Towards more personalized digital health interventions: impact of emotions and physical complaints on goals, action and coping plans to promote physical activity

doi:10.21203/rs.3.rs-2459495/v1

Download PDF

Research Article

Towards more personalized digital health interventions: impact of emotions and physical complaints on goals, action and coping plans to promote physical activity

https://doi.org/10.21203/rs.3.rs-2459495/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Background.

Despite effectiveness of action and coping planning in digital health interventions to promote physical activity (PA), attrition rates remain high. Indeed, support to make plans is often abstract and generic. Nevertheless, people are different, and context varies. Tailored support involving personalized suggestions of specific plans, may be one way to help. As a first step, we designed a study to identify whether dynamic user information (i.e. emotions, physical complaints) relates toward goals, specific action plans (APs) and coping plans (CPs).

Methods.

Each morning for eight consecutive days, students (n = 362) were instructed to create APs and CPs to reach their PA goal and answered questions regarding emotions and physical complaints. To examine associations between emotions, physical complaints and content of the created goals, APs and CPs, generalized linear mixed effect models were fitted using the lme4-package in R.

Results.

Between- and within-individual variations in emotions and physical complaints influenced the content of goals. When experiencing more positive emotions, a higher goal in terms of minutes PA was set (within: P < .001, between: P = .023). If participants had more negative emotions, were more fatigued and experienced more pain, a lower goal in terms of minutes PA was set (within: P < .001; within: P < .001; within: P = .025 respectively). Between- and within-individual variations in positive emotions, fatigue and pain also influenced the content of APs and barriers of CPs. When experiencing more positive emotions, activities outside (within: P = .016, between: P = .011), not at home (within: P = .013, between: P = .007) and with someone else were preferred (within: P < .001, between: P < .001), whereas if participants were more fatigued, they preferred doing activities alone (within: P = .006), and if participants experienced more pain, they preferred light-to-moderate instead of high intensity activities (within: P = .024). Lastly, when participants had more positive emotions, they reported more external barriers (within: P = .004, between: P = .003) whereas when participants were more fatigued and experienced more pain, they reported more internal barriers (within: P = .010, between: P < .001; within: P < .001 respectively).

Conclusions.

This study reveals associations between emotions, physical complaints and the content of goals, APs and CPs participants set. Consequently, the knowledge acquired from these findings might be used to ultimately personalize suggestions of goals, APs and CPs in a more dynamic manner. However, further validation and refinement with experts and experimental designs are needed.

Digital health

e-health

m-health

behaviour change

physical activity

personalization

emotions

physical complaints

Regularly engaging in physical activity (PA) is key to prevent non-communicable diseases such as type 2 diabetes, cardiovascular disease, cancer and obesity (Bassuk & Manson 2005; Haskell et al. 2009; Warburton et al. 2006). Notwithstanding, 23% of adults and 81% of adolescents do not meet the WHO global recommendations on PA (WHO 2018b). Consequently, lifestyle interventions that can promote PA are required. Digital health interventions, an umbrella term for the usage of digital technology to support health (WHO 2018a), are promising because they can reach large groups at a relatively low cost (Broekhuizen et al. 2018; Vandelanotte et al. 2016).

Theories of behavior change (i.e. social cognitive theory, the health belief model, self-determination theory, …) have already proven to be essential to understand PA behaviour (Rhodes et al. 2019) and provide useful frameworks for the development of effective digital health interventions (Webb et al. 2010). An all-encompassing framework is the Health Action Process Approach (HAPA) because it guides individuals through two phases of behaviour change: a motivational phase (pre-intentional), leading to a behavioural intention, and a volitional phase (post-intentional), bridging the gap between intention and the actual behavior (Schwarzer et al. 2011; Schwarzer & Luszczynska 2008). Research already showed that behaviour change techniques (BCTs) such as goal setting (i.e. setting a relevant goal for PA), action planning (i.e. planning what you want to do, where, when and with whom you want to do it) and coping planning (i.e. exploring possible obstacles and think about solutions) are key components in bridging the intention-behaviour gap (Sniehotta et al. 2005; Sniehotta et al. 2006; Sutton 2008). Previous digital health interventions using these BCTs to promote PA such as MyPlan (i.e. weekly cycles of setting goals, action and coping plans) and MyDayPlan (i.e. daily cycles of setting goals, action and coping plans) already proved to be effective (Degroote et al. 2021; Poppe et al. 2019; Schroé et al. 2020). Despite effectiveness, attrition rates in these interventions remained high (Schroé et al. 2022b; Van der Mispel et al. 2017). Overall, digital health interventions indeed suffer from high attrition rates because stopping the intervention is only one click away (Eysenbach 2005; Geraghty et al. 2013). However, another important reason might be due to the fact that support to guide participants through the behavioural processes of goal setting, action and coping planning were abstract, generic and not tailored to the individual in MyPlan and MyDayPlan (Schroé et al. 2022b). In an attempt to tackle this, participants were considered as their own expert in terms of setting goals and making plans: they were instructed to specify the content themselves, allowing to take their own personal situation into account. Notwithstanding, this resulted in a low quality of plans and participants also experienced difficulties in formulating them (Degroote et al. 2020; Poppe et al. 2018). As such, there is a need to tailor this support in a more personalized and dynamic way (Schroé et al. 2022a).

Several systematic reviews already demonstrated that tailoring intervention content to participants’ profiles in digital health interventions positively affected behaviour change (Davis et al. 2020) and health outcomes (Lustria et al. 2013) compared to non-tailored interventions. However, tailoring in most digital health interventions remains largely static. Intervention content is often only tailored to participants’ baseline levels of PA, gender or motivational stage (i.e. having an intention to change or not). Notwithstanding, being physically active demands daily effort and sustained dedication, and might be influenced by more dynamic (i.e. time-dependent) variables that vary from day to day and even throughout the day (Dunton et al. 2010). Indeed, research already showed that PA is influenced by how we feel in the moment and how we interact with our environment (Dunton 2017; Poppe et al. 2021). For example, several studies (Dunton et al. 2010; Schultchen et al. 2019; Vetrovsky et al. 2021) demonstrated that emotions and physical complaints (e.g. fatigue and pain) influence levels of PA. Vetrovsky et al. 2021 (Vetrovsky et al. 2021) showed that morning fatigue in older adults was associated with less daily moderate to vigorous physical activity (MVPA). Also, university students who perceived more stress, more negative emotions and less positive emotions had lower subsequent levels of PA (Schultchen et al. 2019). By understanding how these dynamic variables influence PA behavior, it is possible to determine moments of opportunity or vulnerability to do PA across or even within days, which is important to tailor future digital health interventions (Conroy et al. 2019; Dunton 2017).

To support the development of more dynamically tailored interventions, it is not only of importance to understand how these dynamic variables influence PA levels in itself but also how these dynamic variables influence behavioural processes such as goal setting, action planning and coping planning. According to the HAPA model, situational barriers and opportunities indeed influence behavioural processes (Schwarzer 2008) (e.g. having a social network with friends who supports an individual in being more physically active might influence an individual’s PA goal, action plan and coping plan). Yet, dynamic variables such as emotions and physical complaints can also be seen as situational barriers and opportunities, however investigating whether these dynamic variables influence behavioural processes, has never come to the fore. These insights might contribute to more dynamically tailored support in future digital health interventions (e.g. another type of action plan might be suggested to an individual when they feel more fatigued as compared to other days).

Therefore, the aim of the current study was to investigate whether variations in emotions and physical complaints influence behavioural processes such as goal setting, action and coping planning. More specifically, we assessed whether within-individual and between-individual variations in daily emotions and physical complaints influenced the content of goals, action and coping plans on a daily basis. The assessment of day-to-day variations were chosen to allow insights for dynamically tailored support of goals, action and coping plans on a daily basis.

