Supporting parents of children with chronic conditions : a randomized controlled trial of web-based and self-help ACT interventions

This is a self-archived version of an original article. This version may differ from the original in pagination and typographic details. Author(s): Title: Year: Version: Copyright: Rights: Rights url: Please cite the original version: CC BY-NC-ND 4.0 https://creativecommons.org/licenses/by-nc-nd/4.0/ Supporting parents of children with chronic conditions : a randomized controlled trial of web-based and self-help ACT interventions © 2021 The Authors. Published by Elsevier B.V. Published version Lappalainen, Päivi; Pakkala, Inka; Strömmer, Juho; Sairanen, Essi; Kaipainen, Kirsikka; Lappalainen, Raimo Lappalainen, P., Pakkala, I., Strömmer, J., Sairanen, E., Kaipainen, K., & Lappalainen, R. (2021). Supporting parents of children with chronic conditions : a randomized controlled trial of webbased and self-help ACT interventions. Internet Interventions, 24, Article 100382. https://doi.org/10.1016/j.invent.2021.100382 2021 Internet Interventions 24 (2021) 100382 Available online 16 March 2021 2214-7829/© 2021 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). Supporting parents of children with chronic conditions: A randomized controlled trial of web-based and self-help ACT interventions P¨ aivi Lappalainen a , b , * , Inka Pakkala a , b , Juho Str¨ ommer a , Essi Sairanen c , Kirsikka Kaipainen d , Raimo Lappalainen a a University of Jyv¨ askyl¨ a, Department of Psychology, Jyv¨ askyl¨ a, Finland b The Gerocenter Foundation for Research and Development, Jyv¨ askyl¨ a, Finland c Karlstad University, Department of Social and Psychological Studies, Karlstad, Sweden d Tampere University, Unit of Computing Sciences, Tampere, Finland ARTICLE INFO Keywords: Acceptance and commitment therapy Internet Self-help Parents of children with chronic conditions Burnout Video conferencing ABSTRACT Objective: Many parents of children with chronic conditions and developmental disabilities experience high rates of burnout and psychological distress. The aim of the current study was to examine the effects of two differently delivered interventions based on acceptance and commitment therapy (ACT) on symptoms of burnout, depression, quality of life, psychological flexibility, and mindfulness skills. Method: A total of 110 parents of children aged 0.8 to 17 years with chronic conditions and developmental disabilities participated in a randomized controlled trial lasting 13 weeks with two intervention groups: (1) an iACT intervention, including three psychologist-led video conferencing sessions, and (2) a self-help ACT, which received self-help material: an ACT-based booklet with the possibility of using ACT-based online exercises. Results: For the main outcome measure burnout, a similar decrease was found in both groups. The supported iACT intervention produced significantly larger improvements in depressive symptoms (d =0.49), psychological flexibility (d =0.64), and mindfulness (d =0.55) compared to the self-help ACT intervention. For health-related quality of life, only the dimension of role limitations caused by emotional problems showed a significant difference in favor of the supported iACT (d =0.58). Conclusions: The results suggest that the iACT intervention including three video conferencing sessions with a psychologist produced broader improvements in parents’ psychological well-being than the self-help ACT. Overall, Internet-delivered interventions and video conferencing technology may offer a feasible alternative to psychological support and self-care for parents of children with chronic conditions. Further research is needed to investigate the long-term effects of the current delivery models. 1. Introduction The number of chronic health conditions and disabilities among children has steadily increased in the last decades (Perrin et al., 2014). Children with long-term conditions and complex needs require several hours of additional care and assistance on a daily basis compared to their healthy counterparts (Heyman et al., 2004), thereby posing extra burdens on their parents’ well-being. Parents of children with chronic conditions have been found to experience distress, insecurity, loss of self-esteem, negative emotions, anxiety, disease-related fear, depression, and physical and mental distress (Boman et al., 2004). Burnout symptoms are more typical of parents of chronically ill children than those of healthy children, and mothers, in particular, have been found to suffer from burnout (Lindstr¨ om et al., 2010). Although the level of distress diminishes over time, some symptoms persist for years after a child has been diagnosed with a serious illness (Boman et al., 2003). In addition, studies show that parents with a child with a developmental abnormality perceive their quality of life to be lower than those whose child develops normally (Huang et al., 2014; Xiang et al., 2009). As there is abundant research suggesting that parental distress is associated with a large number of adverse family outcomes, including increased mental health problems in children (Amrock and Weitzman, 2014; Merrill et al., 2007) and poorer long-term adjustment to the child’s illness (Rabineau et al., 2008), there is a need for interventions * Corresponding author at: Department of Psychology, FIN-40014, University of Jyv¨ askyl¨ a, Finland. E-mail address: [email protected] (P. Lappalainen). Contents lists available at ScienceDirect Internet Interventions journal homepage: www.elsevier.com/locate/invent https://doi.org/10.1016/j.invent.2021.100382 Received 8 January 2021; Received in revised form 9 March 2021; Accepted 10 March 2021 Internet Interventions 24 (2021) 100382 2 aimed at reducing parental strain. Parents of children with chronic conditions may engage in a variety of avoidant coping strategies to deal with their child’s illness, such as eliminating or avoiding painful thoughts, feelings, and sensations related to the child’s condition. Avoidant coping strategies may provide temporary relief but may lead to increased distress in the long term (Bardeen and Fergus, 2016; Hayes et al., 2006). Evidence shows that parents who have poor psychological flexibility, i.e., engage in avoidant coping strategies, may