Effects of demographic and behavioral factors on social isolation in a cancer center population during COVID-19 pandemic

ABSTRACT

The COVID-19 pandemic has led to extensive social changes, which may increase the risk of experiencing social isolation, particularly in cancer patients who are at high risk of having poor outcomes if infected with COVID-19. It is unclear if pandemic-related risk mitigation behaviors (e.g., limiting attendance at social gatherings, mask wearing) influence social isolation. We leveraged data from 9,514 patients with cancer, or at high risk for cancer, from Moffitt Cancer Center to examine whether social isolation is associated with sociodemographic factors and risk mitigation behaviors. Eligible patients who had an appointment at the cancer center in the past five years, reported an email address, and had consented to the institutional biobanking protocol were invited to complete a survey regarding demographic, behavioral, and lifestyle factors as well as social isolation measured using the 4-item PROMIS Social Isolation scale. The raw PROMIS score was converted to a T-score and was split at 50, the average for the general U.S. population, and univariable and multivariable logistic regression was performed. Behavioral factors were measured on a Likert scale, ranging from “never” or “not at all”, to “very often” or “a lot”, and were evaluated continuously. Most participants were female (60.5%), Non-Hispanic White (90.3%), and had been diagnosed with cancer (89.6%);mean age was 64 years old. Only 3.5% reported ever testing positive for COVID-19, and 4.6% reported currently smoking. In univariable models, younger age, women, current smokers, and Hispanic ethnicity or Non-White race were associated with higher odds of social isolation. Among risk mitigation behaviors, leaving the house less often, attending social gatherings less often, a greater change in day-to-day life due to the pandemic, less physical contact with others outside their home, and wearing a mask more often were also associated with increased odds of social isolation. In the multivariable model including significant univariate factors, older patients (OR, per one year 0.97;95%CI 0.97-0.98) and males (OR, vs. females 0.64;95%CI 0.58-0.71) had lower odds of social isolation. Further, perceived changes in day-to-day life (OR, per one unit increase 1.64 95%CI 1.56-1.73), leaving the house less often (OR, per one unit increase 0.75;95%CI 0.71-0.80), and attending social gatherings less often (OR, per one unit increase 0.92 95%CI 0.84-0.95) remained associated with social isolation. Overall, behavior change to mitigate risk of COVID-19 infection was associated with more social isolation for cancer patients and survivors. Women, younger patients, and current smokers may be particularly at risk, suggesting targeted interventions to reduce feelings of social isolation in these populations may be warranted.