Health behaviors among head and neck cancer survivors

Potential participants were identified and screened by our research personnel through the tumor registry at the Abramson Cancer Center (ACC) of the University of Pennsylvania as well as affiliated sites (Chester County Hospital, Lancaster General Hospital, Pennsylvania Hospital, Penn Presbyterian Medical Center) to allow for broader reach. Potential participants were eligible if they: (a) had a diagnosis and at least partial treatment of HNC as reported in the registry; (b) were between 18 and 70 years of age; (c) were more than 1-year post-diagnosis; (d) had a recorded confirmation of HPV status; and e) could understand written English.

Potential participants were contacted via mail with an opt out option. Those who did not opt out were contacted by phone or email and asked to participate in a study. If potential participants were unavailable by telephone after two attempts, staff contacted them by mail. Potential participants received an explanation of the study, informed consent documentation, a self-reported questionnaire, and a self-addressed stamped envelope for return of materials. Individuals not returning the materials within the first three weeks received up to two additional phones calls to encourage completion and return of the study materials. This process was repeated on an ongoing, weekly basis. Once data were received, they were entered into a secured database.

Measures

A few specific instruments make up the self-reported questionnaire.

Demographic and Medical Variables – All participants responded to standard demographic questionnaires assessing age, ethnic/racial identification, marital status, number of offspring, educational level, income, insurance status, living situation, and presence of chronic medical comorbidities.

Health Behaviors – HBs included items recommended by Glasgow et al. assessing aerobic activity, strength-flexibility, leisure time activity, smoking, alcohol consumption, and access to a physician [4]. This battery assesses physical activity levels using the RAPA and the Godin Leisure Time Activity Scale [33]. Drinking behavior was assessed using items from the BRFSS [34] and cigarette smoking with three items from national health surveys which query smoking history, current smoking status, and extent of smoking. Eating patterns were assessed with items from the STC-Diet [35] that assesses food patterns vs. nutrient or fat intake. We assessed cancer and non-cancer health surveillance behaviors using tools developed for other, similar applications.

Illness Perceptions – The Illness Perception Questionnaire-Revised (IPQ-R) [36] was used to quantitatively assess attributes of illness. The IPQ-R is a self-report, 84-item questionnaire. It assesses illness identity causality, consequences, timeline, control/cure attributions, illness coherence, and emotional representations, and keyed specifically to the HNC illness experience.

Psychological Distress – We assessed cancer-specific distress using the Impact of Event scale (IES) [37], which has been widely used in a variety of cancer patient and survivor populations. Past studies report acceptable internal consistency and discrimination between situations of varying stressfulness. Instructions for the IES key responses to specific experiences for this study was the HNC experience.

Quality of Life – We assessed QoL in two ways. The University of Washington QoL Questionnaire (UW-QOL) [38] is a brief assessment of 12 individual domains (activity, appearance, chewing, dry mouth, employment, pain, recreation, saliva, shoulder function, speech, swallowing, and taste) relevant to HNC survivors. It also allows for more global ratings of QoL, has been used in numerous examinations of HNC outcomes [39, 40], and non-cancer normative values are available. The RAND Medical Outcomes Study Short-Form-36 Health Survey (SF-36) [41] was used to assess emotional and social functioning and role limitations. This measure has been used extensively among various medical populations, yielding rich normative data [42,43,44,45]. The psychometric properties of the measure are well-established [46,47,48,49,50], and it has been used to demonstrate validate several QoL instruments [51].

Statistical analyses and results

Medical record abstraction included patient demographics, primary treatment, tumor site, and HPV status.

Analysis was restricted to enrolled patients who returned surveys.

A total of 905 individuals were initially screened for eligibility. Of these, 451 were eligible and approached. A total of 102 (23%) patients participated in the study, and usable data were collected from 93 participants. Table 1. presents the demographic characteristics of the sample.

Table 1 Sample characteristics

As can be seen in Table 1, the sample was predominantly white, middle, aged, married, well-educated, and well-resourced. Indeed, fully 63% of the sample reported an annual household income of greater than $100,000. Medical record review found that almost 3/4 of the sample were HPV positive.