experience challenges when adjusting to caring for their child with a chronic condition and may exhibit psychological symptoms such as anxiety, depression, and stress (e.g., Chong et al., 2017). Increasing parents’ ability to take an accepting, non-judging stance toward their own thoughts and feelings and their child’s potential behavioral problems may assist parents in regulating their emotions and choosing more appropriate parenting behaviors (Prevedini et al., 2020). A third wave of cognitive behavioral therapy (CBT) or the process-based CBT approach, such as acceptance and commitment therapy (ACT, Hayes et al., 1999), may be particularly useful in addressing avoidant coping strategies and promoting skills that support children’s socioemotional development (Lunsky et al., 2018). In recent years, there has been an increase in the number of ACTbased interventions for supporting parents of children with chronic conditions. These interventions have varied in duration and have been targeted at parents whose children have been diagnosed with a variety of long-term illnesses, such as asthma (Chong et al., 2017, 2019), autism spectrum disorder (Blackledge and Hayes, 2006; Fung et al., 2018; Joekar et al., 2016; Lunsky et al., 2012; Lunsky et al., 2018; Poddar et al., 2015; Zody, 2017), cerebral palsy (Whittingham et al., 2016; Whittingham et al., 2019; Whittingham et al., 2020), intellectual disabilities (Reid et al., 2016), life-threatening illnesses (Burke et al., 2014), diabetes, and other chronic illnesses (Sairanen et al., 2019). According to recent meta-analyses and reviews (Byrne et al., in press; Parmar et al., 2019; Ruskin et al., in press; Yu et al., 2019), interventions based on ACT, mindfulness, and acceptance show promise in helping parents cope with the stress and psychological distress associated with having children with chronic conditions. As most studies have predominantly examined the impact of individually or group-delivered interventions on the well-being of parents of children with chronic conditions, there is a need for more research on alternative ways to support these parents, who often lack the professional help and services and, owing to time constraints, spend less time on personal care (Brandon, 2007). Furthermore, the barrier of access to support cannot be overcome with the dominant models of delivering psychosocial interventions (Kazdin, 2019). In countries where caring for chronically ill children often coincides with working life, parents may find it particularly difficult or nearly impossible to incorporate the care of a chronically ill child, employment, and self-care into their schedule. Therefore, parents of children with chronic conditions could benefit from novel models of delivering support that are not tied to time or place, allowing them to devote time to self-care whenever they have the time and opportunity to do so. Feasible options to provide parents support are self-help models or web-based interventions combined with video conferencing technology. Currently, there is substantial research supporting the clinical effectiveness of video therapy (Simpson et al., 2020). So far, only few trials have investigated the efficacy of ACT-based online interventions for parents of children with chronic conditions (Sairanen et al., 2019; Whittingham et al., 2016), and an online parent support program for families of infants identified as at high risk of cerebral palsy, which is currently being implemented (Whittingham et al., 2020). The ACTbased web intervention, developed and tested by Sairanen et al. (2019), produced large between-group effect sizes for burnout, depression and anxiety symptoms, and aspects of mindfulness in comparison to a no-treatment (wait-list control) group. As our research lab participated in the development of the Swedish web-based program utilized in the Sairanen et al. (2019) study, the equivalent web program provided in another language (Finnish) and in another country was utilized in the current study. 1.1. Aim of the study As the study of Sairanen et al. concluded that Internet-delivered ACT with the guidance of student therapists was effective at enhancing the psychological well-being of parents, we wanted to confirm these results and, furthermore, investigate whether comparable results could be attained with two differently delivered ACT-based interventions. We were especially interested in investigating the impact of a psychologistsupported iACT intervention using video conferencing technology in comparison to a self-help ACT intervention also including ACT-based exercises online but without any guidance. Thus, the aim of the current non-inferiority study was to examine whether the psychologistsupported iACT intervention with three video conferencing sessions would be superior to the self-help ACT intervention in reducing burnout and depressive symptoms and enhancing psychological flexibility, mindfulness, and quality of life in parents. 2. Methods 2.1. Settings and participants The study was conducted in the Psychology Department Clinic at the University of Jyv¨ askyl¨ a in 2019–2020. Ethical approval for the study was obtained from the Ethical Committee of the Central Finland Health Care District (Board Affiliation: Central Finland Central Hospital. Approval Number 12/U2018 on November, 6, 2018, registered at www. clinicaltrials.gov; ClinicalTrials.gov, Identifier: NCT04250012, Protocol ID 18/26/2018). The participants took part in the study on a voluntary basis and gave written informed consent for their participation. Parents were recruited through advertisements in local newspapers and the Facebook pages of diverse caregiver and family organizations. Interested parents were invited to contact the study team by phone or by sending their contact information via e-mail, after which an appointment for a telephone interview was arranged. In the phone call, preliminary information about the study was provided, and initial inclusion criteria were assessed. Those willing to participate received a link to an online survey, including a questionnaire on burnout that confirmed the suitability of the parent for the study. Parents who fulfilled the following inclusion criteria were eligible to participate: 1) had a child under the age of 18 with a chronic