Participants

Participants in this study were 3rd year bachelor students who studied an academic Master in Clinical Psychology at Ghent University, Belgium. Participants took part in this study by completing an online task ‘MyActionPlan’, which is part of the course ‘Models in Health Psychology’. In order to use participants’ data for research purposes, participants gave informed consent at the start of the online task. If they did not consent, they were asked to complete the online task ‘MyActionPlan’ for educational purposes, but their data was not used for research purposes. The study was conducted from October 2020 till November 2020 and was approved by the Ethical Committee from the Faculty of Psychology and Educational Sciences at Ghent University (registration number: 2020/87). It has to be noted that the study took place during strict corona measures due to the Covid-19 pandemic (i.e. maximizing online education, limitation on close contacts was 4 persons, prohibition on indoor sports, prohibition on outdoor sports with more than 4 people, closing of the fitness centra and swimming pools).

Procedure

The online task of ‘MyActionPlan’ was created with the online survey software LimeSurvey (LimeSurvey Project, Hamburg, Germany). The procedure of the task can be found in Fig. 1. Participants received an e-mail with a link to the online task and received a unique token to log in. First, participants were instructed to fill in an intake survey (day 0). Thereafter, for eight consecutive days (day 1–8), they had to fill in a morning survey, which could only be accessed from 7 AM till 11 AM, and an evening survey, which could only be accessed from 8 PM till 12 PM. The evening questionnaire was not used for analysis in this study as it only evaluated the achievement of the goal in general and not whether they actually performed the action and coping plans. The task and all questionnaires are stored in OSF (Schroé 2022).

Measurements

Demographic information

Participants age, sex, height and weight were assessed in the intake survey.

Goal setting, action and coping planning

In the morning surveys, participants were guided through the process of goal setting, action planning and coping planning. Participants were instructed to set a relevant goal regarding PA in minutes per day (e.g. 30 minutes of PA per day) or in steps per day (e.g. 10 000 steps per day). This question was accompanied with background information regarding the PA guidelines (WHO 2018b). In order to achieve their goal, participants were instructed to create one or several action and coping plans. The following open questions along with a standard example were asked to guide participants through the process of action planning: “What are you going to do today to achieve your goal?” (standard example: “I will go running”), “When will you do this?” (standard example: “When coming home after work”), “Where will you do this?” (standard example: “In the wood near my home”), “With whom will you do this or will you do the activity alone?” (standard example: “With my partner”). Thereafter, the following open questions along with a standard example were asked to guide participants through the process of coping planning: “What could be a barrier for you to carry out your action plan?” (standard example: “I might forget about my plan”) and “What could be a possible solution for that barrier?” (standard example: “I will set a reminder in my agenda”). Participants were provided the opportunity to create up to three action plans, and up to two coping plans per action plan.

Emotions and physical complaints

In the morning survey also emotions and physical complaints were assessed. The items for emotions were selected from items that are frequently used for the Experience Sampling Method (ESM) (Delespaul 1995). Four positive emotions (i.e. cheerfulness, relaxation, enthusiasm, satisfaction) and four negative emotions (i.e. insecurity, anxiousness, irritation, feeling down) were assessed. An example of an item is: “How relaxed are you at this moment?”. All emotions were rated on a 7-point response scale (e.g. 1 = not at all relaxed to 7 = completely relaxed). The mean scores for both the positive and negative emotions were used in the analyses as they showed good internal consistency (α = 0.89 and α = 0.81 respectively). The items for the physical complaints were based on the Patient Health Questionnaire-15 (Kroenke et al. 2002).Two physical complaints were assessed: fatigue by the item “How fatigued are you at this moment?” (1 = not fatigued at all to 7 = extremely fatigued) and pain with the item “How much pain do you feel at this moment” (1 = no pain at all to 7 = strenuous pain).

Data-analysis

Coding and inter-rater reliability

In preparation for analysis, answers to the open questions of the action and coping plans were coded into different variables. A research team of six researchers (the researchers HS, MB and four master thesis students) coded all action and coping plans of participants following a codebook. The codebook was developed by the researchers HS and MB and was made more accurate during three rounds of discussions with all researchers. During each round, +- 50 action plans and coping plans were coded by all researchers, the coded plans were then compared with indication of the main differences and difficulties, and afterwards a discussion with all researchers took place where adjustments to the codebook were made accordingly. Finally, the action plans were coded into the following variables; ‘Type of activity’ (coding was based on the Physical Activity Concept Ontology (PACO) (Kim et al. 2019), e.g. walking was coded as 420), ‘Moment of the activity’ (e.g. 3 = doing activity in the afternoon), ‘Location of the activity’ (e.g. 7 = doing activity in the nature), ‘With whom doing the activity’ (e.g. 1 = doing activity alone), ‘PA domain’ (coding was based on the 4 domains in which one can perform PA (Strath et al. 2013), e.g. 3 = doing PA at work). The coping plans were coded into the following variables; ‘Type of barrier’ (e.g. 2 = weather) and ‘Type of solution’ (coding was based on the Self-Enactable Techniques of Knittle et al (Knittle et al. 2020); e.g. when someone prepared his/her equipment in advance to not forget the plan, this was coded as 10 = adding objects to the environment). For the coping plans, the variable ‘Type of solution’ was not used for analysis in this study because of two reasons. First, for the analyses all dependent variables of interest were dichotomized (see statistics). We judged that there was no logical categorization for the Self-Enactable Techniques. Second, solutions are already likely to be associated with the barriers in itself (Schroé et al. 2022a), so we decided after careful internal consideration that investigating the association between emotions/physical complaints and solutions was less important. The quality of the action plans and barriers was also coded in terms of instrumentality (e.g. 1 = plan is instrumental). Instrumentality of an action plan reflects whether the plan was suitable to achieve the intended PA goal (Reinwand et al. 2016). Similarly, instrumentality of the barrier reflects whether the barrier was suitable for the action plan.

Each master thesis student coded a part of the dataset in Microsoft Excel. A sample of 13.7% (n = 396 of 2896 days) of the action plans and barriers was independently double-coded. The intraclass correlation coefficient (ICC) statistic was used to quantify the level of consistency between the coding of each variable of the action plans and barriers. ICCs were computed using IBM SPSS Version 27, using a two-way mixed effect model (Koo & Li 2016). Cicchetti’s ICC numerical ranges and descriptors were used for inter-rater reliability: “excellent” (ICC ≥ 0.75), “good” (ICC = 0.60– 0.74), “fair” (ICC = 0.40–0.59) and “poor” (ICC < 0.40) (Cicchetti 2001). Overall, the ICC of all variables can be considered to be at least fair (Table 1).

Table 1

Inter-rater reliability for all variables of the action plans and barriers of the coping plans.
Variable Description		ICC (with Cicchetti’s descriptors)
Action plans	Type of activity	0.984 [0.982–0.987] (excellent)
	Moment of the activity	0.960 [0.953–0.967] (excellent)
	Location of the activity	0.946 [0.936–0.954] (excellent)
	With whom doing the activity	0.987 [0.984–0.989] (excellent)
	PA domain	0.840 [0.810–0.864] (excellent)
	Instrumentality action plan	0.777 [0.737–0.811] (excellent)
Coping plans	Barrier	0.910 [0.896–0.923] (excellent)
Coping plans	Instrumentality barrier	0.471 [0.386–0.544] (fair)
ICC = intraclass correlation coefficient

Statistical analysis

Analyses were conducted in R version 4.0.1 (Team 2020). First, descriptive statistics were calculated for the total sample of participants, for the content of goals, the action plans and barriers, and for the emotions and physical complaints. Only action plans and barriers that were evaluated as instrumental, were used for analysis.

