Personalized Physical Activity Coaching: A Machine Learning Approach.

Living a sedentary lifestyle is one of the major causes of numerous health problems. To encourage employees to lead a less sedentary life, the Hanze University started a health promotion program. One of the interventions in the program was the use of an activity tracker to record participants' daily step count. The daily step count served as input for a fortnightly coaching session. In this paper, we investigate the possibility of automating part of the coaching procedure on physical activity by providing personalized feedback throughout the day on a participant's progress in achieving a personal step goal. The gathered step count data was used to train eight different machine learning algorithms to make hourly estimations of the probability of achieving a personalized, daily steps threshold. In 80% of the individual cases, the Random Forest algorithm was the best performing algorithm (mean accuracy = 0.93, range = 0.88-0.99, and mean F1-score = 0.90, range = 0.87-0.94). To demonstrate the practical usefulness of these models, we developed a proof-of-concept Web application that provides personalized feedback about whether a participant is expected to reach his or her daily threshold. We argue that the use of machine learning could become an invaluable asset in the process of automated personalized coaching. The individualized algorithms allow for predicting physical activity during the day and provides the possibility to intervene in time.

Do activity monitors increase physical activity in adults with overweight or obesity? A systematic review and meta-analysis.

OBJECTIVE: To systematically assess contemporary knowledge regarding behavioral physical activity interventions including an activity monitor (BPAI+) in adults with overweight or obesity. METHODS: PubMed/MEDLINE, Embase, CINAHL, PsycINFO, CENTRAL, and PEDro were searched for eligible full-text articles up to 1 July 2015. Studies eligible for inclusion were (randomized) controlled trials describing physical activity outcomes in adults with overweight or obesity. Methodological quality was independently assessed employing the Cochrane Collaboration's tool for risk of bias. RESULTS: Fourteen studies (1,157 participants) were included for systematic review and 11 for meta-analysis. A positive trend in BPAI+ effects on several measures of physical activity was ascertained compared with both wait list or usual care and behavioral physical activity interventions without an activity monitor (BPAI-). No convincing evidence of BPAI+ effectiveness on weight loss was found compared with BPAI-. CONCLUSIONS: Behavioral physical activity interventions with an activity monitor increase physical activity in adults with overweight or obesity. Also, adding an activity monitor to behavioral physical activity interventions appears to increase the effect on physical activity, although current evidence has not yet provided conclusive evidence for its effectiveness.

Written pain neuroscience education in fibromyalgia: a multicenter randomized controlled trial.

Mounting evidence supports the use of face-to-face pain neuroscience education for the treatment of chronic pain patients. This study aimed at examining whether written education about pain neuroscience improves illness perceptions, catastrophizing, and health status in patients with fibromyalgia. A double-blind, multicenter randomized controlled clinical trial with 6-month follow-up was conducted. Patients with FM (n = 114) that consented to participate were randomly allocated to receive either written pain neuroscience education or written relaxation training. Written pain neuroscience education comprised of a booklet with pain neuroscience education plus a telephone call to clarify any difficulties; the relaxation group received a booklet with relaxation education and a telephone call. The revised illness perception questionnaire, Pain Catastrophizing Scale, and fibromyalgia impact questionnaire were used as outcome measures. Both patients and assessors were blinded. Repeated-measures analyses with last observation carried forward principle were performed. Cohen's d effect sizes (ES) were calculated for all within-group changes and between-group differences. The results reveal that written pain neuroscience education does not change the impact of FM on daily life, catastrophizing, or perceived symptoms of patients with FM. Compared with written relaxation training, written pain neuroscience education improved beliefs in a chronic timeline of FM (P = 0.03; ES = 0.50), but it does not impact upon other domains of illness perceptions. Compared with written relaxation training, written pain neuroscience education slightly improved illness perceptions of patients with FM, but it did not impart clinically meaningful effects on pain, catastrophizing, or the impact of FM on daily life. Face-to-face sessions of pain neuroscience education are required to change inappropriate cognitions and perceived health in patients with FM.

Not throwing out the baby with the bathwater: lessons from the Fibromyalgia Impact Questionnaire.