condition or developmental disability; 2) scored greater than 2.75 points on the Shirom-Melamed Burnout Questionnaire (SMBQ), indicating mild symptoms of burnout (Lundgren-Nilsson et al., 2012; Melamed et al., 1999); 3) had no parallel psychological treatment; 4) did not suffer from a serious mental disorder; 5) had sufficient Finnish language skills; and 6) had access to the Internet through a computer or other device. The data were collected in three phases, with the first wave starting in February 2019 and the last post-measurements taking place in May/ June 2020. The randomization in the two groups followed each data collection using a randomization tool (see www.random.org). A total of 119 interested parents contacted the study team (see study flow in Fig. 1). Of these, nine were excluded: four of them did not meet the burnout (SMBQ) inclusion criteria; four dropped out before the premeasurement; and one could not participate due to time constraints. Altogether, 110 parents comprising 103 females (93.6%) and seven males (6.4%) completed the pre-measurements, including background information (supported iACT, n =57; ACT self-help, n =53, Table 1). The mean age of the parents was 40.1 years (SD =6.68, range 27–55), and close to 70% had a university-level education. A significant difference between the groups was noted in the age of the parents, with the parents in the iACT group being slightly younger (M =38.9 years; SD = 6.85) compared to those in the self-help ACT group (M =41.5 years; SD =6.29). However, we considered this difference (39 vs. 41.5 years) as P. Lappalainen et al. Internet Interventions 24 (2021) 100382 3 clinically non-significant. Sample demographics are presented in Table 1. 2.2. Interventions 2.2.1. Description of the supported iACT intervention The iACT group (n =57) participated in a web-based ACT intervention, including a web-program called Uupu and three online meetings with a psychologist via the doxy.me telemedicine solution. The parents received credentials to the program website (www.uupu.fi), and the intervention lasted 13 weeks, of which the web-based program lasted 10 weeks and included five modules (2 weeks each), according to an ACT approach: (1) What is important to you? (2) meaningful actions; (3) present moment; (4) distance to thoughts and feelings; and (5) acceptance and self-compassion (see Table 2). The modules were composed of ACT-related content such as text, pictures, audio exercises, and videos. The content of the web-based program was identical to that in the study of Sairanen et al. (2019), except for changes related to data security according to the wishes of the Ethical Committee: 1) the discussion forum was left out; 2) the program did not include any possibility to write reflections or store any input on the program platform; and 3) email reminders were not sent to the participants. Instead, the parents were encouraged to write and reflect on the two-weekly home assignment for each module in a diary in paper form, which was sent to them by regular mail before the start of the intervention. In addition, the intervention included three 45-minute online meetings with a clinical psychologist using the Doxy.me telecommunication application (www.doxy.me). The two psychologists, who worked as coaches, had ample experience in ACT. The first virtual session was arranged at the beginning of the intervention—before access to the web-program—the second after the parents had worked through the first two modules (i.e., four weeks from the start), and the third online meeting took place at the end of the intervention (i.e., weeks 11–12). During the first appointment, each parent’s set of problems, assets, and circumstances were mapped using the modified psychosocial interview model (see Strosahl et al., 2012), including questions related to family and work life, leisure, and health behaviors. Based on the interview, the psychologist prepared a vector-graphic case formulation of the parent’s overall situation (according to the Functional Analytic Clinical Case Model, FACCM; Haynes and O’Brien, 2000; Haynes et al., 2011), which was sent to the parent via regular mail. During the second appointment, the case formulation was discussed, and according to the content of the first two modules, concrete shortand long-term goals were discussed and set to guide the parent toward a valued life. The third online meeting was devoted to sharing experiences and lessons learned, the need for additional practice, and take-home messages for the parent. 2.2.2. Description of the ACT self-help material The parents in the self-help group (n =53) received a 20-page selfhelp booklet highlighting an ACT approach, which was sent by regular mail (see Table 3). In addition, they were encouraged to visit a website called Oivamieli (www.oivamieli.fi), a web page accessible to the public that included a series of ACT-based exercises (to read or listen) on topics such as values, acceptance, and the present moment (see Table 4). The Fig. 1. Participant flow. P. Lappalainen et al. Internet Interventions 24 (2021) 100382 4 participants in this group did not have access to the same website as the supported iACT group, neither did they receive professional support during the intervention period. 2.3. Measures 2.3.1. Symptom and quality of life measures The main outcome measure was the Shirom-Melamed Burnout Questionnaire (SMBQ, Lundgren-Nilsson et al., 2012; Melamed et al., 1999). The questionnaire contains 22 statements measuring different aspects of burnout. The measure comprises of four subscales: emotional exhaustion, physical fatigue, listlessness, and tension. The statements are answered on a 7-point Likert scale, where the response options range from 1 =“almost never” to 7 =“almost always.” Based on the SMBQ, burnout can be divided into low (2.75–3.74), high (3.75–4.46), and pathological (≥4.47) levels. The validity and reliability of the SMBQ has been found to be good in previous studies (Lundgren-Nilsson et al., 2012). In this study, Cronbach’s alpha at baseline was excellent, 0.92. Depressive symptoms were measured by the Patient Health Questionnaire-9 (PHQ-9) questionnaire (Kroenke et al., 2001), which is a dual-purpose instrument for making preliminary diagnoses and assessing the severity of depressive