Second, to examine the variability of the emotions and physical complaints, generalized linear mixed effect models were implemented using the lme4-package (Bates et al. 2014). These models can take the hierarchical structure of the data into account (i.e. within participants, within days). Levels of the between subject variance (i.e. differences between different subjects) and within subject variance (i.e. differences within the same subject, but between different days) of the positive and negative emotions and physical complaints were calculated by running intercept-only models (i.e. models only including a fixed and random intercept) with the emotions and physical complaints as outcome variables.

Third, to examine the effect of the emotions and physical complaints on the content of the goals, action plans and barriers, generalized linear mixed effect models were fitted that allowed separate within-and between subject effects. To obtain within-subject effects, each independent variable (positive emotions, negative emotions, fatigue and pain) was centered within subject (i.e. the individual’s daily score was subtracted from the individual’s mean score). As such, dependencies between the intercept and the average levels of predictors within individuals were avoided (Enders & Tofighi 2007). The within-subject predictors yield an index of daily change. The between-subject effects reflect the individual’s mean level of the independent variable and has the same value for all measurements of one subject.

For each dependent variable, a separate model was fitted. For the continuous variables (i.e. content of goals in number of minutes PA and content of goals in number of steps), the normality assumption was checked visually by inspecting the residuals versus fitted values plot and the quantile-quantile plot. If normality could not be assumed, the Bayesian Information Criterion (BIC) of models with different variance and link functions (i.e. Gamma with log, Gamma with identity, Poisson with log, and negative binomial with log) were compared and the model with the lowest BIC value was selected (Germeys & De Gieter 2018; Poppe et al. 2021). The content of the goals (i.e. choice between type of goal in steps versus minutes PA), action plans and barriers were categorized into dummy variables before they were entered in the models. ‘Type of activity’ was categorized into ‘light to moderate intensity PA’ (0) versus ‘high intensity PA’ (1), which was based on the MET-guidelines (Ainsworth et al. 2000), and ‘sports or leisure activity’ (0) versus ‘daily living activity’ (1), which was based on the PA domains (Strath et al. 2013). ‘Moment of the activity’ was categorized in ‘doing the activity during the day’ (0) versus ‘doing the activity in the evening’ (1). ‘Location of the activity’ was categorized into ‘home’ (0) versus ‘not home’ (1), and ‘inside’ (0) versus ‘outside’ (1). ‘With whom doing the activity’ was categorized into ‘alone’ (0) versus ‘with someone else’ (1). Categorization of ‘moment of the activity’, ‘location of the activity’, ‘with whom doing the activity’ are in line with categories of the PACO ontology (Kim et al. 2019). ‘Type of barrier’ was categorized into ‘external barrier’ (0) versus ‘internal barrier’ (1), which is in line with internal and external locus of control (Rotter & Mulry 1965). A Gamma model with log link function was used for the content of goals participants set in number of minutes PA. In order to run the models for the number of steps, the number of steps were divided by 1000 to avoid very large values. Notwithstanding, the Poisson with log model and the negative binomial with log model failed to convert. Consequently, a Gamma model with an identity link function was used for the content of goals participants set in number of steps. For the type of goal participants chose (steps versus minutes PA), a binomial model with logit function was used. The same model was also used for action plans and barriers. P values < .05 were considered statistically significant. Beta’s for the continuous variables, ODDs ratios (OR) for the dummy variables and 95% confidence intervals were also reported.

Descriptive statistics of the study sample

In total, 362 students agreed to participate in the study. All participants responded to all questionnaires on all days (intake survey, eight morning surveys, eight evening surveys). The mean age of the participants was 21 years (SD = 3.36), 88.7% (321/362) of participants were female, and the mean Body Mass Index in kg/m² was 22.05 (SD = 3.31). Tables 2, 3, and 4 present the descriptive characteristics of the content of goals, action plans and barriers, respectively. For the goals, a total of 2894 goals were set by participants, of which 2827 (97.7%) were relevant goals in terms of steps or minutes per day and were thus used in further analyses. For the action plans, a total of 4063 action plans were created by participants, of which 3855 action plans (94.8%) were evaluated as instrumental. For the barriers, a total of 5056 barriers were created by participants, of which 4946 barriers (97.9%) were evaluated as instrumental. As can be seen in Tables 3 and 4, the total number per variable of the action plans and barriers is different (e.g. different total number for ‘Type of activity’, ‘Moment of Activity’), which might be attributed to two possibilities. First, during the coding process itself, some entries were overlooked due to the large amount of data and were consequently not encoded by the researchers. Second, during the categorization process, some data was not included in the categories (e.g. category ‘not specified’ for the variable ‘Moment of the activity’ was not included in the dummy-variable ‘during the day’ vs ‘in the evening’).

Table 2

Characteristics of the goals participants’ set.
Goals (n = 2827)
Characteristics	N (%)	Mean	SD	Range
Goal chosen in steps per day	486 (17.2)	8971	2403	3000-20 000
Goal chosen in minutes PA per day	2341 (82.8)	52	34	10–480

Table 3

Characteristics of the action plans.
Instrumental action plans (n_total= 3855; n_ap1=2736, n_ap2=948, n_ap3=171)
Type of activity	n	%	Moment of the activity	n	%	Location of the activity	n	%	With whom doing the activity	n	%
Walking	1656	43.4	Not specified	957	25.4	Home	1358	35.7	Alone	2281	60.2
Work-out	837	21.9	Afternoon	852	22.6	Nature	848	22.3	Same generation	906	23.9
Running	379	9.9	Evening	710	18.8	Neighborhood	698	18.4	Different generation	268	7.1
Cycling	337	8.8	Noon	674	17.9	On the road	379	10	Group	229	6
Household activities	264	6.9	Before noon	575	15.3	City	248	6,5	Pet	104	2.7
Other	154	4	Total	3768	100	Garden	105	2.8	Total	3788	100
Dance/ gymnastics/ pole dance/martial art	86	2.3	Moment in 2 categories*			Sport facilities (outside)	41	1.1	With whom in 2 categories*
Stair climbing	60	1.6	During the day	2101	74.7	Work	39	1	Alone	2385	63
Ball sports	31	0.8	Evening	710	25.3	Not specified	34	0.9	Not alone	1403	37
Racket sports	11	0.3	Total	2811	100	Other	30	0.8	Total	3788	100
Swimming	4	0.1				Sport facilities (inside)	23	0.6
Total	3819	100				Total	3803	100
Type of activity in 2 categories (intensity)*						Location in two categories*
Light to moderate intensity	2654	69.5				Outside	2319	62
High intensity	1164	30.5				Inside	1420	38
Total	3818	100				Total	3739	100
Type of activity in 2 categories (PA domain)*						Location in two categories*
Sports or leisure activity	2973	78.6				Not home	2276	60.9
Daily living activity	808	21.4				Home	1463	39.1
Total	3781	100				Total	3739	100

ap1 = action plan number 1; ap2 = action plan number 2; ap3 = action plan number 3; *information on the categorization of the variables; light to moderate intensity = walking, work-out (yoga), cycling (outdoor), household activities, other (canoeing, horseback riding, physical active game, skate, inline skate, skateboarding, stretching), dance, gymnastics, martial art, stair climbing; high intensity = work-out (weightlifting, circuit training, calisthenic, jump rope), other (ski), pole dancing, cycling (stationary), running, ball sports, racket sports, swimming; sports or leisure activity = PA domain ‘sports or leisure activity’, daily living activity = PA domain ‘active transport’, ‘household activities’ and ‘at work’; during the day = before noon, noon, afternoon; evening = evening; outside = nature, neighborhood, on the road, city, garden, sport facilities (outside); inside = home, sports facilities (inside), work; home = home, garden; not home = nature, neighborhood, on the road, city, work, sport facilities (outside) ; alone = alone, pet; not alone = same generation, different generation, group.

