Graduation Year

2026

Document Type

Dissertation

Degree

Ph.D.

Degree Name

Doctor of Philosophy (Ph.D.)

Degree Granting Department

School of Aging Studies

Major Professor

William E. Haley, Ph.D.

Committee Member

Brent J. Small, Ph.D.

Committee Member

Karen L. Fingerman, Ph.D.

Committee Member

Alison Salloum, Ph.D.

Committee Member

Gizem Hueluer, Ph.D.

Keywords

Inflammation, Social connections, Social relationships, Biomarkers, Survival Analysis

Abstract

Cancer is the second leading cause of death both in the United States and globally and for individuals over sixty-five. In 2026, approximately 2,114,850 new cancer cases are projected to emerge according to the National Center for Health Statistics (Siegel et al., 2026). Cancer incidence is increasing, making older adults sixty-five and older at risk of frailty and mortality following cancer. Although social isolation, loneliness, lack of social support, and social strain have been found to be risk factors for poor health and mortality, it is not clear the extent to which social relationships are associated with incident cancer, and incident frailty and all-mortality for people with cancer, and whether inflammation plays a mediating role in these associations. Therefore, this dissertation was completed using longitudinal data from the Health and Retirement Study (HRS 2006-2020) to address the following aims: to assess whether baseline and changes in social isolation, loneliness, and social strain are associated with incident cancer and whether inflammation by C-Reactive Protein (CRP) explains these associations. Another aim was to investigate among participants with cancer incidence, whether levels of and changes in social isolation, loneliness, social support and strain relate to incident frailty and all-cause mortality, and whether CRP is a causal mechanism driving these associations.

From a sample of 12,762 individuals without a history of cancer, survival analyses were run to assess how social factors such as social isolation, loneliness and social strain independently and cumulatively related to cancer incidence. Baseline and changes over time in these social factors were independently examined in association to incident cancer. Associations between the combined effects of high social isolation, high loneliness, and moderate social strain were assessed at study baseline (2006/08) and T2 (2010/12) in relation to incident cancer at follow-up. Mediation by CRP was investigated for significant associations only. After inclusion/exclusion criteria, 903 participants developed cancer. We found social isolation at baseline was associated with increased risk of cancer at follow-up. This paper also identified that the combined effects of high social isolation, loneliness and social strain imposed a greater risk for cancer incidence. High inflammation did not mediate the association between social isolation and cancer incidence.

Among participants with cancer, 412 participants developed incident frailty and 328 had all-cause mortality at follow-up. Survival analyses were run to assess how social factors such as social isolation, loneliness and social strain and support independently and cumulatively related to incident frailty and all-cause mortality following cancer onset. Baseline and changes over time in these social factors were independently examined in association to incident frailty and mortality. Next, associations between the combined effects of high social isolation, high loneliness, and moderate social strain and social support were assessed at study baseline (2006/08) and T2 (2010/12) in relation to incident frailty and all-cause mortality following cancer. Finally, a possible causal mechanism of inflammation as measured by CRP was investigated for significant associations in unadjusted and adjusted models. We found loneliness at baseline and increases in loneliness increased the risk of incident frailty after cancer; and that both social isolation and loneliness at baseline and changes over time in these indicators were associated with higher all-cause mortality after cancer, independent of sociodemographic and health variables. The combined effects of high social isolation, loneliness, social strain and support at baseline and T2 did not have a significant effect on incident frailty but combined effects at T2 did have a significant effect with respect to all-cause mortality, independent of age and sociodemographic variables only. This paper identified high CRP as a causal mechanism in the association of high loneliness at baseline and incident frailty. CRP did not have a mediating role in the associations between social isolation and all-cause mortality nor loneliness and all-cause mortality following cancer. Social support and strain do not have a significant role in incident frailty or all-cause mortality. Future work should further consider possible public health and clinical interventions to prevent adverse effects of social isolation and loneliness on cancer occurrence, and on frailty and survivorship after onset of cancer.

Download

Share

COinS