disorders. The PHQ-9 contains nine Table 1 Characteristics of parents. Parental variables All (n = 110) iACT (n = 57) Self-help ACT (n = 53) Age [mean (SD)] 40.1 (6.69) range 27–55 38.9 (6.9) range 27–53 41.5 (6.3)* range 27–55 Sex Female 103 (93.6%) 54 (94.7%) 49 (92.5%) Male 7 (6.4%) 3 (5.3%) 4 (7.5%) Marital status Married or living together 91 (82.7%) 48 (84.2%) 43 (81.1%) Single parent 19 (17.3%) 9 (15.8%) 10 (18.9%) Education a Primary level 3 (2.7%) 3 (5.3%) – Secondary level 31 (28.2%) 13 (22.8%) 18 (34%) University level 76 (69.1%) 41 (71.9%) 35 (66%) Working status ** Full time 48 (44%) 23 (41.1%) 25 (47.2%) Part time 25 (22.9%) 12 (21.4%) 13 (24.5%) Unemployed 5 (4.6%) 2 (3.6%) 3 (5.7%) Student 11 (10.1%) 6 (10.7%) 5 (9.4%) Sick leave 2 (1.8%) 2 (3.6%) – Caregiver/maternity leave 18 (16.5%) 11 (19.6%) 7 (13.2%) Has own disease or diagnosis 49 (44.5%) 23 (40.4%) 26 (49.1%) Child’s age [mean, (SD)] 9.0 (4.29) range 0.8–17 8.5 (4.3) range 0.8–17 9.6 (4.2) range 1–17 Child’s gender Male 76 (69.1%) 41 (71.9%) 35 (66%) Female 35 (30.9%) 16 (28.1%) 18 (34%) Child’s condition *** Developmental disability 36 (33.3%) 16 (29.1%) 20 (37.7%) Autism spectrum disorder 20 (18.5%) 14 (25.5%) 6 (11.3%) ADHD 9 (8.3%) 5 (9.1%) 4 (7.5%) Diabetes 8 (7.4%) 3 (5.5%) 5 (9.4%) Gastrointestinal diseases 8 (7.4%) 3 (5.5%) 5 (9.4%) Coronary diseases 6 (5.6%) 3 (5.5%) 3 (5.7%) Muscle diseases 5 (4.6%) 4 (7.3%) 1 (1.9%) Other 16 (14.8%) 7 (12.7%) 9 (17%) a Primary >9 years; secondary 9–12 years; university 12 >years (university, college, etc.) * Missing information: n =52. ** Missing information: n =56. *** Missing information: n =55. Table 2 Content of the web-based program (supported iACT group). Modules (ACT processes) Program content Introduction Welcome to the program Module 1: What is important to you? (Values) Text, video on “Important to you” experiential exercises: Curtains in your life’s window A meaningful moment 90th birthday party Assignment, weeks 1–2 Reflect on your life values, and formulate them. Choose one or more areas of life that feel/s right for you right now. Do what it takes! Module 2: Meaningful actions! (Value-based actions) Text, experiential exercises: Do it now! Relieving stress Focus on breathing A place to rest Tired Passengers on the bus But-stories On your life’s journey Assignment, weeks 3–4 Choose at least two important life values, and make a plan. Write down your plan, as in the example below: My life value (what matters to me) is: … What do I choose to do? Module 3: Present moment (Present moment) Text, video, experiential exercises: Senses Breathing exercise Mindful listening Mindfulness in daily life Mindfully washing the dishes Mindful eating Mindful walking Mindful body Assignment, weeks 5–6 1. Apply the SOAL (stop, observe, accept, and let go) method in your everyday life. 2. Choose a daily chore, and do it mindfully. 3. How can you influence your well-being? What actions are you willing to choose during the next two weeks Module 4: Distance to thoughts and feelings (Defusion) Text, video, experiential exercises: The little man Leaves in the stream The sky and the weather Observer Assignment, weeks 7–8 1. Stop once in a while, and focus on what you are feeling or thinking. Label your thoughts, feelings, and body sensations as they appear: Now, I’ve got the feeling that I’m sad Now, I’ve got the feeling to … Now, the memory that comes to me … Now, it feels like this in the body …. 2. Train distance to your thoughts by asking yourself Who is in charge here; me or my thoughts? Or Is this thought helpful? Has it helped me to live a good life? Module 5: Acceptance (Acceptance and selfcompassion) Text, video, experiential exercises: K¨ oydenveto Hillomunkit Broken machine Let it be Compassionate hand Stone on the beach A good friend The guesthouse Assignment, weeks 9–10 1. Is there anything in your life that you need to learn to accept? Apply the general model of acceptance to what you have difficulty accepting: 1. Note and describe the present moment. 2. Accept what you note. 3. Accept your painful and unpleasant thoughts and feelings. 4. Do not avoid. 2. Be compassionate with yourself and others. Which small acts of kindness can improve your wellbeing and that of others? Closing words The journey continues… P. Lappalainen et al. Internet Interventions 24 (2021) 100382 5 items, which are answered on a 4-point Likert scale, where the answer options range from 0 =“not at all” and 3 =“almost every day.” The total score on the PHQ-9 ranges from 0 to 27, with a higher score indicating more depressive symptoms. PHQ-9 scores of 5 (5–9), 10 (10–14), 15 (15–19), and 20 (20–27) represent mild, moderate, moderately severe, and severe depression, respectively. The validity and reliability of the PHQ-9 meter have been found to be good (Kroenke et al., 2001). In this study, Cronbach’s alpha indicated high reliability, 0.78. Health-related quality of life was measured with the Finnish version of the SF-36 measure (Dempster and Donnelly, 2000; Ware and Sherbourne, 1992) RAND-36 (Aalto et al., 1999). The RAND-36 measures the following eight dimensions: physical functioning, role limitations due to physical health problems, bodily pain, social functioning, general mental health, psychological distress and well-being, role limitations due to emotional problems, and vitality, energy. Each dimension is measured with Likert scales and “Yes” or “No” questions and forms its own entity. The original numeric values are recoded to give a value of 0–100. Regarding the scoring, index values were calculated, which were obtained by summing the number of questions for each dimension divided by the number of questions answered, with high scores indicating good health and quality of life. The validity and reliability of the RAND-36 have been found to be good in the Finnish population (Aalto et al., 1999). Cronbach’s alpha for the RAND-36 in this study was high, 0.89. 