Table 4. Characteristics of the barriers.
Instrumental barriers (n_total= 4946; n_cp1a=2661, n_cp1b=940, n_cp2a=920, n_cp2b=244, n_cp3a=142, n_cp3b=39)
Type of barrier	n	%	Type of barrier	n	%
Weather	1364	27.6	Problem with location	81	1.6
Don't want to	862	17.5	Dark	58	1.2
Tired	609	12.3	Mental Health	14	0.3
No time	600	12.2	Link is gone	6	0.1
Physical health	236	4.8	No barrier	6	0.1
Other	230	4.7	Total	4935	100
No social support anymore	226	4.6
Forgotten	223	4.5	Type of barrier in two categories*
Problem with materials	157	3.2	Internal barrier	2658	56.6
Something came up	149	3	External barrier	2035	43.4
Too difficult	114	2.3	Total	4693	100
cp1a = first barrier of action plan 1; cp1b = second barrier of action plan 1; cp2a = first barrier of action plan 2; cp2b = second barrier of action plan 2; cp3a = first barrier of action plan 3; cp3b = second barrier of action plan 3; information on the categorization of the variables; internal barrier* = don’t want to, tired, no time, physical health, forgotten, too difficult, mental health; external barrier = weather, no social support anymore, problem with materials, something came up, problem with location, dark

Variability of emotions and physical complaints

Table 5 presents the mean scores and variability of the emotions and physical complaints. Overall, slightly more variability in emotions and physical complaints was found between days (within participants) compared to between participants, except for negative emotions where slightly more variability was found between participants. Especially for the physical complaint fatigue, the majority of variability (67.8%) was explained by differences between days (within participants).

Table 5

Mean scores and variability of the emotions and physical complains.
Variable	Mean (SD); range	Between-subject variance (%)	Within-subject variance (%)
Positive emotions	3.91 (0.85); [1.72–6.25]	45.6	54.4
Negative emotions	2.32 (0.84); [1-5.91]	55.1	44.9
Fatigue	3.41 (0.94); [1.5–6.63]	32.2	67.8
Pain	2.09 (0.93); [1-6.38]	47.0	53.0
Note: Range of the items: 1 (absolutely not) to 7 (very much)

Associations between emotions, physical complaints and the content of goals, action plans and barriers

Table 6 presents the results of the associations between emotions (i.e. positive and negative emotions), physical complaints (i.e. fatigue and pain) and the content of the goals, action plans and barriers.

Associations between positive emotions and the content of the goals, action plans and barriers

The results show that positive emotions were associated with the content of goals, action plans and barriers (Table 6). More specifically, both between- and within-individual differences in positive emotions were associated with the goal participants set in number of minutes PA, the location of the activity, whom participants did the activity with and barriers for doing the activity.

On days that participants had more positive emotions, the goal participants set in minutes PA was on average 5% higher as compared to other days (P < .001, β[CI] = 1.05 [1.03–1.07]). Likewise, participants who had on average more positive emotions compared to others, set their goal in minutes PA 7% higher as compared to others (P = .023, β[CI] = 1.07 [1.01–1.14]).

On days that participants had more positive emotions as compared to other days, they were more likely to choose an activity outside (P = .016, OR[CI] = 1.12 [1.02–1.23]), to plan activities not at home (P = .013, OR[CI] = 1.12 [1.02–1.23]) and to plan activities with at least one other person (P < .001, OR[CI] = 1.27 [1.16–1.38]). Likewise, participants who had on average more positive emotions compared to others were more likely to choose an activity outside (P = .011, OR[CI] = 1.23 [1.05–1.44]), to plan activities not at home (P = .007, OR[CI] = 1.23 [1.06–1.43];) and to plan activities with at least one other person (P < .001, OR[CI] = 1.28 [1.12–1.46]). Finally, on days that participants experienced more positive emotions, they were less likely to report internal barriers compared to external barriers for the chosen activity (P = .004, OR[CI] = 0.89 [0.83–0.96]). Similarly, participants who experienced on average more positive emotions than other participants were less likely to report internal compared to external barriers (P = .003, OR[CI] = 0.85 [0.76–0.94]).

Associations between negative emotions and the content of the goals, action plans and barriers

The results show that negative emotions were associated with the content of goals, but not with the content of the action plans and barriers (Table 6). More specifically, within-individual differences in negative emotions were associated with the goal participants set in number of minutes PA. On days when participants had more negative emotions, the goal participants set in minutes PA was on average 4% lower as compared to other days (P < .001, β[CI] = 0.96 [0.94–0.98]).

Associations between fatigue and the content of the goals, action plans and barriers

The results show that fatigue was associated with the content of goals, action plans and barriers (Table 6).

On days when participants were more fatigued, the goal participants set in minutes PA was on average 2% lower as compared to other days (P < .001, β[CI] = 0.98 [0.96–0.99]).

On days when participants reported to be more fatigued, they were less likely to plan the activity with at least one other person compared to plan the activity alone (P = .006, OR[CI] = 0.92 [0.86–0.98]). Moreover, on days on which participants were more fatigued, they were more likely to report internal compared to external barriers (P = .010, OR[CI] = 1.07 [1.02–1.13]). Likewise, participants who were on average more fatigued compared to others were also more likely to report internal barriers compared to external barriers for the chosen activity (P < .001, OR[CI] = 1.2 [1.09–1.32]).

Associations between pain and the content of the goals, action plans and barriers

The results show that pain was associated with the content of goals, action plans and barriers (Table 6). More specifically, within-individual differences in pain were associated with the goal participants set in number of minutes PA, the type of activity and the barriers for doing the activity.

On days when participants reported to have more pain, the goal participants set in minutes PA was on average 2% lower as compared to other days (P = .025, β[CI] = 0.98 [0.96-1.00]).

On days when participants reported to have more pain compared to other days, they were less likely to plan a vigorous intensity activity compared to a light to moderate intensity activity (P = .024, OR [CI] = 0.90 [0.83–0.99]) and were more likely to report internal barriers compared to external barriers for the chosen activity (P < .001, OR[CI] = 1.19 [1.08–1.24]).

