence at the time of diagnosis, as well as the ZIP code of the treatment facility in which they received care. If the ZIP code of the treatment facility is unknown, the participant can give the city and state, in which a ZIP code of the central point of the city will be found and used for analysis. Using Microsoft Excel, and the list of latitudes and longitudes by ZIP code, provided by the United States Census Bureau (find citation), the distance in miles between the place of residence and place of treatment will be determined. This distance calculation can be used to help evaluate treatment adherence and the distance to treatment. Residential histories will also be used to determine participant rurality. Regarding the many competing definitions and classification schemes for rurality, Hall et al. found dichotomous definitions mask heterogeneity relevant to health research and studies of accessibility. The investigators will instead use the Rural-Urban Commuting Area (RUCA) codes developed by the Office of Rural Health Policy of the Health Resources and Services Administration and the Economic Research Service of the United States Department of Agriculture (USDA). RUCA codes combine information on population density, urbanicity, and daily commuting patterns to classify census tracts into 22 distinct codes, which can then be consolidated into more manageable classifications.

Descriptive statistics will be presented as mean (standard deviation) and median (inter quartile range) for continuous variables and as frequency (percentage) for categorical variables. Data visualization tools such as histogram, boxplot, scatterplot and line plot will be employed to assess the trends and outliers in the data. The investigators will explore the associations between treatment adherence and each measure of social determinants (such as poverty to income ratio, racism, education, and proximity to treatment facility) using mixed effect logistic regression. The model will be sequentially adjusted for the effect of non-modifiable and modifiable confounders such as age, cancer stage, treatment type, and insurance status.

Recruitment of participants is expected to be ongoing for ~12 months.