2.3.2. Process measures Psychological flexibility was measured with the Comprehensive assessment of ACT processes (CompACT; Francis et al., 2016), a 23-item questionnaire that includes three subscales: openness to experience (CompACT-OE), behavioral awareness (CompACT-BA), and valued action (CompACT-VA). The items are answered on a 7-point Likert scale ranging from 0 (strongly disagree) to 6 (strongly agree), with higher scores representing greater levels of psychological flexibility (openness, awareness, activation). The total score of the CompACT ranges from 0 to 138, with CompACT-OE ranging from 0 to 60, CompACT-BA ranging from 0 to 30, and CompACT-VA ranging from 0 to 48. In this study, the CompACT showed good internal consistency for the total score ( α = 0.83), α =0.79 for CompACT-OE, α =0.76 for CompACT-VA, and α = 0.80 for CompACT-BA. Mindfulness skills were assessed using the Five Facet Mindfulness Questionnaire (FFMQ, Baer et al., 2006), which measures five aspects of mindfulness: observing (noticing internal and external experiences), describing (naming and labeling internal experiences), acting with awareness (paying attentions to one’s activities in the present moment), non-judging of inner experiences (having a non-evaluative stance toward inner experiences), non-reactivity to inner experiences (ability to let thoughts and feelings to come and go without struggle). The scale comprises 39 items rated on a 5-point Likert-type scale ranging from 1 (never or very rarely true) to 5 (very often or always true), with higher scores indicating higher levels of mindfulness. The structure, reliability, and validity of FFMQ have been found to be good (Baer et al., 2006). In this study, the FFMQ showed high reliability for the total score ( α = 0.87), with α =0.74 for observing, α =0.91 for describing, α =0.79 for non-judging and non-reacting, and α =0.82 for acting with awareness. Table 3 Content of self-help booklet (self-help ACT group). Chapter Content Introduction to the value, acceptance, and mindfulness-based approach (2 pages) The mindful mind (4 pages) Being present: Be present; live in this moment. Thoughts as traps and double-edged sword. The skill of being present The wise mind (6 pages) Observing mind: Take a flexible stance on your thoughts. You are not your thoughts (3 pages). Accepting mind: What is acceptance? Accept your thoughts and emotions. Accept what you cannot change (3 pages). The valuing mind (3 pages) What are values? Reflect on your own values and act accordingly. What is important to you? Healthy body (2 pages) Relaxation Exercise Mindful eating Table 4 Content of the self-help ACT material on the web. Process MP3 exercises 1. The mind Present moment Focus on what you are doing. Introduction Follow your breath Mindful sitting Connect with sounds around you Drop the anchor Mindful chores Summary Observing You are not the same as your mind Introduction Observer The bird Leaves in the stream Terrier thoughts The little man I have a thought Summary Acceptance Accept what you cannot change Introduction Warm jelly doughnuts Tug-of-war Stone on the beach Broken machine Compassionate hand Summary 2. Values and choices Values Find what is important to you Introduction Your life now Epitaph Important in life Value domains Six months to live Notice the little things in your life Summary Value-based actions Act according to your values Introduction Set a goal Value-based action Do it now! Mindful listening Appreciate a loved one Passengers in the bus I cannot do it Summary 3. Body wellness The sensing body Feel the calm and peace in your body Introduction Breath in, breath out Release your tensions Be a mountain Accept your body Massage the pain away Relaxation exercise Summary Mindful eating Nutrition to your body, mind and heart Introduction Mindful eating Who is hungry? Individual eating rhythm The value of food choices Barriers to eating choices Mindful groceries Summary The moving body Take care of your body Introduction The value of exercising Barriers to exercising Choose the stairs! Dance your troubles away Mindful exercise Summary P. Lappalainen et al. Internet Interventions 24 (2021) 100382 6 2.3.3. User activity and intervention satisfaction In addition to the standardized instruments, we collected data on user activity and participant satisfaction. In the supported web-based group, user activity was determined as a percentage of the Uupu web program, which each parent had completed based on the log data. In the self-help group, the usage of the publicly accessible Oivamieli website could not be tracked, and instead, the parents were asked to report with “Yes” or “No” whether they had used the self-help booklet and the website, including the MP3 exercises. Overall satisfaction with the support received (web-based group) was measured using 0–10 visual rating scales (0 =not at all satisfied with the support model, 10 =very satisfied with the support model). Both groups of parents were also asked to evaluate with “Yes” or “No” whether they would recommend this kind of support (web +psychologist support or the self-help material) to other parents. 2.4. Statistical analysis Latent change score (LCS) modelling (Wald test) was used to analyze the Group ×Time interaction, i.e., whether the groups changed differently from the preto post-measurement (Klopack and Wickrama, 2020). LCS modelling applies structural equation modelling (SEM) and is equivalent to repeated measures ANOVA. It accounts for missing values at random (MAR) and includes all available data. The parameters were estimated using the full maximum likelihood (MLF) method (robust maximum likelihood (MLR) estimation in Mplus) (Muth´ en and Muth´ en, 1998–2015). Effect sizes (ESs) are reported using Cohen’s d. This was calculated from the baseline to post-measurements within, and corrected between, groups to estimate the ESs using estimated values. ES was calculated by dividing the difference between the means by the average of the standard deviations. When calculating the between-group ES at postmeasurement, we took into account the possible pre-measurement difference between the groups. An ES of 0.2 is considered small, 0.5 medium or moderate, and 0.8 large (Roth and Fonagy, 2005). The power analysis suggested that in the current study we needed a sample size of 128 participants given the effect size (d) =0.25, α =0.05, power =0.80. Thus, the between group effect size d >0.20 was the lowest effect size of interest, and our sample n =110 was large enough to detect the difference between the groups. 