Table 6

Associations between emotions, physical complaints and content of goals, action plans and barriers.
Characteristics of goals. action plans and barriers	Positive emotions				Negative emotions				Fatigue				Pain
	Between		Within		Between		Within		Between		Within		Between		Within
	β / OR [CI]	P-value	β /OR [CI]	P-value	β / OR [CI]	P-value	β /OR [CI]	P-value	β / OR [CI]	P-value	β /OR [CI]	P-value	β / OR [CI]	P-value	β/OR [CI]	P-value
Type of goal
Number of minutes PA (continuous)^a	1.07 [1.01; 1.14]	0.023*	1.05 [1.03; 1.07]	< 0.001*	0.98 [0.92; 1.04]	0.476	0.96 [0.94; 0.98]	< 0.001*	0.98 [0.93; 1.04]	0.504	0.98 [0.96; 0.99]	< 0.001*	0.98 [0.93; 1.04]	0.592	0.98 [0.96; 1.00]	0.025*
Number of steps (continuous)^b	487.0 [-90.0; 1064.0]	0.098	79.3 [-94.1; 252.7]	0.370	92.9 [-592.2; 406.4]	0.715	77.2 [-125.8: 280.1]	0.456	-121.0 [-608.4; 366.4]	0.627	45.2 [-80.7; 171.2]	0.481	-212.4 [-702.9; 278.1]	0.396	-120.0 [-243.7; 24.1]	0.101
Type of goal in steps (0) vs minutes PA (1)^c	0.94 [0.65; 1.36]	0.758	0.91 [0.78; 1.07]	0.273	1.05 [0.74; 1.51]	0.771	1.05 [0.87; 1.25]	0.631	0.91 [0.66; 1.27]	0.577	1.06 [0.96; 1.18]	0.266	1.15 [0.83; 1.60]	0.394	0.99 [0.85; 1.14]	0.833
Type of activity
Moderate intensity (0) vs vigorous intensity (1)^c	1.04 [0.90; 1.19]	0.616	0.93 [0.85; 1.02]	0.121	0.95 [0.83; 1.10]	0.497	0.95 [0.85; 1.06]	0.388	1.00 [0.89; 1.14]	0.924	1.05 [0.98; 1.12]	0.139	1.00 [0.88; 1.13]	0.952	0.90 [0.83; 0.99]	0.024*
Daily living activity (0) vs leisure activity (1)^c	0.97 [0.85; 1.11]	0.645	0.98 [0.89; 1.08]	0.692	1.00 [0.87; 1.14]	0.948	0.96 [0.86; 1.09]	0.598	1.01 [0.90; 1.14]	0.822	1.02 [0.95; 1.10]	0.569	1.12 [0.99; 1.28]	0.061	0.95 [0.87; 1.04]	0.330
Moment of the activity
During the day (0) vs evening (1)^c	0.91 [0.76; 1.09]	0.304	0.96 [0.85; 1.09]	0.533	0.97 [0.81; 1.16]	0.747	1.02 [0.89; 1.18]	0.761	1.16 [0.99; 1.36]	0.068	0.99 [0.91; 1.07]	0.810	0.90 [0.76; 1.06]	0.194	1.00 [0.90; 1.11]	0.930
Location of the activity
Inside (0) vs outside (1)^c	1.23 [1.05; 1.44]	0.011*	1.12 [1.02; 1.23]	0.016*	1.07 [0.91; 1.25]	0.419	0.95 [0.85; 1.06]	0.375	0.89 [0.77; 1.03]	0.114	0.95 [0.89; 1.02]	0.136	0.98 [0.85; 1.13]	0.767	0.98 [0.90; 1.07]	0.699
Home (0) vs not home (1)^c	1.23 [1.06; 1.43]	0.007*	1.12 [1.02; 1.23]	0.013*	1.00 [0.86; 1.17]	0.955	0.93 [0.84; 1.04]	0.186	0.87 [0.76; 1.00]	0.050	0.96 [0.90; 1.02	0.192	0.98 [0.86; 1.13]	0.802	0.99 [0.91; 1.08]	0.808
With whom doing the activity
Alone (0) vs with somebody else (1)^c	1.28 [1.12; 1.46]	< 0.001*	1.27 [1.16; 1.38]	< 0.001*	0.93 [0.81; 1.06]	0.255	0.92 [0.83; 1.03]	0.132	0.95 [0.85; 1.07]	0.422	0.92 [0.86; 0.98]	0.006*	1.05 [0.93; 1.18]	0.449	0.95 [0.88; 1.04]	0.238
Type of barrier
External barrier (0) vs internal barrier (1)^c	0.85 [0.76; 0.94]	0.003*	0.89 [0.83; 0.96]	0.004*	1.02 [0.92; 1.14]	0.718	1.03 [0.94; 1.13]	0.486	1.20 [1.09; 1.32]	< 0.001*	1.07 [1.02; 1.13]	0.010*	1.1 [1.00; 1.21]	0.058	1.19 [1.08; 1.24]	< 0.001*

^* P < 0.05 = significant predictors of attrition. β/OR = beta for each predictor with continuous outcome variable, odds ratio for each predictor with dummy coded outcome variable. CI = 95% confidence interval. ^aGamma model with log link function was used, ^bGamma model with identity link function was used, ^cBinomial model with logit link function was used.

The aim of the current study was to examine whether emotions and physical complaints influenced behavioural processes such as goal setting, action and coping planning. Insights gained from this study may contribute to more dynamic personalized support in the process of goal setting, action and coping planning in future digital health interventions.

Our study revealed that both emotions and physical complaints had an influence on the content of goals participants set. When participants had more positive emotions, they set a higher goal in terms of minutes PA, whereas if participants had more negative emotions, were more fatigued and experienced more pain, they set a lower goal in terms of minutes PA. Positive emotions, fatigue and pain also had an influence on the content of the action plans and the barriers of the coping plans participants created. When participants had more positive emotions, they preferred to choose activities outside, not at home and with someone else, whereas if participants were more fatigued, they preferred doing activities alone, and if participants experienced more pain, they preferred light to moderate instead of high intensity activities. Moreover, when participants had more positive emotions, they reported more external barriers, whereas when participants were more fatigued and experienced more pain, they reported more internal barriers. Although some variations in emotions and physical complaints between individuals influenced the above described behavioural processes of goal setting, action and coping planning, our findings suggest that mainly variations in emotions and physical complaints within individuals between days accounted for the influence on these behavioural processes. This is also supported by the fact that overall more variability in emotions and physical complaints (expect for negative emotions) was found within individuals than between individuals.

Results of this study may have implications for the way in which behaviour change theories are used and provide a first step to inform the development of personalized digital interventions. First, the majority of behaviour change theories, including HAPA, are predominantly used as static or linear frameworks, i.e. determinants are assumed to remain stable throughout the intervention or study (Riley et al. 2011). Several authors argued to transform these rather static frameworks into more dynamic frameworks so that they can inform the development of more dynamic digital health interventions (Rhodes et al. 2019; Riley et al. 2011). Findings from this study show that time-dependent variations in the HAPA determinant ‘situational barriers and opportunities’ (i.e. emotions and physical complaints) dynamically influenced other behavioural processes. The same might be applicable to other HAPA-determinants (e.g. intention or self-efficacy) or other theoretical frameworks. As such, this study highlights the importance to use theoretical frameworks and their determinants in a more dynamic way.

Second, the present study gives some first insights that may inform future digital ‘personalized’ interventions. Since variations in emotions and physical complaints within individuals influenced the content of goals, action plans and barriers, personalized suggestions may be provided to each individual in a more dynamic way depending on their emotional or physical state. For instance, suggestions of goals can be based on an individual’s actual PA level and can then be adjusted depending on their emotions or physical complaints of that day (i.e. higher PA goal when someone experiences positive emotions, lower PA goal when someone experiences negative emotions, fatigue or pain ). Likewise, the suggestions made about the content of the action plans in terms of intensity of the activity, location and company could be adjusted based on an individual’s emotional or physical state. Finally, also suggestions for coping plans may be personalized based on an individual’s emotional and physical state: external barriers may be suggested when someone feels more positive and internal barriers when someone feels more fatigued or experiences more pain. Previous research revealed that participants often experience difficulties in formulating coping plans, and providing tailored pre-formulated coping plans would be helpful to increase their quality and consequently their effectiveness (Degroote et al. 2020). Selecting suggestions based on an individual’s emotional and physical state could narrow down the number of suggestions.

