Adherence To Prophylactic Swallowing Exercises in Head and Neck Cancer Patients (PRESTO-TRIAL) - the ROLE of Personality Characteristics

Abstract

Prophylactic swallowing exercises (PSE) can mitigate (the consequences of) radiation associated dysphagia (RAD) in head and neck cancer (HNC) patients, provided patients adhere to a high-intensity program. However, patient adherence, a multifactorial concept influenced by several confounders, has been found to vary considerably between studies and individuals. PSE-related adherence confounders such as gender, smoking at baseline, baseline depression and high baseline fatigue levels have already been postulated. The influence of personality on adherence to PSE has not yet been investigated. Therefore, the aim of this secondary analysis of the Presto-trial is to examine the influence of personality traits on overall adherence rates to a PSE-program. One hundred fifty patients with a stage III or IVA-B (TNM7) newly diagnosed squamous cell carcinoma of the oropharynx -treated with (chemo)radiotherapy (C)RT - were included in the PRESTO-trial. One hundred sixteen out of the 150 participants completed the NEO Five Factor Inventory (NEO-FFI) personality questionnaire prior to starting RT. The NEO-FFI is a validated 60 item self-report questionnaire with 12 items per personality trait (neuroticism, extraversion, altruïsm, conscientiousness and openness). Statistical analysis was performed using simple and multiple regression models for the 5 personality traits. This study shows that the personality trait openness significantly influences overall adherence rates to PSE in HNC (p = 0.047). Higher scores on the personality trait openness increase overall adherence rates (β = 0.058, SE = 0.0290). Although, this singular statistical result has a very small effect size (OR 1.06), meaning that it is not likely a meaningful difference to consider any practice change. The personality traits extraversion, neuroticism, conscientiousness and altruism are not significantly related to overall adherence during PSE. This study is the first to explore the relationship between personality traits and adherence to PSE in HNC. While openness showed a statistically significant association with overall adherence, the effect size was very small and is unlikely to be of clinical relevance. No other personality traits demonstrated a significant relationship with adherence.

Introduction/Background

A severe complication following Head and Neck Cancer (HNC) concomitant (chemo) radiotherapy (C-(C)RT) is Radiation Associated Dysphagia (RAD). RAD is caused by muscle fibrosis, atrophy and by disuse following (C)RT and related toxicities [1, 2]. These swallowing disorders are considered to be the most important long-term adverse effect of radiotherapy for head and neck cancer and have a major impact on the patients’ quality of life (QoL) [3].

Since swallowing rehabilitation in RAD-patients is very challenging, experts worldwide acknowledge the importance of prevention of RAD [4–10]. Over the last few years multiple research studies investigated the effect of prophylactic swallowing exercises (PSE); i.e. swallowing therapy prior to and during concomitant CRT (CCRT). There is increasing evidence that PSE can reduce the severity of RAD [6–11]. Shown benefits include faster transition to oral diets and shorter duration of gastrostomy tube use [7, 8, 11–15], less structural deterioration in head and neck musculature [6], improved functional swallowing outcomes [6, 7, 9, 15, 16], and superior quality of life scores [10, 16].

However, a high level of exercise repetitions and committed adherence rates are essential to achieve benefits of PSE [8, 17]. It has been shown that adherence rates for PSE are generally low to high and vary between studies and individuals. In a study by Messing, Pisano et al., in which HNC patients underwent PSE during 7 weeks, adherence rates progressively decreased from 54 to 64% during the first 2 weeks, to 37% during week 3 and further to 17% in week 7 [18]. In a study by Shinn et al. only 13% of 109 patients classified as fully adherent [19].

Since intensive PSE puts an additional burden to the patient during an emotionally and physically demanding period, these low adherence rates do not come unexpectedly [7]. Considering the inherent capacity of PSE to have a major impact on the patients’ swallowing function and QoL, it is crucial to gain insight in factors that influence patients’ adherence to PSE programs. So far, research addressing this issue is limited.