Results for aim 1

Describe the range of HBs relevant to both recurrence and survivorship among HNC survivors and determine whether these differ by HPV status and demographics.

Smoking and Alcohol use. Although current smoking was rare (7%), historical use of tobacco was common (48%). Current alcohol use was also common (65%). Among those who reported alcohol use, there was an average of 12 days per month in which alcohol was consumed, though this ranged from 1 to 30 days. On those days in which alcohol was consumed participants reported an average of 2.1 drinks per day. Binge drinking (consuming > 5 drinks in one sitting) was reported by 22% of those who drank, with an average of 3.5 binge drinking sessions per month among those reporting any binge drinking. There was no relationship between HPV and smoking status. Women were more likely than men, however, to be current smokers (p < 0.05). There was no relationship between current alcohol use status, days drinking per month, or number of drinks on average per drinking session and either HPV status or gender.

Diet and Weight. Scores on Starting the Conversation—Diet (SCT-Diet), our measure of nutritional behavior can range from 0 to 16 with lower scores indicating better nutrition. Overall, the sample mean was 7.1 (SD = 2.4) with scores ranging from 0 to 13. BMI was calculated from self-reported height and weight. Forty-one percent of the overall sample were Normal or Underweight, while 59% were overweight or obese.

There was no relationship between HPV status or gender with respect to nutritional behavior. Men had a higher BMI (M = 27.3) than women (24.0; F = 7.01, p < 0.01), and were more likely to be obese (p < 0.05).

Physical Activity. With respect to the overall sample, 59% reported adequate aerobic exercise and 64% adequate strength and flexibility on the RAPA. With respect to leisure time activity, 18% would be classified as sedentary, 19% as moderately active, and 64% as active. Individuals with HPV positive status were marginally more likely to meet criteria for adequate strength and flexibility than those with negative status (p = 0.07), although there was no difference with respect to HPV status for aerobic activity or leisure time activity status. Men were more likely than women to report adequate aerobic activity (p < 0.01), though there was no gender difference for strength and flexibility or leisure time activity status.

Medical Health Behaviors. All participants reported having a current primary care physician and 92% reported having had a physical examination in the previous 12 months, precluding examination of differences.

Results for aim 2

Describe QoL and distress outcomes among HNC survivors and determine whether these differ by HPV status and demographics.

Quality of Life. Table 2. presents the overall mean scores for the SF-36 subscales assessing general health-related QoL as well as normative values from a sample of US. QoL on the SF-36 is similar to population values for Role Limitations due to physical functioning, social functioning, mental health, and role limitations due to mental functioning. HNC survivors reported higher physical functioning, less bodily pain, but lower vitality and general health than normative values. There were no significant differences between groups based on HPV status. Compared to men, women reported significantly more Bodily Pain, and greater disruption of Physical Functioning, Role Limitations due to Physical Functioning, and Social Functioning (all p < 0.05).

Table 2 SF-36 values and normative values

With respect to the HNC cancer specific UW-QOL measure, composite scores reflecting Physical and Social Functioning were computed. Scores can range from 0 to 100 with higher values reflecting better QoL. Physical Functioning was high (M = 80.5, SD = 13.7) and significantly higher (p < 0.001) than the normative value of 71 found among a large sample of similar patients. Similarly, Social Functioning was high (M = 79.1, SD = 16.9) and significantly higher than the normative value of 74 (p < 0.01). There were no significant differences between groups based on HPV status. Males reported significantly higher Physical (M = 82.1 vs. M = 72.9; p < 0.05) and Social Functioning (M = 81.9 vs. M = 67.8; p < 0.05) compared to females.

Cancer Specific Distress. Scores on the IES were computed for Intrusion and Avoidance subscales as well as the overall composite which was then used to categorize level of distress. Table 3. presents the categorical data. There were no significant differences between groups based on HPV status. Females reported significantly higher Intrusion (M = 11.9 vs. M = 6.0; p < 0.01) Avoidance (M = 13.1 vs. M = 6.6; p < 0.01) and overall cancer-specific distress (M = 25.0 vs. M = 12.6; p < 0.01) compared to males.