The Fibromyalgia Impact Questionnaire (FIQ) is the most frequently used questionnaire in patients with fibromyalgia in the last 20 years. Recently, a revised version of the FIQ has been published. In this study, we examined the factor structure of the original version using explorative and confirmative factor analyses in a representative group of about 500 Dutch and Belgian patients with fibromyalgia, in which the work-related item was excluded. The FIQ consisting of a three-factor structure with a functional domain (ten items), physical symptom domain (six items), and mental symptom domain (two items), which is the most accurate. From the data presented, it is concluded that the FIQ is a solid, worldwide-used questionnaire with a history of 20 years consisting of three domains. When constructing a revised FIQ, the results of this study can be incorporated.

Do illness perceptions of people with chronic low back pain differ from people without chronic low back pain?

OBJECTIVES: To determine why some people develop chronic low back pain, and whether illness perceptions are an important risk factor in the transition from acute to chronic low back pain. DESIGN: Cross-sectional study. PARTICIPANTS: Four hundred and two members of the general Dutch population, with and without chronic low back pain. MAIN OUTCOME MEASURES: Sociodemographics and the translated version of the illness perception questionnaire-revised, adapted for back pain. RESULTS: Of the sample, 115 (29%) individuals had chronic low back pain (>6 months) and 287 (71%) did not have chronic low back pain. Many of the participants with chronic low back pain believed that one 'wrong' movement can potentially lead to more severe problems, and that X-rays or computer tomography scans can determine the cause of the pain. Many of the participants with chronic low back pain did not perceive a relationship between psychosocial factors and low back pain. CONCLUSIONS: Illness perceptions differed between individuals with and without chronic low back pain. In the subacute phase, healthcare professionals could assess illness perceptions and, if necessary, incorporate them into the management of patients with low back pain.

Patients with more severe symptoms benefit the most from an intensive multimodal programme in patients with fibromyalgia.

PURPOSE: Patients with fibromyalgia (FM) experience symptoms over a long period of time impacting their quality of life (QoL). Patients are often treated in multimodal programmes that combine physical and cognitive treatment modalities. Purpose of this study was to identify prognostic factors of effectiveness of a multimodal programme. METHOD: A prospective study was performed with a group of 87 patients with FM who had participated in a multimodal programme. The Revised Illness Perception Questionnaire (IPQ) and the Fibromyalgia Impact Questionnaire (FIQ) were used. Criterion for clinically relevant improvement was a decline in total FIQ score of 12.5 points or more after the treatment programme. Investigated determinants of improvement of QoL were patient characteristics, illness perceptions (IP) and QoL at baseline. RESULTS: QoL of 34 patients with FM made a clinically relevant improvement after the programme. There was no difference in age, number of years with pain, number of years diagnosed or IP compared to the group that did not improve. The group of patients with an improved QoL after the programme reported severe impact on daily living, highest intensity of pain and most depression at baseline. CONCLUSIONS: Total FIQ score on QoL, intensity of pain, morning tiredness and depression can be used as prognostic factors to pre-select patients with FM for a multimodal treatment. IP were not adequate to predict treatment outcome. An intensive multimodal programme seemed most suitable for patients with severe symptoms and limitations.

Illness perceptions in patients with fibromyalgia and their relationship to quality of life and catastrophizing.

OBJECTIVE: In the last decade, illness perceptions have been identified as important in the treatment of fibromyalgia (FM). The aim of the present study was to examine illness perceptions and use of the revised Illness Perception Questionnaire in patients with FM (IPQ-R-FM) and their relationship to quality of life and catastrophizing. METHODS: A domain with specific causal attributions related to FM was added to the IPQ-R-FM. The psychometric properties of the IPQ-R-FM dimensions and attribution scales were examined. The causal domain, in which patients describe the most important perceived causes for their FM, was analyzed. To analyze the relationship with quality of life and catastrophizing, the Fibromyalgia Impact Questionnaire and the Pain Catastrophizing Scale were used. RESULTS: Fifty-one outpatients completed the questionnaires on 2 occasions, 3 weeks apart. FM was considered to be chronic and to have serious consequences; patients perceived little personal control and did not expect medical treatment to be effective. The psychometric properties of the IPQ-R-FM were found to be adequate. Patients most frequently attributed the causes of FM to an external somatic source (58%). Quality of life was related to experiencing more consequences attributable to FM. Catastrophizing was related to a limited understanding of the symptoms of FM, the more cyclical nature of FM, and an emotional representation. CONCLUSION: The IPQ-R-FM is a useful tool to assess illness perceptions in patients with FM. Illness perceptions are related to quality of life and catastrophizing; therefore, it seems important to assess and integrate illness perceptions into the management of patients with FM.