Different possible adherence confounders can be distinguished. These confounders can globally be divided into 3 groups, PSE-related confounders (training intensity and service mode (tele practice, home practice, ….)), treatment-related confounders (e.g. CRT vs. RT) and patient-related confounders [5, 10, 19–23].

The presumption that patient characteristics have an impact on patient adherence to PSE has been supported by studies identifying gender (male subjects), smoking at baseline, baseline depression and high baseline fatigue levels as negative predictors of adherence [24]. A study by Shinn et al. that especially focused on underlying reasons for non-adherence to PSE stated that not having a swallowing problem at the time and a lack of understanding the need for exercises were additive factors for non-adherent behavior [19]. Other non – adherence reasons were finding exercises difficult, forgetting to do them, too busy, pain, nausea and fatigue [24].

A far less investigated domain is the influence of personality on adherence to (preventive) strengthening/physical exercises. Adherence to exercise protocols can be defined as the extent to which an individual’s behavior correlates with the agreed plan (repetition, intensity and frequency) of the proposed exercise intervention [25]. As personality comprises a number of broad dimensions that represent dispositional characteristics that reflect relatively enduring patterns of thoughts, feelings and behaviors, personality may directly have an influence on adherence [26].

There is accumulating evidence that personality may be linked to health outcomes in a variety of ways [27]. Personality may reflect differences in biological systems [28], it may reflect how individuals respond to illness (via health-seeking behavior, treatment adherence, social support) [29] and personality may be related to health-promoting (use of preventive healthcare) or health-damaging behavior (smoking, alcohol consumption, vaccination) [27, 30–32].

Research on the link between personality and adherence to curative health care have already been published. A great amount of research has already been carried out on the influence of personality on adherence to medication intake. In summary, low level of self-conscientiousness, and extraversion seems to have a negative impact on adherence to medication intake in different patient groups. While higher conscientiousness and agreeableness seems to be associated with better adherence to medication intake rates in different patient groups [32–46].

Far less studies were conducted on the effect of personality on adherence to exercise/physical programs. A study by Courneya et al. on adherence to exercises in cancer survivors revealed the personality trait extraversion as an independent predictor of adherence to exercise/physical programs. A study by Carsten et al. in university student population revealed positive associations for extraversion and conscientiousness on physical activity, while negative emotionality yielded inverse relationships with PA variables. Regarding adherence to physical activity/exercises, in summary, extraversion and conscientiousness seems to be a positive predictor [47–49].

Research on the link between personality and adherence to preventive healthcare on the other hand is rather limited. A longitudinal study by Nolan et al. on the link between personality and preventive healthcare in ageing people revealed that openness to experience is a significant predictor of breast lump check and mammogram in women aged 65+ [50]. Another study by Iwasa et al. also found that openness was the only personality trait associated with preventive health checkups in a sample of older Japanese individuals [51]. Padhi et al. revealed openness to be associated with use of preventive health care services [52].

A model often used to describe personality is the Big Five. The Big Five is a theory that describes personality in terms of a taxonomy of five general dimensions or clusters of personality traits: Openness, Conscientiousness, Extroversion, Agreeableness, and Neuroticism. These five traits represent broad domains of human behavior and account for differences in both personality and decision making [53].

1. Extraversion reflects a person’s propensity for sociability, assertiveness, high energy levels, seeking excitement and experiencing positive emotions.

2. Neuroticism reflects a person’s level of emotional stability. Typically defined as a tendency towards anxiety, depression, self-doubt and other negative feelings.

3. Conscientiousness exhibits traits such as organization, dutifulness, self-discipline, and a strong orientation towards performance.

4. Openness/Open-mindedness (to experience/intellect) refers to being perceptive, creative, reflective, valuing imagination and aesthetics, and heaving an appreciation for new experiences.

5. Altruism/Agreeableness as a personality trait includes attributes such as trust, altruism, kindness, affection and other prosocial behaviors.