3. Results 3.1. Baseline psychological distress and dropout rates Regarding burnout (SMBQ), the majority of the parents in the iACT group were classified as experiencing pathological levels of burnout (≥4.47, n =34, 59.6%); 12 parents (21.1%) had high burnout (3.75–4.46); and 11 parents (19.3%) experienced mild burnout (2.75–3.74). In the self-help ACT group, the equivalent values were 35 (66%), 13 (24.5%), and 5 (9.4%). With regard to depression (PHQ-9), mild depression was reported by 21 (36.8%) parents in the iACT and 26 (49.1%) in the self-help ACT group; moderate depression was reported by 18 parents (31.6%) in the iACT and 14 (26.4%) in the self-help group; moderately severe to severe depression was experienced by 10 parents (17.6%) in the iACT group and six parents (11.4%) in the self-help ACT group. Out of the 110 parents who participated in the pre-measurement, a total of 28 (25.5%) dropped out at the post-measurement, leaving 42 parents in the supported iACT group (dropout 26.3%) and 40 in the selfhelp ACT (dropout 24.53%) at post-measurement. Fig. 1 presents the flow of the study. 3.2. Symptom, quality of life, and process measures The effects of the two interventions on the symptom and process measures are summarized in Table 5, and those regarding the quality of life measures are summarized in Table 6. For the main outcome measure, symptoms of burnout (SMBQ), there was no significant interaction effect. Instead, the groups showed relatively similar and significant within group decreases (p <0.01) in burnout symptoms (SMBQ), and the within group effect sizes were medium in both groups (iACT: d = 0.63–0.99; self-help ACT: d =0.48–1.10). In terms of depressive symptoms, a comparison revealed that the two groups changed differently (PHQ-9; a significant interaction effect, p =0.021), with a larger decrease in the supported iACT group. For the process measures, we observed the following significant interaction effects: the supported iACT produced larger changes for the total CompACT score (p =0.039) and its subscale valued action (p =0.039) as well as for the total mindfulness score (FFMQ; p =0.047) and its subscale describing (p =0.007). These significant interactions showed medium-sized differences (d = 0.49–0.64) in favor of the supported iACT intervention. For healthrelated quality of life (RAND-36), both groups showed significant positive changes in vitality and general mental health (Table 5). The withingroup ESs indicated large changes in the iACT (d =0.94/0.84 for vitality and general mental health, respectively) and small to medium changes in the self-help ACT for vitality and mental health (d =0.42–0.55). However, the RAND subscale role limitations caused by emotional problems showed a significant medium-sized (d =0.58) difference between the groups in favor of the supported iACT (significant interaction effect, p = 0.021). An investigation of the betweenand within-group ESs revealed that the changes in several symptom, process, and quality of life measures were somewhat larger in the supported iACT than in the self-help ACT. 3.3. User activity In the supported iACT group, a great majority (69%, n =29/42) of the parents completed the whole program or most of it (76–100%); 11.9% (n =5) completed between 51 and 75%; and 19.1% completed 50% or less. Overall, approximately 80% of the parents completed more than 50% of the web intervention. Further, the log data indicated that approximately 62% had been using the web program over five to eight weeks (45.2%) or longer (16.7%, over 9–12 weeks). For the self-help ACT, no user activity data were available, but following the intervention, 62.5% of the parents reported that they had used the self-help booklet, while 32.5% had used the website with the ACT-based exercises. More precisely, 30% of the parents in the self-help ACT group (n = 12/40) had used both the self-help booklet and the website. 3.4. Intervention satisfaction The parents in the supported iACT gave a rating of 7.8 (scale 0–10) for their satisfaction with the overall support. In the self-help ACT, their satisfaction with the self-help material was rated 5.4, on average, using the same scale. The participants were also asked to evaluate with “Yes” or “No” whether they would recommend this kind of support (web + psychologist support) to other parents of children with chronic conditions. In the iACT group, 95% of the parents would recommend the intervention to other parents, whereas in the self-help ACT, 65% of them would recommend the self-help material to other parents. 4. Discussion Parents of children with chronic conditions have been found to experience physical and mental distress, which can have harmful consequences for the entire family. The aim of the present study was to examine two types of interventions with the purpose of supporting such families, namely the impact of a psychologist-supported iACT intervention and self-help ACT regarding burnout, depressive symptoms, quality of life, and psychological flexibility among parents of children with chronic conditions. With regard to the main outcome measure P. Lappalainen et al. Internet Interventions 24 (2021) 100382 7 burnout, the results showed that both the psychologist-supported iACT and the self-help ACT intervention reduced burnout symptoms. However, regarding depressive symptoms, psychological flexibility, mindfulness, and in one dimension of health-related quality of life, the supported iACT produced significantly larger improvements compared to the self-help ACT intervention. In these measures, the between group Table 5 Estimated sample statistics (mean ±standard deviation (SD)) and the effect of the intervention on outcomes. Supported iACT Self-help ACT Interaction p a d d Pre Post p b d e Pre Post p c d e SMBQ 4.51 ±0.88 3.57 ±1.08 <0.01 0.99 4.76 ±0.75 4.04 ±1.11 <0.01 0.77 0.27 0.29 Phys.fatigue 4.68 ±0.98 3.52 ±1.19 <0.01 0.84 4.82 ±0.81 3.99 ±1.15 <0.01 1.10 0.14 0.39 Listlessness 4.67 ±0.85 3.84 ±1.06 <0.01 0.89 5.02 ±0.95 4.20 ±1.28 <0.01 0.72 0.86 0.05 Tension 4.25 ±1.18 3.45 ±1.19 <0.01 0.68 4.60 ±0.99 4.04 ±1.15 <0.01 0.55 0.47 0.19 Emot.exhaust. 