When incorporating these insights into future personalized interventions, it is important to take the following points into account. First, a balance will have to be found between reinforcing existing patterns and challenging the participants to break through these patterns. For example individuals with chronic (low back) pain will avoid physical activities because it might be painful (Boutevillain et al. 2017), however evidence within chronic low back pain patients shows that participating in these activities even reduce pain on the long term (Chou & Huffman 2007). Likewise, suggesting to do an activity alone when someone is more fatigued compared to other days might not be the best suggestion. Although the individual might prefer to do an activity alone when they are fatigued, there might be a higher chance of carrying out the plan if they do it with someone else. Moreover, being physically active could even reduce negative affect (Schultchen et al. 2019). This is known as the exploration-exploitation dilemma: what an individual already knows is not necessarily the best (i.e. exploitation), exploring new possibilities and opportunities is also important (i.e. exploration) (Berger-Tal et al. 2014). Second, the association between an individual’s emotional and physical state on the one hand and the content of goals and plans on the other hand will undoubtedly not be the same for every individual. For example, we might not assume that everyone who experiences positive emotions will benefit from external barriers as a suggestion. There will always be a certain level of uncertainty. Therefore, there should always be an option to choose other suggestions when the proposed suggestions do not fulfill the individual’s needs. Since the current study also gave insights in which kind of goals, action plans and barriers participants’ created to do PA, it is possible to provide suggestions that were more frequently reported (e.g. the barriers ‘weather’, ‘don’t want to’ and ‘tired’).

Although the results of this study provide some first insights into dynamic personalized suggestions, no firm conclusions can be drawn since it was not investigated whether the found associations in this study actually contributed to successful achievement of the PA plan or not (within the evening questionnaire it was evaluated whether participants achieved their goal in general, but it was not investigated whether the PA plan was successfully achieved or not). Therefore, further investigation is warranted which may involve several aspects. First, future studies should involve different stakeholders (e.g. domain experts, end users) to further validate and refine the associations found in this study before using them as suggestions in digital health interventions. This might be done in workshops where certain scenarios can be discussed (e.g. discussing a possible PA plan of a user who experiences negative emotions that day). Second, future studies might perform N-of-1 experimental designs (also called single- subjects designs or single-case designs) (McDonald et al. 2017) to investigate which suggestions are most beneficial for PA. N-of-1 designs usually test interventions or intervention components by experimentally manipulating the intervention (component) within participants, within a certain period of time, in a number of subjects (McDonald et al. 2017). For instance, during a study which may last for example 30 days, personal information that varies daily (e.g. emotions and physical complaints) might be accessed via short self-reports. On randomly allocated days, random suggestions of plans (not personalized) might be delivered, whilst on other days, personalized suggestions might be provided (personalized to daily variations based on the self-reports). When measuring the PA levels using accelerometry, and asking each evening whether they enacted on their PA plan suggestions, it will be possible to identify whether these personalized suggestions were effective and valuable for the promotion of PA. Third, the two previous aspects further described the validation and refinement of the associations found in this study. However in order to develop digital interventions with more dynamic personalized suggestions, these associations should be embedded in a sort of computerized system to actually recommend suggestions to participants (Çelik Ertuğrul & Elçi 2020). Previous work from our research group already stated that the use of a knowledge-based system would be appropriate (Schroé et al. 2022a). A knowledge-based system is a computer program that contains a knowledge base which represents knowledge in a particular domain (Pratiwi et al. 2018) and reasons with that information via intelligent algorithms in order to deliver personalized suggestions. Based on our findings from the current study, the knowledge-base may contain associations between emotions, physical complains and content of goals, action plans and coping plans. Admittedly, such a knowledge-base should contain more information from possible end-users to deliver dynamic personalized support, since exploring how emotions and physical complaints relate towards goals, action and coping plans is only a small part of the puzzle. Previous work already investigated how personal user information (e.g. gender, age, motivational stage, etc) related towards the content of action and coping plans using a clustering method (Schroé et al. 2022a). Although this approach was promising, this data was relatively stable over time and would only result in personalized suggestions at the level of the user profile. As such, the findings of the current paper contribute to more dynamic personalized suggestions which can vary from day to day. In order to make further progress in the field, future studies should also explore how other dynamic contextual user information (e.g. changing home and work environment, weather of the day, agenda of the day) relates toward the content of goals, action and coping plans.

Strengths and limitations

This study has several strengths. First, to the best of our knowledge, this is the first paper assessing whether variations in daily emotions and physical complaints influenced behavioural processes such as goal setting, action and coping planning. These variations were already investigated in relation towards physical activity levels in itself (Schultchen et al. 2019; Vetrovsky et al. 2021), but not in relation to behavioural processes of physical activity. In doing so, the findings from the current study contribute towards more dynamic personalized support in digital health interventions. Second, both between- as well as within-individual variations in emotions and physical complaints were assessed, which enables to distinguish which suggestions might be delivered at an inter- or intra-individual level. Third, the study sample was large (n = 362 students) and completion rates of the morning questionnaires were high, which provided us with a large dataset of goals, action and coping plans.

This study also has a number of limitations. First, since the current study sample only included university students (i.e. Master in Clinical Psychology) of which 88,7% were female, the findings of this study cannot be generalized to a broader population. Results may be different in other target groups. For instance, research already showed that older adults show less variability in positive and negative emotions than younger adults (Röcke et al. 2009). Second, the study took place when strict corona measures were installed by the Belgian Government, this consequently affected the content of action and coping plans participants could create. Notwithstanding, this study might also be seen as a conceptual study which gives insights in how knowledge can be acquired to provide more personalized suggestions in future digital health interventions. Third, emotions and physical complaints may not only vary across days but also within days (May et al. 2018; Pe & Kuppens 2012), yet this was not explored in this study since emotions and physical complaints were only assessed once in the morning. It may well be that participants for example experience positive emotions in the morning and more negative emotions emerge throughout the day (e.g. because it is a stressful day at work). Consequently, the given suggestion in the morning might be less applicable. Since daily cycles of goal setting, action and coping planning were found to be effective in MyDayPlan (Degroote et al. 2021), we do not argue for shorter cycles as this might not be feasible and be a burden on participants. Notwithstanding, advanced designs such as ecological momentary assessment (EMA), which works with repeated measurements throughout the day (Dunton 2017), might inform how changing emotions and physical complaints influence plans or activities in a smaller timeframe. Accordingly, this information may help to adapt plans throughout the day in future interventions. Fourth, since our study sample was homogeneous (i.e. highly educated female university students), no moderation effects were examined in this study (e.g. gender, age, socio-economic status). However, moderation from other variables remains a possibility. For example, motivational stage of participants (Lippke et al. 2005; Schwarzer et al. 2011) might play a role. It may well be that intenders (i.e. participants who already developed an intention to change their behaviour but did not perform PA regularly yet) react differently when they experience negative emotions for example compared to actors (i.e. participants who already perform PA regularly).

BCT

behaviour change technique

EMA

ecological momentary assessment

HAPA

health action process approach

physical activity

Ethics approval and consent to participate

The research protocol of the current study was approved by the Ethical Committee from the Faculty of Psychology and Educational Sciences at Ghent University (registration number: 2020/87). The authors confirm that all methods were carried out in accordance with relevant guidelines and regulations. Participants provided informed consent regarding the study measures before enrolling in the study.

Consent for publication

Not applicable.

Availability of data and materials

Data supporting the results reported in this article are stored at the University of Ghent, Belgium. The datasets used during the current study are available from the corresponding author on reasonable request.

Competing interests

The authors declare that they have no competing interests.

Funding

The research of author HS is funded by the Special Research Fund of Ghent University (BOF). The funding body is not involved in the study design, collection, management, analysis and interpretation of data or in writing of the manuscript.

Authors’ contributions

All authors were involved in designing the study. The authors HS and ADP analyzed and interpret the data. The author HS drafted the manuscript. DVD, ADP, MB and GC critically revised the manuscript. All authors read and approved the final manuscript.