To the best of our knowledge the relationship between personality and adherence to PSE in HNC has not yet been investigated.

The presto-Trial, a previously published, multicentric, randomized-controlled trial investigating the effect of adherence to PSE during HNC radiotherapy revealed important insights to improve dysphagia care in head and neck cancer patients treated with radiotherapy [54]. The primary results showed that patients were more likely to adhere to preventive swallowing exercises when they exercised face-to-face with the therapist. Furthermore, patients who completed more than 75% of the exercise protocol demonstrated significantly greater improvements in terms of tongue strength and swallowing function. Primary results did not focus on analyzing factors that may influence adherence rates were not analyzed as primary results. Therefore, this study - a secondary analysis of the Presto-trial, - aims to investigate the effect of specific personality traits on adherence to a high-intensity PSE-program, regardless of treatment arm.

Methods

Study Population

One hundred fifty patients with a stage III or IVA-B (TNM7) newly diagnosed squamous cell carcinoma of the oropharynx, treated with (chemo)radiotherapy with or without induction chemotherapy, were included. Patients needed to demonstrate sufficient cognitive and language abilities. This assessment was made based on the initial counseling session with the aim that patients had to have sufficient cognitive and linguistic abilities to understand and perform the protocol correctly. The presence of a recurrent carcinoma or metastasis from another carcinoma and previous CRT or surgery in the head-neck region with possible impact on swallowing function, were exclusion criteria [54].

Study Design

This study is a secondary analysis of a multicentric RCT study: the prophylactic swallowing exercises in head-and-neck cancer patients treated with (chemo)radiotherapy trial - (PRESTO trial).

Participants

All participants were randomly assigned to one of the three exercise groups, namely: the control group (group 1), app group (group 2) or therapist group (group 3). Group 1 performed the exercises at home without supervision of a speech and language pathologist (SLP). Group 2 practiced at home with instructions by videos via an application on a tablet. Both groups had one counseling session per week by a SLP - focusing on remeasuring patients. Group 3 was given daily face-to-face therapy by a SLP and also had a remeasurement of tongue strength and chin tuck strength once a week during one of the 5 sessions. All participants of all groups trained five times a week only during the first 4 weeks of CRT (Fig. 1). Restricting the duration to the first four weeks of CRT is a method to increase compliance since literature shows that the last 2–3 weeks of therapy the feasibility of completing exercises decreases.

Fig. 1.

flow chart of the exercise program and study design

Exercise Program

The prophylactic swallowing exercise program was the same in all groups and comprised evidence-based exercises targeting the main muscle groups involved in swallowing: muscles involved in tongue strength, pharyngeal contraction and laryngeal elevation/upper oesophageal sphincter opening. These strengthening exercises included tongue-strengthening exercises (TSE) and chin-tuck against resistance (CTAR) combined with an effortful swallow, alternating every day. Tongue-strengthening exercises consisted of 120 tongue presses every exercise day and were divided into 12 sets of 10 repetitions with the target level set at 80% of the maximum amount of pressure that can be generated in one repetition. Each CTAR session consisted of 150 chin-tucks against resistance at a target level of 60–70% of the maximum chin-tuck strength that could be generated. These chin-tucks were divided into 30 sets of five repetitions with an effortful swallow at every fifth repetition. Target levels for both exercises were based on feedback from the devices used for maximum pressure measurements.

Outcome Measure

Adherence

For the primary outcomes of the PRESTO-study, the degree of adherence was expressed as the percentage of completed repetitions per week (%reps) and by means of 4 different categories, defined by Wall et al. (2017), including negligible practice (< 25% reps), low practice (25–50% reps), moderate practice (50–75% reps) and high practice (> 75% reps) [17, 24]. As the primary results showed that patients with adherence rates of 75% or more performed significantly better in terms of swallowing function and tongue strength, this secondary analysis was conducted based on two groups: those with adherence rates of 75% or more, and those with adherence rates of less than 75%.