4.33 ±1.25 3.53 ±1.40 <0.01 0.63 4.56 ±1.04 4.00 ±1.44 <0.01 0.48 0.35 0.21 PHQ-9 9.61 ±4.97 6.61 ±5.13 <0.01 0.63 9.21 ±4.40 8.15 ±4.75 0.26 0.19 0.02 0.49 CompACT 77.20 ±18.65 91.61 ±18.16 <0.01 0.76 75.24 ±14.01 78.95 ±11.72 0.22 0.27 0.04 0.64 OE 32.59 ±11.25 40.18 ±10.20 <0.01 0.65 33.08 ±8.54 35.84 ±8.34 0.15 0.28 0.09 0.47 BA 13.75 ±6.17 15.43 ±5.97 0.16 0.30 12.53 ±5.25 12.21 ±3.20 0.74 0.07 0.18 0.37 VA 30.86 ±7.13 36.00 ±5.76 <0.01 0.74 29.63 ±6.44 30.90 ±4.99 0.18 0.25 0.04 0.49 FFMQ 125.72 ±16.62 138.07 ±17.69 <0.01 0.72 126.37 ±14.00 129.47 ±13.87 0.15 0.27 0.05 0.55 OBS 25.82 ±5.36 28.39 ±4.75 <0.01 0.49 26.41 ±4.92 26.63 ±5.32 0.31 0.19 0.18 0.30 DESC 28.46 ±5.92 31.29 ±7.09 <0.01 0.40 28.98 ±5.49 28.37 ±5.10 0.41 0.15 <0.01 0.59 AWA 22.75 ±5.12 24.43 ±5.25 0.04 0.34 21.55 ±4.06 22.55 ±3.70 0.35 0.25 0.74 0.17 NJ 28.04 ±7.34 31.11 ±7.39 0.02 0.41 28.29 ±6.53 29.90 ±6.05 0.12 0.28 0.39 0.20 NR 20.66 ±4.40 22.86 ±3.17 0.01 0.58 21.14 ±3.54 22.05 ±4.07 0.39 0.19 0.21 0.38 SMBQ =Shirom-Melamed Burnout Questionnaire. SMBQ Phys.fatigue =physical fatigue. SMBQ Emot.exhaust. =emotional exhaustion. PHQ-9 =Patient Health Questionnaire (depressive symptoms). CompACT =comprehensive assessment of ACT processes. CompACT-OE =openness to experiences. CompACT-BA =behavioral awareness. CompACT-VA =valued action. FFMQ =Five Facet Mindfulness Questionnaire. FFMQ-OBS =observing. FFMQ-DESC =describing. FFMQ-AWA =acting with awareness. FFMQ-NJ =non-judgement. FFMQ-NR =non-reactivity. a p-Value for differences in changes between the groups using estimated parameters (Latent Change Score, Wald test). Bold text indicates significant p-value <0.05. b p-Values of the change between pre and post within the supported iACT group. c p-Values of the change between pre and post within self-help ACT group. d Cohen’s d from baseline to post-measures between the supported iACT group and the self-help ACT group using estimated parameters. e Cohen’s d from baseline to post-measures within the group using estimated parameters. Table 6 Estimated sample statistics (mean ±standard deviation (SD)) and the effect of the intervention on Health-related quality of life (RAND-36). RAND-36 Supported iACT Self-help ACT Interaction p a d d Pre Post p b d e Pre Post p c d e Physical function. 85.61 ±16.61 86.19 ±20.41 0.05 0.13 85.00 ±14.01 86.83 ±14.35 0.90 0.02 0.33 0.13 Role limit./phys. health 75.00 ±10.69 77.98 ±33.27 0.27 0.14 61.79 ±37.49 62.81 ±36.68 0.97 0.01 0.56 0.12 Role limit./emot. probl. 39.18 ±35.39 61.11 ±37.03 <0.01 0.60 45.28 ±38.90 46.34 ±41.57 0.99 0.00 0.02 0.58 Vitality 34.91 ±14.49 50.00 ±19.24 <0.01 0.94 33.89 ±14.88 44.02 ±21.56 <0.01 0.55 0.17 0.38 Mental health 54.18 ±14.80 66.00 ±13.81 <0.01 0.84 51.93 ±14.27 57.76 ±16.03 0.01 0.42 0.10 0.39 Social functioning 58.77 ±22.70 69.05 ±27.45 <0.01 0.46 60.82 ±21.38 65.94 ±18.75 0.15 0.24 0.20 0.30 Bodily pain 67.50 ±26.04 72.08 ±28.08 0.06 0.22 66.70 ±22.10 71.95 ±20.69 0.12 0.26 0.99 0.02 Health perceptions 54.12 ±18.52 59.17 ±21.01 0.03 0.22 51.60 ±18.78 57.76 ±20.13 0.07 0.26 0.96 0.02 Physical function. =physical functioning. Role limitat./phys. health =role limitations due to physical health problems. Role limitat./emot. probl. =role limitations due to emotional problems. Vitality =vitality, energy. Mental health =general mental health covering psychological distress & well-being. Social functioning. Bodily pain. Health perceptions =general health perceptions. a p-Value for differences in changes between the groups using estimated parameters (latent change score, Wald test). Bold text indicates significant p-value <0.05. b p-Values of the change between pre and post within supported iACT group. c p-Values of the change between pre and post within self-help ACT group. d Cohen’s d from baseline to post-measures between the iACT group and the self-help ACT group using estimated parameters. e Cohen’s d from baseline to post-measures within the group using estimated parameters. P. Lappalainen et al. Internet Interventions 24 (2021) 100382 8 effect sizes showed medium-sized differences (d =0.49–0.64) in favor of the supported iACT intervention. In an examination of psychological flexibility (CompACT) and mindfulness skills (FFMQ) at the subscale level, the valued action and describing (labeling feelings, thoughts and experiences with words) subscales showed larger changes in the iACT group compared to the self-help ACT group. In an examination of the withingroup effect sizes, small or very small effects in mindfulness skills were generally observed in the ACT self-help group, however, larger changes were recorded in the supported iACT group. This may be due to the fact that only one third of the parents undergoing the self-help approach had undertaken the experiential exercises of teaching value-, acceptance and mindfulness skills, which may have reflected in the outcomes in these constructs. One explanation why no significant changes in other mindfulness and flexibility skills were detected in the supported iACT group could be that skills such as observing (noticing or attending to internal feelings and thoughts), acting with awareness (attending to what is happening in the present), and non-reacting (allowing emotions and thoughts to come and go) may require more practice and take more time to appear, which is similar to the argument regarding the subskill of non-judging (Hofer et al., 2018). The same may apply to other psychological flexibility skills measured by CompACT, such as openness to experience and behavioral awareness, which are very close, if not overlapping, constructs to mindfulness skills. The change in valued actions may be due to the fact, as a construct, it is easier to understand and to promote compared to mindfulness constructs, which require more practice. This, potentially, suggest that the parents – who were reminded by the psychologist – had more actively committed to exercises dealing with value work, and, consequently, may have engaged in value-based actions. Interestingly, both interventions had an impact on burnout symptoms. In the past years, research has been inconclusive in terms of whether or not burnout and depression are the same construct or not and whether burnout is a risk factor in developing depression (Kaschka et al., 2011; Koutsimani et al., 2019). According to Koutsimani et al. (2019), burnout and depression seem to be different constructs that share some common characteristics and may develop in parallel. Given that burnout is an antecedent of depression, a possible explanation