Table 3 Cancer specific distressResults for aim 3

Describe illness perceptions underlying cancer and non-cancer future events and determine whether these differ by HPV status and demographics.

The IPQ-R assesses the common sense model illness perceptions: Identity (symptoms associated with the disorder), timeline (the expected trajectory or cyclical nature of the disorder), consequences (the anticipated outcome of the disorder), and treatment and personal controllability (what can be done to control the threat posed by the disorder and by whom), illness coherence (an individual’s sense that their illness representation is coherent and useful), and emotional representation (negative affective reactions to the illness). Higher scores in the dimensions including identity, timeline, consequences, and emotional representation reflect more strongly held beliefs about the number of symptoms attributed to the illness, the chronicity of the condition, the negative consequences of the illness, and negative emotional reactions to the illness. High scores on the personal control, treatment control and coherence dimensions, represent positive beliefs about the controllability of the illness and a personal understanding of the condition. See Table 4. for the overall group descriptive statistics for the illness perception variables.

Table 4 CSM illness perceptions

With respect to HPV status, individuals with non-HPV related HNC reported significantly stronger beliefs in the chronicity of their condition than individuals with HPV related disease (M = 18.4 vs. M = 15.4, p < 0.05). No other differences were noted between HPV status groups. There were no significant differences between groups as a function of gender.

Results for aim 4

Assess the relationship between illness perception, QoL, distress, and HB among HNC survivors classified by HPV status, and determine if population stratification is necessary in intervention design. Unfortunately, the low prevalence of non-HPV related HNC and few females make stratification by these variables for the following analyses highly unreliable due to power considerations and risk for overfitting of models. Thus, only overall group relationships were examined and reported.

Table 5. presents the correlations between illness perception variables and HBs. As can be seen, illness perceptions appear remarkably unrelated to HBs among HNC survivors with the exception of Personal Control beliefs which are better dietary behaviors and greater leisure time activity.

Table 5 Correlations between illness perception variables and HBs

Table 6. presents the correlations between illness perception variables and QoL as assessed by the SF-36. The results demonstrate illness perceptions that relate significantly and moderately to many aspects of QoL. Identity, expected chronicity, negative consequences, and a negative emotional representation of HNC relate negatively to all aspects of QoL. The degree to which HNC is seen as cyclical relates negatively to all QoL domains except physical functioning, while the degree to which the illness makes sense to the individual relates positively to these domains. A sense of personal control over the disease and its treatment result(ed) in better QoL in all areas except physical functioning and limitations due to emotional function. Treatment control also was unrelated to limitations due to physical functioning.

Table 6 Correlations between illness perception variables and QoL as assessed by the SF-36

Table 7. shows the relationship between illness perceptions and HNC specific QoL on the UW-QoL measure. All illness perceptions were associated with HNC-specific QoL. Increasing beliefs in identity, chronicity and cyclical nature of the disease timeline, negative consequence, and negative emotional representation were related to worse physical and social QoL. A sense that one could exert control over the disease or its treatment and that one could make sense of HNC were related to better physical and social QoL.

Table 7 Correlations between illness perceptions and HNC specific QoL

Table 8 shows the correlations between illness perceptions and cancer-specific distress as assessed by the IES total score. Results are identical for the IES Intrusion and Avoidance subscales. Increasing beliefs in Identity, chronicity and cyclical nature of the disease timeline, negative consequence, and negative emotional representation were related to more cancer-specific distress. A sense that one could exert control over the disease and that one could make sense of HNC were related to decreased cancer-specific distress.

Table 8 Correlations between illness perceptions and cancer-specific distress as assessed by the IES total score

Table 9. presents the correlations between HBs and indices of general QoL, HNC-specific QoL, and Distress. Among HBs, only aerobic exercise activity was associated with QoL. In this case, greater aerobic activity was associated with better physical functioning, fewer physical limitations, greater vitality, better general health, and better cancer specific social QoL. No HBs were associated with cancer-specific distress.

Table 9 Correlations between HBs and indices of general QoL, HNC-specific QoL, and Distress

留言 (0)

沒有登入
gif