Academic Commons

Theses Doctoral

Identifying distinct trajectories of health behaviors after a breast cancer diagnosis

Shi, Zaixing

Breast cancer (BC) survivors are at increased risk of cancer recurrence, a second cancer, and non-cancer comorbidities. Previous studies suggest that many women adopt a spontaneous change in lifestyle after a BC diagnosis in hope of achieving a better survival outcome. While this observation has led to the suggestion that a BC diagnosis is a “teachable moment” for improving health behaviors, other conflicting studies report that BC survivors do not make positive changes in health behaviors following a breast cancer diagnosis. Although previous studies suggest that receipt of cancer chemotherapy and hormonal therapy is associated with weight loss or weight gain, the association between post-diagnosis weight change with changes in lifestyle has not been studied in detail. The majority of prior studies of post-diagnosis changes in behavior and weight have examined the mean change between two time points, and therefore may over simplify the trajectory of change over time due to lack of more granular data. New methods are needed to examine the distribution and correlates of behavior/weight trajectories following the BC diagnosis.
In my dissertation, a systematic literature review was conducted to evaluate the evidence regarding the frequency, magnitude and pattern of post-diagnosis changes in diet [fruit/vegetable (F/V), dietary fat], physical activity [moderate to vigorous physical activity (MVPA) and sedentary behaviors], alcohol intake, and body weight among BC survivors. A total of 66 studies were included in the systematic review. These studies suggest that after a breast cancer diagnosis, women are less likely to engage in MVPA and more likely to reduce alcohol intake. Previous studies suggested that women may experience weight change after a BC diagnosis, although there were strong evidence showing both weight gain and weight loss were common. The reports of changes in diet and sedentary behavior following a BC diagnosis are limited and inconclusive about the direction of change. The results of the review suggested that there is wide variation in post-diagnosis lifestyle changes among BC survivors. However, very few studies have investigated the variability in multiple behavior trajectories following a BC diagnosis.
In this dissertation, I made use of a population of 4,505 women newly diagnosed with a BC and enrolled in the Kaiser Permanente Northern California Pathways Study. I used a combination of statistical methods, including a semi-parametric, group-based trajectory modeling and a non-parametric K-means for longitudinal data analysis, to identify latent trajectories groups that are unobserved clusters of individuals following similar trajectories of a behavior. These analyses tested the hypotheses that in the 24 months following a breast cancer diagnosis, women follow a mixture of lifestyle (F/V, dietary fat, MVPA, sedentary behavior, alcohol) and body mass index (BMI) trajectories, which can be stable, temporarily increase or temporarily decrease. My analysis identified multiple distinct trajectories of lifestyle behaviors and BMI during the first 24 months after a BC diagnosis. The trajectory analysis results suggest that the large majority of women maintained their lifestyles following a BC diagnosis. Socioeconomic status, dispositional optimism, perceived social support, and the severity of CIPN during active treatment were associated with the post-diagnosis trajectories of. Furthermore, the BMI trajectories were stable over the first 24 months following a BC diagnosis. The BMI trajectories were associated with trajectories of F/V, dietary fat intake, MVPA, sedentary behavior and alcohol intake over the same period, independent of demographic characteristics, tumor characteristics and cancer treatment received.
In summary, previous studies suggest that women may spent fewer time on MVPA and drink less alcohol after a BC diagnosis, while both weight gain and loss are common post diagnosis. In a trajectory analysis of 4505 BC survivors enrolled in the Pathways Study, I did not observe any latent trajectory of meaningful change in health behavior or BMI in the first 24 months after a BC diagnosis in the Pathways Study. Instead, my analysis suggests that most women maintained their body weight following a BC diagnosis. The BMI trajectories were strongly associated with trajectory of F/V, dietary fat intake, MVPA, sedentary behavior, and alcohol intake over the same period, independent of demographic characteristics, tumor characteristics and receipt of cancer therapies. These results suggest that there is an absence of spontaneous changes in lifestyle behaviors after BC diagnosis and the importance of maintaining a healthy lifestyle in weight management after a BC diagnosis. Future studies should examine the associations of these health behaviors and BMI trajectories and BC prognosis to better understand the effect of post-diagnosis changes in lifestyle and weight on BC-specific and all-cause mortality.

Files

  • thumnail for Shi_columbia_0054D_14167.pdf Shi_columbia_0054D_14167.pdf application/pdf 4.4 MB Download File

More About This Work

Academic Units
Epidemiology
Thesis Advisors
Greenlee, Heather
Degree
Ph.D., Columbia University
Published Here
August 25, 2017
Academic Commons provides global access to research and scholarship produced at Columbia University, Barnard College, Teachers College, Union Theological Seminary and Jewish Theological Seminary. Academic Commons is managed by the Columbia University Libraries.