Personality Characteristics

To examine the BIG 5 personality traits of the patient the NEO Five Factor Inventory (NEO-FFI) was used [55]. The NEO-FFI is a well-validated 60-item self-report questionnaire with 12 items measuring each of the five personality domains. The NEO FFI questionnaire was completed by patients during the first (baseline) counseling session. Completing this questionnaire takes approximately 10 min. One hundred sixteen out of the 150 patients were willing to complete the personality questionnaire at baseline.

Patients responded to all statements of the tool by selecting one of the five response-options (totally disagree, disagree, neutral, agree or totally disagree) according to their agreement with the statement. The researcher will assign domain scores to all 5 personality domains by totaling the scores the respondent assigned to the items in the questionnaire (= absolute scores). Because each item is scored on a 0–4 scale, absolute scores for each personality domain range from 0 to 48.

Afterwards, the absolute scores were converted into stanine scores based on the NEO-FFI normative tables. This gives us a profile that defines, per personality domain, the extent to which that domain applies to a particular person. A stanine score ranges from very low (1) to very high (9) with standard interpretive categories: low scores (1–3), average scores (4–6), and high scores (7–9).

Statistical Analytical Framework

All reported results of this secondary analysis are based on analyses with adherence in two groups: overall adherence > 75% and overall adherence < 75%.

Binary logistic regression model with outcome variable adherence group (overall adherence > 75% and overall adherence < 75%) and dependent variables personality traits and therapy were fitted. First, simple models were fitted with one explanatory variable per model. The final multiple regression mode, controlled for therapy group was built using the step-up selection procedure based on p-values of model coefficients and Akaike Information Criterion (AICs).

Descriptives (Numbers, Means, lower and upper Quartiles (Q1, Q3)) of distribution were analyzed for all personality traits.

Results

Demographics

Demographic and medical information for the sample is detailed in Table 3. T classification, N classification and HPV information was missing data for 1 patient. The mean age of participants was 63,4 (SD = 8,49). 75% of the patients were men. Given the fact that head and neck cancer is more common in males than in females (with a male to female ratio of 2.5), this skewed male-female distribution within our patient group is not surprising [56]. Medical information indicates that 84% of the patients received CRT. 58 of the patients (50%) were HPV positive. Regarding the T classification 55 of the patients had a T1 or T2, while for the N classification only 5 patients had a N0, 16 patients had a N2 and 94 patients had a N2-3.

Table 3.

patients, disease and treatment characteristics

		N	%
Gender	Male	87	75
	Female	29	25
T classification	1–2	55	48
	3–4	60	52
N classification	0	5	4
	1	16	14
	2–3	94	82
Treatment	RT	18	16
	CRT	98	84
HPV status	Positive	58	50
	Negative	57	50
Overall adherence to PSE	< 75%	60	52
	> 75%	56	48

N number, M mean, SD standard deviation, RT radiotherapy, CRT chemoradiotherapy

Simple Models

Regarding the simple models no significant results were obtained for personality traits. Significant results were only observed for therapy group, as descripted in the article about the primary results by Margot Baudelet et al. [17]

Main results – multiple Regression Model

The results of the final, multiple regression model, controlling for therapy group, demonstrates a significant effect of the personality trait openness on adherence (Table 1). A higher score on this personality trait leads to better adherence to PSE. No significant relation between extraversion, neuroticism, conscientiousness, agreeableness and adherence was found.

Table 1.