could be that the self-help ACT intervention had started to have an impact on burnout, but had not yet affected depressive symptoms. An alternative explanation is that provided by Baer et al. (2012) who stated that before learning any mindfulness skills, other factors such as coming to a treatment group and sharing with fellow participants with similar problems may lead to small decreases in perceived stress. In this study, being allocated to a research group and received attention may have also induced changes in burnout in the self-help ACT group. As mindfulness and flexibility skills begin to develop, additional decreases in burnout may be expected (see Baer et al., 2012). With regard to the amount of support, our findings are consistent with previous research suggesting that unsupported treatments are less effective than treatments that incorporate support (e.g., Andersson and Cuijpers, 2009; Farrand and Woodford, 2013). The current study is also in line with meta-analyses that have compared guided Internet-based treatments with unsupported control conditions and indicated larger effect sizes in favor of guided Internet-based treatments in which a therapist or coach supports the client through the treatment program (between-group d =0.42–0.78; Andersson and Cuijpers, 2009; Richards and Richardson, 2012). At the same time, the results also indicate that pure self-help ACT may produce positive outcomes in parents. Despite of the differential design, this study can be compared with that of Hofer et al. (2018) which concluded that ACT delivered through a self-help book without contact from a therapist reduced stress and burnout for individuals among working population compared to those in a waitinglist control group. An important aspect was the satisfactory user activity in both intervention groups. In the supported web-intervention group, approximately 70% completed all or nearly all of the program, and only 5% (n =2) took very little advantage of the program (1–25% usage observed). Conversely, self-help material, which was provided with no support, was utilized by over 60% of the parents. These numbers can be regarded as encouraging. To attain even better adherence to the web intervention (iACT) and self-help material (self-help ACT), some parents may have needed more professional support, peer support, and/or emailreminders inviting them back to the program. Therefore, there is a need to examine in greater detail parents’ usage activity in terms of reflecting their engagement and patterns of usage in the current iACT and self-help ACT interventions. In addition, future studies should pay attention to remaining sensitive to parental needs and wishes for support. Support for families—as Coyne et al. (2020) suggest—should be provided in small, doable doses. Even when the dose is small, it can still make a difference (Coyne et al., 2020). The current interventions are one step in this direction, providing parents with flexible tools for managing their own distress and promoting self-care. The study by Sairanen et al. (2019) offers an interesting comparison to the current study as it applied a comparable web-based ACT intervention in a different country and language. Juxtaposing the results of the current study with those of the equivalent study by Sairanen et al. (2019), comparable changes were observed in the context of burnout (within-group effect size, d w =0.99 (FI) versus 0.70 (SWE)). Interestingly, these ESs were comparable to burnout decrease in the self-help group (d w =0.77). Although different measures for depression were used, the magnitude of the change was also similar in both iACT interventions (d w =0.63 (FI) versus 0.45 (SWE)). Moreover, regarding mindfulness, the impact was equal in the two interventions (d w =0.63 (FI) versus 0.72 (SWE)). In addition, the percentages for dropouts were equal at post-measurement, with 26% (n =15) in the current iACT intervention compared with 27% in the Swedish treatment group. Both percentages are in accordance with the findings of a meta-analysis of computer-based interventions for depression, showing dropout rates of 28% for therapist-support treatments (Richards and Richardson, 2012). Interestingly, the dropout rate for the self-help group (25%) was comparable to that of the supported iACT group and, therefore, was not consistent with dropout evidence for non-supported treatments (74%, Richards and Richardson, 2012). In conclusion, the current study confirmed the results observed in the previous study by Sairanen et al. (2019), which showed that a psychologist supported ACT intervention delivered via the Internet can decrease symptoms of burnout and depression among parents of children with chronic conditions. 4.1. Limitations Several limitations need to be considered in interpreting the results of this study. First, the recruitment method and data collection may limit the generalizability of the results. By conducting the recruitment through family and parent organizations, we may have excluded parents who were not actively engaged in these organizations and who may not use the Internet. Second, the current study reports only the preand post-intervention results. However, the ongoing follow-up period will determine whether the changes were maintained over a longer period of time. Third, the lack of a waiting-list control group (WLC) that did not undergo intervention suggests that we cannot draw firm conclusions from the results of the study, neither can we exclude that other factors might have impacted the results. It is possible that the improvements in the outcomes of the two groups were caused by repeated measures or spontaneous recovery effects. However, changes in a WLC group were verified in the Swedish study by Sairanen et al., 2019, showing no or minimal changes in the no-intervention group, with a web-based ACT intervention showing significantly larger improvements. Fourth, all the assessment scales were self-report questionnaires. Ecological momentary assessment through smartphones could have given additional information on the impact of the study. Fifth, given the large number of significant test performed, and the increased risk of type I and type II errors (and in purpose to ensure that our conclusions were reliable), the P. Lappalainen et al.