Ainsworth BE, Haskell WL, Whitt MC, Irwin ML, Swartz AM, Strath SJ, O'Brien WL, Bassett DR, Jr., Schmitz KH, Emplaincourt PO, Jacobs DR, Jr., and Leon AS. 2000. Compendium of physical activities: an update of activity codes and MET intensities. Medicine and science in sports and exercise 32:S498-504. 10.1097/00005768-200009001-00009
Bassuk SS, and Manson JE. 2005. Epidemiological evidence for the role of physical activity in reducing risk of type 2 diabetes and cardiovascular disease. Journal of applied physiology (Bethesda, Md : 1985) 99:1193-1204. 10.1152/japplphysiol.00160.2005
Bates D, Mächler M, Bolker B, and Walker S. 2014. Fitting linear mixed-effects models using lme4. Available at http://arxiv.org/abs/14065823.
Berger-Tal O, Nathan J, Meron E, and Saltz D. 2014. The Exploration-Exploitation Dilemma: A Multidisciplinary Framework. PloS one 9:e95693. 10.1371/journal.pone.0095693
Boutevillain L, Dupeyron A, Rouch C, Richard E, and Coudeyre E. 2017. Facilitators and barriers to physical activity in people with chronic low back pain: A qualitative study. PloS one 12:e0179826. 10.1371/journal.pone.0179826
Broekhuizen K, Simmons D, Devlieger R, van Assche A, Jans G, Galjaard S, Corcoy R, Adelantado JM, Dunne F, Desoye G, Harreiter J, Kautzky-Willer A, Damm P, Mathiesen ER, Jensen DM, Andersen LL, Lapolla A, Dalfra MG, Bertolotto A, Wender-Ozegowska E, Zawiejska A, Hill D, Snoek FJ, Jelsma JGM, Bosmans JE, van Poppel MNM, and van Dongen JM. 2018. Cost-effectiveness of healthy eating and/or physical activity promotion in pregnant women at increased risk of gestational diabetes mellitus: economic evaluation alongside the DALI study, a European multicenter randomized controlled trial. The international journal of behavioral nutrition and physical activity 15:23. 10.1186/s12966-018-0643-y
Çelik Ertuğrul D, and Elçi A. 2020. A survey on semanticized and personalized health recommender systems. Expert Systems 37:e12519. https://doi.org/10.1111/exsy.12519
Chou R, and Huffman LH. 2007. Nonpharmacologic therapies for acute and chronic low back pain: a review of the evidence for an American Pain Society/American College of Physicians clinical practice guideline. Annals of internal medicine 147:492-504. 10.7326/0003-4819-147-7-200710020-00007
Cicchetti DV. 2001. Methodological Commentary The Precision of Reliability and Validity Estimates Re-Visited: Distinguishing Between Clinical and Statistical Significance of Sample Size Requirements. Journal of Clinical and Experimental Neuropsychology 23:695-700. 10.1076/jcen.23.5.695.1249
Conroy DE, Hojjatinia S, Lagoa CM, Yang C-H, Lanza ST, and Smyth JM. 2019. Personalized models of physical activity responses to text message micro-interventions: A proof-of-concept application of control systems engineering methods. Psychology of Sport and Exercise 41:172-180. https://doi.org/10.1016/j.psychsport.2018.06.011
Davis A, Sweigart R, and Ellis R. 2020. A systematic review of tailored mHealth interventions for physical activity promotion among adults. Translational behavioral medicine 10:1221-1232. 10.1093/tbm/ibz190
Degroote L, De Paepe A, De Bourdeaudhuij I, Van Dyck D, and Crombez G. 2021. Effectiveness of the mHealth intervention ‘MyDayPlan’ to increase physical activity: an aggregated single case approach. International Journal of Behavioral Nutrition and Physical Activity 18:92. 10.1186/s12966-021-01163-2
Degroote L, Van Dyck D, De Bourdeaudhuij I, De Paepe A, and Crombez G. 2020. Acceptability and feasibility of the mHealth intervention ‘MyDayPlan’ to increase physical activity in a general adult population. BMC Public Health 20:1032. 10.1186/s12889-020-09148-9
Delespaul PAEG. 1995. Assessing Schizophrenia in Daily Life. The Experience Sampling Method.
Dunton GF. 2017. Ecological Momentary Assessment in Physical Activity Research. Exercise and Sport Sciences Reviews 45:48-54. 10.1249/jes.0000000000000092
Dunton GF, Atienza AA, Castro CM, and King AC. 2010. Using Ecological Momentary Assessment to Examine Antecedents and Correlates of Physical Activity Bouts in Adults Age 50+ Years: A Pilot Study. Annals of Behavioral Medicine 38:249-255. 10.1007/s12160-009-9141-4
Enders CK, and Tofighi D. 2007. Centering predictor variables in cross-sectional multilevel models: a new look at an old issue. Psychological methods 12:121-138. 10.1037/1082-989x.12.2.121
Eysenbach G. 2005. The Law of Attrition. J Med Internet Res 7:e11. 10.2196/jmir.7.1.e11
Geraghty AW, Torres LD, Leykin Y, Pérez-Stable EJ, and Muñoz RF. 2013. Understanding attrition from international Internet health interventions: a step towards global eHealth. Health promotion international 28:442-452. 10.1093/heapro/das029
Germeys L, and De Gieter S. 2018. A diary study on the role of psychological detachment in the spillover of self-control demands to employees’ ego depletion and the crossover to their partner. European Journal of Work and Organizational Psychology 27:140-152. 10.1080/1359432X.2017.1417259
Haskell WL, Blair SN, and Hill JO. 2009. Physical activity: health outcomes and importance for public health policy. Preventive medicine 49:280-282. 10.1016/j.ypmed.2009.05.002
Kim H, Mentzer J, and Taira R. 2019. Developing a Physical Activity Ontology to Support the Interoperability of Physical Activity Data. Journal of medical Internet research 21:e12776. 10.2196/12776
Knittle K, Heino M, Marques MM, Stenius M, Beattie M, Ehbrecht F, Hagger MS, Hardeman W, and Hankonen N. 2020. The compendium of self-enactable techniques to change and self-manage motivation and behaviour v.1.0. Nature Human Behaviour 4:215-223. 10.1038/s41562-019-0798-9
Koo TK, and Li MY. 2016. A Guideline of Selecting and Reporting Intraclass Correlation Coefficients for Reliability Research. Journal of Chiropractic Medicine 15:155-163. https://doi.org/10.1016/j.jcm.2016.02.012
Kroenke K, Spitzer RL, and Williams JBW. 2002. The PHQ-15: Validity of a New Measure for Evaluating the Severity of Somatic Symptoms. Psychosomatic Medicine 64.
Lippke S, Ziegelmann JP, and Schwarzer R. 2005. Stage-specific adoption and maintenance of physical activity: Testing a three-stage model. Psychology of Sport and Exercise 6:585-603. 10.1016/j.psychsport.2004.11.002
Lustria ML, Noar SM, Cortese J, Van Stee SK, Glueckauf RL, and Lee J. 2013. A meta-analysis of web-delivered tailored health behavior change interventions. Journal of health communication 18:1039-1069. 10.1080/10810730.2013.768727
May M, Junghaenel DU, Ono M, Stone AA, and Schneider S. 2018. Ecological Momentary Assessment Methodology in Chronic Pain Research: A Systematic Review. The journal of pain 19:699-716. 10.1016/j.jpain.2018.01.006
McDonald S, Quinn F, Vieira R, O'Brien N, White M, Johnston DW, and Sniehotta FF. 2017. The state of the art and future opportunities for using longitudinal n-of-1 methods in health behaviour research: a systematic literature overview. Health Psychology Review 11:307-323. 10.1080/17437199.2017.1316672
Pe ML, and Kuppens P. 2012. The dynamic interplay between emotions in daily life: Augmentation, blunting, and the role of appraisal overlap. Emotion 12:1320-1328. 10.1037/a0028262
