Physical Fitness and Brain - Interventional Study (PHYSBI)

Overall objective To examine whether physical exercise reduces anxiety symptoms, improves cognition and decreases need of sick leave in patients with anxiety disorders in primary care.

Background Mental illness is the most common reason for unemployment in people of working age and this is a growing patient group in Swedish primary care today. A survey by the European Commission showed that 17% of the Swedish population had sought help over the past year for psychiatric problems. These figures are above average for European countries and are increasing. Sweden's costs for mental illness is around MSEK 70,000 annually - an increase of MSEK 20,000 over ten years - and the proportion of sick leave days due to a psychiatric diagnosis is currently 40%. Mental illness reduces quality of life substantially. For example, 18-year-olds with anxiety disorders have increased risk of future marginalization and have a two-fold increase in mortality not only due to suicide but several different death causes, emphasizing the importance of finding effective treatment strategies.

The ability of the adult brain to change both structurally and functionally is called brain plasticity. Both in humans and in animal models physical exercise has a major impact on brain plasticity. Physical exercise can increase neurogenesis and improve memory functions in rats, and circulating insulin-like growth factor I (IGF-I) can promote this effect. Brain-derived neurotrophic factor (BDNF), and vascular endothelial growth factor (VEGF) are also involved. In addition, there is evidence that physical activity modifies neuronal activity, increases the formation of synapses, and reduces inflammation markers, which may be correlated with neuroprotection.

The investigators have previously demonstrated an association between physical fitness and anxiety disorders, suggesting positive preventive benefits of physical exercise. The beneficial effects of exercise interventions have to date mainly been demonstrated in persons with depression.

The specific aims of this project are:

1. To test a 12 week physical exercise intervention for persons treated for anxiety disorders within primary care. The exposure of interest is intensity of physical exercise; outcomes include anxiety symptom burden, cognitive ability and sickness leave.
2. To gain knowledge regarding potential mechanisms by comparing serum levels of specific hormones (characterized and associated with brain plasticity in animal models) before and after different intensities of exercise.

Relationship to the research frontline In contrast to the literature on depression there are very few clinical studies on physical exercise as an intervention for anxiety. All six trials performed hitherto were identified as being at a high risk of bias. Higher quality trials are thus required. There is a need to investigate the role of intensity, volume, frequency and type of exercise on anxiety symptoms, cognitive function and working ability in people with anxiety disorders. Also, the putative mechanisms behind the potential effects on anxiety symptoms are not elucidated and deserve attention.

A novel feature of this study is the examination of different intensities of physical fitness-improving exercise in patients with anxiety disorders. This has not previously been performed and would yield valuable insight into dose-response associations between physical fitness and effects on the brain. Objective measurements of physical fitness and cognition are a great strength of the proposed study. Analyzing specific hormones and cytokines before and after the intervention, in relation to symptoms and functional data, may provide additional information about mechanistic aspects, treatment effect and prognosis. The proposed research also offers opportunities to develop effective treatment protocols to reduce the costs of sick-leave, an area of increasing interest world-wide.

In summary, the proposed study has the potential to provide novel and important contributions to existing scientific knowledge regarding how physical exercise influences anxiety, cognitive function and working ability in patients with anxiety disorders in primary care.

Work plan The study cohort will be patients diagnosed with anxiety disorders at six primary care centres in Gothenburg, Southern Bohuslän and Halland including patients with panic syndrome (F41.0), generalized anxiety (F41.1), mixed anxiety- and depression conditions (F41.2 and F41.3), as well as anxiety UNS (F41.9). Diagnoses and comorbidities will be determined by the study psychiatrist using the M.I.N.I (the Mini International Neuropsychiatric Interview, Swedish version 6.0.0d DSM-IV). Intervention and follow-ups will take place at Primary Care Rehab.