Binary logistic regression models for adherence. Left panel: simple logistic models with one dependent variable per model. Right panel: final multiple model obtained after Stepwise selection and controlling for group

	Simple models				Final model
	β (SE)	p	OR	95%CI	β (SE)	p	OR	95%CI
Extraversion	−0.003 (0.026)	0.91	1.00	0.946–1.050	-	-
Neuroticism	−0.015 (0.024)	0.52	0.985	0.939–1.032	-	-
Conscientious-ness	−0.034 (0.027)	0.21	0.967	0.917–1.020	-	-
Openness	0.032 (0.026)	0.23	1.033	0.980–1.088	0.058 (0.0290)	0.047	1.06	1.00–1.12
Agreeableness	0.002 (0.032)	0.95	1.002	0.941–1.067	-	-
GROUP (ref = therapist group)		< 0.001				< 0.001
paper	−1.43 (0.49)	0.003	0.24	0.093–0.62	−1.46 (0.49)	0.003	0.23	0.089–0.60
app	−2.22 (0.49)	< 0.001	0.11	0.042–0.29	−2.40 (0.51)	< 0.001	0.091	0.034–0.25

β = standardized regression coefficients, with standard error (SE), OR = Odds Ratio, representing the ODDS for higher order outcome category (OR > 1 = higher odds for higher outcome category (= adherence)

Descriptive Statistics of all Personality Traits

Table 2 representing the global tendencies for all personality traits when looking at overall adherence in two groups (overall adherence < 75% and overall adherence > 75%).

Table 2.

Descriptive statistics (number (N), mean (M), the lower quartile (Q1) and upper quartile (Q3)) of distribution for all personality traits. Adherence in two groups (overall adherence < 75% and OA > 75%)

	Extraversion		Neuroticism		Conscientiousness		Openness		Agreeableness
	OA < 75%	OA > 75%	OA < 75%	OA > 75%	OA < 75%	OA > 75%	OA < 75%	OA > 75%	OA < 75%	OA > 75%
Number Participants	60	56	60	56	60	56	60	56	60	56
Mean NEO-FFI score	40,75	40,80	30,63	29,98	47,48	46,02	34,48	35,52	44,38	44,57
Q1	36,25	36,00	27,00	24,75	42,25	41,25	30,00	31,00	41,00	40,25
Q3	45,00	45,00	35,00	35,00	51,75	50,00	38,75	39,00	48,75	49,00
SD	6,14	6,68	6,63	8,03	6,56	6,50	6,59	6,49	5,43	5,24

Discussion

In order to benefit from the positive effects of PSE, high adherence rates are very important [17]. Consequently, in order to keep the adherence of patients to PSE high, it is very important to get a clear view of possible factors influencing adherence to PSE.

This study is the first study to examine the relationship between patients personality and adherence to preventive swallowing exercises in the HNC population. In the current study, no clear relation between personality and adherence rate could be found. Only openness influenced adherence significantly, with a limited odds ratio of 1.06. The higher the score on the personality trait openness the better the compliance rates. None of the other investigated personality traits had a statistically significant impact on adherence rates. This finding is remarkable since in literature regarding studies focusing on adherence to exercise programs most frequently personality traits extraversion and consciousness are positively correlated with higher adherence rates [47–49].

A possible explanation for the small but statistically significant correlation between openness and adherence may be patients coping strategies. Coping is often defined as efforts to prevent or diminish threat, harm and loss, or reduce associated distress [57]. Personality does influence coping in many ways, however, some of which occur prior to coping, such as the frequency of exposure to stressors, the type of stressors experienced and appraisals [58].

Openness, extraversion and conscientiousness all relate to perceiving events as challenges rather than threats and to positive appraisals of coping resources [59]. Literature analyzing the relationship between personality and coping mechanisms demonstrate that the personality trait neuroticism predicts exposure to interpersonal stress, and tendencies to appraise events as highly threatening and low coping resources as low [60–63]. Whereas conscientiousness and agreeableness predict low stress exposure [58, 64, 65]. Meta analyses link optimism, extraversion, agreeableness, conscientiousness and openness to more engagement coping, while neuroticism is often linked to disengagement coping [57]. Engagement coping includes problem focused coping and some forms of emotion focused coping: support seeking, emotional regulation, acceptance and cognitive restructuring. Disengagement coping includes responses such as avoidance, denial and wishful thinking [57]. As such we hypothesize that persons with higher openness have a more positive mindset and feel more in control and able to do something themselves. That could explain the higher adherence rates among individuals with higher scores on the personality trait openness.