Poppe L, De Bourdeaudhuij I, Verloigne M, Shadid S, Van Cauwenberg J, Compernolle S, and Crombez G. 2019. Efficacy of a Self-Regulation-Based Electronic and Mobile Health Intervention Targeting an Active Lifestyle in Adults Having Type 2 Diabetes and in Adults Aged 50 Years or Older: Two Randomized Controlled Trials. Journal of medical Internet research 21:e13363. 10.2196/13363
Poppe L, De Paepe AL, Van Ryckeghem DML, Van Dyck D, Maes I, and Crombez G. 2021. The impact of mental and somatic stressors on physical activity and sedentary behaviour in adults with type 2 diabetes mellitus: a diary study. PeerJ 9:e11579. 10.7717/peerj.11579
Poppe L, Van der Mispel C, Crombez G, De Bourdeaudhuij I, Schroé H, and Verloigne M. 2018. How Users Experience and Use an eHealth Intervention Based on Self-Regulation: Mixed-Methods Study. Journal of medical Internet research 20:e10412. 10.2196/10412
Pratiwi PS, Xu Y, Li, Yuefeng, Trost SG, Clanchy KM, and Tjondronegoro DW. 2018. User profile ontology to support personalization for e-Coaching systems. International Workshop on ACM International Conference on Information and Knowledge Management (CIKM).
Reinwand DA, Crutzen R, Storm V, Wienert J, Kuhlmann T, de Vries H, and Lippke S. 2016. Generating and predicting high quality action plans to facilitate physical activity and fruit and vegetable consumption: results from an experimental arm of a randomised controlled trial. BMC Public Health 16:317. 10.1186/s12889-016-2975-3
Rhodes RE, McEwan D, and Rebar AL. 2019. Theories of physical activity behaviour change: A history and synthesis of approaches. Psychology of Sport and Exercise 42:100-109. https://doi.org/10.1016/j.psychsport.2018.11.010
Riley WT, Rivera DE, Atienza AA, Nilsen W, Allison SM, and Mermelstein R. 2011. Health behavior models in the age of mobile interventions: are our theories up to the task? Translational behavioral medicine 1:53-71. 10.1007/s13142-011-0021-7
Röcke C, Li SC, and Smith J. 2009. Intraindividual variability in positive and negative affect over 45 days: do older adults fluctuate less than young adults? Psychology and aging 24:863-878. 10.1037/a0016276
Rotter JB, and Mulry RC. 1965. Internal versus external control of reinforcement and decision time. Journal of personality and social psychology 2:598-604. 10.1037/h0022473
Schroé H. 2022. MyActionPlan: Towards more perzonalized digital health interventions. 31 March 2022 ed. OSF.
Schroé H, Carlier S, Van Dyck D, De Backere F, and Crombez G. 2022a. Towards more personalized digital health interventions: a clustering method of action and coping plans to promote physical activity. BMC Public Health 22:2325. 10.1186/s12889-022-14455-4
Schroé H, Crombez G, De Bourdeaudhuij I, and Van Dyck D. 2022b. Investigating When, Which, and Why Users Stop Using a Digital Health Intervention to Promote an Active Lifestyle: Secondary Analysis With A Focus on Health Action Process Approach-Based Psychological Determinants. JMIR mHealth and uHealth 10:e30583. 10.2196/30583
Schroé H, Van Dyck D, De Paepe A, Poppe L, Loh WW, Verloigne M, Loeys T, De Bourdeaudhuij I, and Crombez G. 2020. Which behaviour change techniques are effective to promote physical activity and reduce sedentary behaviour in adults: a factorial randomized trial of an e- and m-health intervention. International Journal of Behavioral Nutrition and Physical Activity 17:127. 10.1186/s12966-020-01001-x
Schultchen D, Reichenberger J, Mittl T, Weh TRM, Smyth JM, Blechert J, and Pollatos O. 2019. Bidirectional relationship of stress and affect with physical activity and healthy eating. British Journal of Health Psychology 24:315-333. 10.1111/bjhp.12355
Schwarzer R. 2008. Modeling Health Behavior Change: How to Predict and Modify the Adoption and Maintenance of Health Behaviors. Applied Psychology 57:1-29. https://doi.org/10.1111/j.1464-0597.2007.00325.x
Schwarzer R, Lippke S, and Luszczynska A. 2011. Mechanisms of health behavior change in persons with chronic illness or disability: the Health Action Process Approach (HAPA). Rehabilitation psychology 56:161-170. 10.1037/a0024509
Schwarzer R, and Luszczynska A. 2008. How to overcome health-compromising behaviors: The health action process approach. European Psychologist 13:141-151. 10.1027/1016-9040.13.2.141
Sniehotta FF, Scholz U, and Schwarzer R. 2005. Bridging the intention–behaviour gap: Planning, self-efficacy, and action control in the adoption and maintenance of physical exercise. Psychology & Health 20:143-160. 10.1080/08870440512331317670
Sniehotta FF, Scholz U, and Schwarzer R. 2006. Action plans and coping plans for physical exercise: A longitudinal intervention study in cardiac rehabilitation. British Journal of Health Psychology 11:23-37. https://doi.org/10.1348/135910705X43804
Strath SJ, Kaminsky LA, Ainsworth BE, Ekelund U, Freedson PS, Gary RA, Richardson CR, Smith DT, and Swartz AM. 2013. Guide to the Assessment of Physical Activity: Clinical and Research Applications. Circulation 128:2259-2279. doi:10.1161/01.cir.0000435708.67487.da
Sutton S. 2008. How does the Health Action Process Approach (HAPA) Bridge the Intention–Behavior Gap? An Examination of the Model's Causal Structure. Applied Psychology 57:66-74. https://doi.org/10.1111/j.1464-0597.2007.00326.x
Team RC. 2020. A language and environment for statistical computing. . Available at https://www.r-project.org.
Van der Mispel C, Poppe L, Crombez G, Verloigne M, and De Bourdeaudhuij I. 2017. A Self-Regulation-Based eHealth Intervention to Promote a Healthy Lifestyle: Investigating User and Website Characteristics Related to Attrition. J Med Internet Res 19:e241. 10.2196/jmir.7277
Vandelanotte C, Müller AM, Short CE, Hingle M, Nathan N, Williams SL, Lopez ML, Parekh S, and Maher CA. 2016. Past, Present, and Future of eHealth and mHealth Research to Improve Physical Activity and Dietary Behaviors. Journal of nutrition education and behavior 48:219-228.e211. 10.1016/j.jneb.2015.12.006
Vetrovsky T, Omcirk D, Malecek J, Stastny P, Steffl M, and Tufano JJ. 2021. Morning fatigue and structured exercise interact to affect non-exercise physical activity of fit and healthy older adults. BMC geriatrics 21:179. 10.1186/s12877-021-02131-y
Warburton DE, Nicol CW, and Bredin SS. 2006. Health benefits of physical activity: the evidence. CMAJ : Canadian Medical Association journal = journal de l'Association medicale canadienne 174:801-809. 10.1503/cmaj.051351
Webb TL, Joseph J, Yardley L, and Michie S. 2010. Using the Internet to Promote Health Behavior Change: A Systematic Review and Meta-analysis of the Impact of Theoretical Basis, Use of Behavior Change Techniques, and Mode of Delivery on Efficacy. J Med Internet Res 12:e4. 10.2196/jmir.1376
WHO. 2018a. Classification of digital health interventions.
WHO. 2018b. Global action plan on physical activity 2018-2030: More active people for a healthier world

Download PDF

Version 1

posted

You are reading this latest preprint version

Towards more personalized digital health interventions: impact of emotions and physical complaints on goals, action and coping plans to promote physical activity

Status:

Version 1

Abstract

Background.

Methods.

Results.

Conclusions.

Figures

Introduction

Materials & Methods

Statistical analysis

Results

Discussion

List Of Abbreviations

Declarations

References

Status:

Version 1