The investigators will exclude patients from the baseline who have physical difficulties in performing a physical exercise program, pathological ECG, previous psychiatric illnesses, ongoing abuse, increased risk of suicide as determined by the GP, those under 18, pregnant women and those over 65 years of age. An additional exclusion criterion will be physical activity level at baseline exceeding one exercise occasion per week. Both people with and without ongoing treatment with psychoactive medication will be included. Supportive contact for example with a health center nurse is not an exclusion criterion but persons with ongoing psychotherapy will not be included unless the participants, as well as the GP, are willing to take a break in therapy for the duration of the 12 week study.

Recruitment takes place at primary care centres and thus represents the individuals who seek help in primary care for anxiety. Potential participants are contacted by a study psychiatrist who has no healthcare provider relationship to the patient. Intervention and follow-ups will take place at Primary Care Rehab. Each individual must sign a statement of informed consent.

At baseline, 12 weeks and 1 year, the following data will be collected by the study physician:

• MINI International Neuropsychiatric Interview (Swedish version 6.0.0d DSM-IV)
• Use of psychoactive prescription drugs including dosage and duration
• Sick leave
• Blood pressure
• BMI
• Self-report questions about alcohol (AUDIT self-assessment scale), smoking and diet
• Established psychiatric self-assessment scales: Beck Anxiety Inventory (BAI), Montgomery Åsberg Depression Rating Scale (MADRAS) and the quality of life instrument EQ-5D.
• Test Design Fluency, which is part of the D-Kefs battery of tests and is well known and used by many psychologists within education and psychiatry. This is a standardized test, carried out within a few minutes that measures executive abilities such as creativity, response inhibition and cognitive flexibility.
• Objective measurements of physical fitness in the form of estimated maximal oxygen uptake capacity (VO2max) using Åstrand's submaximal bicycle test.
• Blood samples will be taken, by a nurse at the health center, to analyze the potential correlation of the following hormones/cytokines with symptoms and function data: S-IGF-I, S-BDNF, S-VEGF and high sensitivity CRP. The blood samples are stored in a separate, locked -70 freezer at c-lab Sahlgrenska. IVO has been notified of the establishment of a biobank (No 946).

Patients will be randomized into three groups: 1) Intervention I: 12 week exercise program with low-intensity fitness training 3 times per week. 2) Intervention II: 12 week exercise program with moderate to high-intensity fitness training 3 times per week. 3) Control group. The physiotherapists at Primary Care Rehab will perform the submaximal bicycle test and personally design the exercise program in the gym. The patients will have a one-to-one session with the physiotherapists on 2 occasions and afterwards several people will exercise at the same time in the gym.

A statistical power analysis has been performed by statistician Linus Schiöler. Based on other studies, an effect size of 0.5 difference between the groups is clinically relevant, which equates to a 5 point difference between the groups using the BAI scale. Standard deviations from a previous study using BAI as the outcome measurement were in the interval of 5-10. SD 9 was used in the calculations. Based on these assumptions, at least 60 patients in each of the treatment arms should be included to reach 80% power at the 5% significance level. Health-economic analyses will be performed using primary data generated within the project concerning health-related quality of life effects caused by the intervention and information about healthcare resources utilized. A health-economic model that depicts the clinical trial will be constructed and used for computing cost-utility measures. Moreover, the model will be constructed for performing extensive probabilistic sensitivity analyses.

Ethical considerations The Ethics Committee of the University of Gothenburg approved the study: PHYSBI intervention study: EPN Göteborg Dnr: 300-16. The control group will have a physiotherapy session once and will be given general advice about physical activity. There will be a possibility to individualize treatments i.e. the same intensity but different exercises to better fit the needs and preferences of different individuals. Coding and anonymization of data occur after data collection is complete.

Collaborations The principal investigator for this project Ass Prof Maria Åberg is the chair of the steering committee for Physical Fitness and Brain Epidemiological studies (PHYSBE) which is a multidisciplinary umbrella organization for several projects. Prof Kjell Torén from Occupational and Environmental Medicine provides epidemiological competence. Prof Margda Waern provides psychiatric expertise. Physiotherapist Louise Danielsson and Prof Kristian Bolin at the Department of Economics and Statistics have experience of clinical intervention studies.

The results will be published in international renowned peer-reviewed research journals.