Although no significant associations were found for extraversion, neuroticism, conscientiousness and agreeableness, the descriptive statistics of these personality traits (Table 2) suggest patterns that resemble tendencies described in previous studies. For neuroticism and agreeableness, the distribution of scores across adherence groups appears consistent with observations reported in research on adherence to medication intake [27–41]. No comparable patterns were observed for extraversion and the opposite tendence was observed for the personality trait conscientiousness. While in other studies higher scores on conscientiousness were related with higher adherence rates [28, 36–40], the inversed relationship can be visually distinguished in Table 2. These observations are purely descriptive and should be interpreted with caution, as no statistically significant effects were detected.

A study by Courneya et al. (2002) on adherence to exercises in cancer survivors revealed the personality trait extraversion as an independent predictor of adherence in the exercise condition [47]. Another study of Courneya et al. (2004) within the population of colorectal cancer identified only non-personality predictors of adherence to exercises [48]. Baseline exercise stage of change, employment status, treatment protocol and perceived behavioral control were identified as predictors. Likewise, our study within the HNC cancer population demonstrates that personality has a limited impact on adherence rates to PSE in head and neck cancer.

Consequently, the need for additional support to adherence to PSE cannot be predicted by the NEO FFI-scores for personality traits extraversion, neuroticism, conscientiousness, openness and agreeableness. Overall, there are limited differences in personality traits between the 2 adherence groups overall adherence 75- and overall adherence 75+. Therefore, measuring personality traits has little clinical impact.

Studies that focus on cancer population groups and examined exercise adherence confounders to physical activity’s do show that there are possible other confounders to predict adherence. The study by Courneya et al. on adherence to exercises in cancer survivors revealed three independent predictors of adherence to exercise: sex, normative beliefs and perceived behavioral control (PBC). PBC is defined as the perceived ease or difficulty of performing behavior [47]. Research of Buffart et al. (2018) investigated demographic, clinical, lifestyle-related, and social-cognitive correlates of physical activity in an HNC population showed that there are other non-personality confounders to predict adherence. Results show that physical activity intention was significantly higher in HNC survivors with a history of exercising, who had a more positive attitude, subjective norm, and perceived behavioral control. Patients with higher physical activity intention, higher perceived behavioral control, a lower age, and without unintentional weight loss or comorbidities had higher physical activity behavior [66].

For our data collection it would also be worthwhile to explore what other possible confounders (social network, personal health state) affects adherence rates. Providing better knowledge of possible confounders, would allow better estimation of the necessary counselling and encouragement in the future when patients are enrolled in exercise protocols. Accordingly, more contact hours and guidance from a speech and language therapist could improve adherence of patients at risk for low adherence behavior. In that way, also patients at risk for low adherence would benefit from the positive effects of PSE.

This research and similar future research will help establish an empirical basis for improving adherence to exercise training protocols.

Limitations

There are some limitations that need to be considered when interpreting the findings of this study. First of all, due to the fact that less open-minded people may be less attracted to participate in scientific research, less open-minded patients may not adequately be represented in our patient population. Secondary, the total number of patients that were questioned to participate in the bigger PRESTO-study was not recorded. The percentage of patients that refused participation is therefore unknown. Of the 150 patients that participated in the PRESTO-study only 116 were willing to complete the personality questionnaire at baseline.

Conclusion

Personality traits that influence overall adherence to PSE in HNC have been identified in this study for the first time. Results demonstrate that higher scores on the personality trait openness were significantly associated with greater overall adherence rates. However, it should be mentioned that this significant result is not clinically relevant given the effect size.

Data Availability

The datasets generated during the current study are not publicly available since they contain patient data and the Informed Consent does not include sharing data publicly. They are available from the corresponding author upon reasonable request. All clinical record forms are collected and managed using REDCap (Research Electronic Data Capture).