2017 Section 7 Green Book

high school degree or less. Approximately 63% were categorized as overweight or obese, and 55% had one or more comorbid conditions. Head and neck cancers included 53% oral, 38% laryngeal, and 10% pharyngeal cancer, with 17% in stage III and 39% in stage IV. Approximately 57% received surgery, 77% radiation, and 41% chemotherapy. Given that only 10% of the cancers were pharyngeal, HPV status is not likely to play a major role in this group of patients. Table II includes a summary of the adjusted mean FACT HRQOL scores for those factors that were statisti- cally significantly related to the FACT measure for each HRQOL domain at the significance level of 0.05. Physical Well-Being Statistically significant factors associated with better physical well-being included older age, higher education, private insurance, no current tobacco use, somewhat or very much current alcohol use, no comorbidities, early stage cancer, currently without a feeding tube, and received surgery. The interaction between receiving radia- tion therapy and time was statistically significant. Although radiation therapy had a strong effect on HRQOL early after completion of therapy, QOL returned to that of patients not receiving radiation therapy by 4 years after treatment. Emotional Well-Being Statistically significant factors associated with bet- ter emotional well-being included being of African- American race, older age, higher education, not cur- rently using tobacco, no comorbid conditions, oral cancer (compared to laryngeal cancer), and early stage of can- cer. Both the interaction between feeding tube and time and between lymph node status and time were statisti- cally significant (quadratic). Figure 1 shows that the emotional well-being for those who continued to need a feeding tube decreased over time, whereas those who never needed a feeding tube gradually increased. For lymph node status, emotional well-being gradually decreased over time, whereas those without lymph node involvement had slightly better well-being. This may be because of the late fibrosis seen in the necks of patients, with advanced nodal disease treated with multimodality therapy. Social Well-Being Statistically significant factors associated with bet- ter social well-being included being female, living with a spouse or significant other, not currently using tobacco, and received surgery. Functional Well-Being Statistically significant factors associated with bet- ter functional well-being include African American race, older age, higher education level, private insurance cov- erage, not currently using tobacco, no comorbid condi- tion, oral (compared to laryngeal) cancer, early stage of

cancer, and not currently having a feeding tube. The interaction between receiving radiation and time was statistically significant. Those receiving radiation ther- apy had poorer functional status early in treatment, and then over time returned to similar functional levels as those who did not receive radiation treatment approxi- mately 2 years after diagnosis. Head and Neck Cancer Symptoms Statistically significant factors associated with reduced HNC symptoms include older age, higher education, having private insurance options, not currently using tobacco, some- what or very much current alcohol use, no comorbid condi- tions, early stage cancer, and no feeding tube currently. There were significant time interactions with race (quad- ratic), income (quadratic), number supported in household (linear), and received radiation therapy (quadratic). Symp- toms for African Americans improved more quickly over time than non-Hispanic whites, who only reported gradual improvement in symptoms. Figure 2 shows that those who received radiation therapy experienced more symptoms dur- ing and posttreatment with improvement over time com- pared to those who did not undergo radiation therapy. The interactions of time with income and time with number sup- ported in the household were not clinically meaningful in terms of their differences (results not shown). DISCUSSION Having a HNC negatively impacts an individual’s health-related quality of life. This study followed HNC survivors over 5 years and found a number of demo- graphic, behavioral, and clinical factors that were associ- ated with different levels of HRQOL. Many of the HRQOL initial negative impacts, however, improve over time. There were several demographic factors that were consistently associated with HRQOL. Older patients reported better physical, emotional, and functional well- being, and fewer symptoms than younger HNC patients, even after adjusting for comorbidity status and treat- ment (when a significant factor). This conflicts with Hammerlid et al. 21 and Penedo et al.’s 9 finding that younger individuals had better posttreatment HRQOL, and the Ronis et al. 15 study that found no relationship between age and HRQOL. Our population-based study findings, with a larger number of participants, could reflect that younger patients often receive more intense multimodality therapy 39 or are less accepting of the cos- metic and functional repercussions of therapy. Higher educational status was also associated with a better long-term HRQOL, which could reflect that education (as an indicator of socioeconomic status) is associated with better access to care and support networks. Studies by Fang et al. 13 and Kugaya et al. 18 also supported the positive association between education level and HRQOL; however, Ronis et al. 15 found no relationship in a sample from Michigan. Private insurance was associ- ated with better physical and functional well-being and less symptoms compared with no or government insur- ance. Lack of insurance is associated with greater

Reeve et al.: Factors Associated With Quality of Life

132