University of Wisconsin Meditation & Exercise Cold Study

Acute respiratory infection (ARI), including common cold and influenza, is a leading cause of morbidity and mortality, and has a major economic impact. Both mental and physical health are linked to ARI burden. For example, people who report more negative emotion and higher stress are more likely to get ARI. Exercise affects the immune system, improves physical and mental health, and may protect against ARI illness. Mindfulness meditation reduces perceived stress, influences the immune system, and may protect against ARI. The investigators' own recent NCCAM-funded trial randomized 154 people to 3 groups: 1) meditation, 2) moderate intensity exercise, or 3) wait-list control. For the 149 people followed to study completion, there were 40 ARI episodes and 453 days of illness in the control group, 27 episodes and 257 days of ARI illness in the meditation group, and 26 episodes and 241 days of ARI illness in the exercise group. Corresponding reductions in ARI-related work days lost to ARI illness were observed. The proposed research will build upon these findings with refined methodology in a larger sample to: A) determine whether these findings are replicable, and B) investigate potential explanatory pathways.

This research will use state-of-the-art randomized controlled trial (RCT) methodology to assess potential effects of meditation or exercise on ARI outcomes. This will be a 5-year project, with 4 yearly cohorts of n=99 per cohort randomized into 3 groups of n=33 each. Assuming 9% loss to follow-up, the final sample size will be n=360 study participants, with n=120 in each comparison group. Participants will be randomized to 1 of 3 groups: 1) an 8-week training program in mindfulness meditation, 2) an attention, duration and location-matched program in progressive exercise, or 3) a non-interventional wait-list control group. Each cohort will be observed for 8 months comprising the annual cold and flu season. Enrollment, randomization and study interventions will begin in September of each year. Participants will be monitored by weekly self-report through May. Summers will be used for data cleaning, preliminary analyses, and for recruiting the next year's cohort.

The primary goal of this project is to determine whether behavioral training in mindfulness meditation or moderate intensity sustained exercise can lead to reductions in acute respiratory infection (ARI) illness, such as common cold and influenza like illness. The primary outcome will be severity-weighted total days of ARI illness (global severity), calculated as trapezoidal approximation to area under the time severity curve during ARI illness, with severity assessed once daily using self-reports on the validated Wisconsin Upper Respiratory Symptom Survey (WURSS-24).

Computer-assisted weekly monitoring will assure that ARI illness episodes are detected, and will serve to document secondary outcomes, including health care utilization and work or school absenteeism. Visits to health care facilities and time lost from work and school will be documented, then classified as ARI-related (or not) by personnel blinded to allocation. Questionnaire measures assessing perceived stress, self-efficacy, sleep quality, depression, and general mental and physical health will also be analyzed as secondary outcomes. Degree of stress reduction, mindfulness, positive and negative emotion, social support, self-efficacy and sleep quality will be analyzed as potential mediators of effects of interventions on outcomes.

Laboratory assessed objective measures will primarily serve to corroborate self-reports of disease severity, but will also be analyzed as potential mediators of effects of behavioral interventions on ARI illness incidence, duration, and severity. As potential mediators, pro-inflammatory cytokines (CRP, IL-6, IL-8, IP-10) will be assessed as change from baseline to one month after the 8 week behavioral interventions finish. These will serve as indicators of a pro-inflammatory state. Repeating these assays 3 months later will assess whether pro-inflammatory changes from baseline will be sustained. Cytokines from samples taken during ARI illness will be assessed as corroborating biomarkers of disease severity. Identification of viral agents using multiplex PCR will also serve to corroborate ARI self-reports.

Self-report measures of perceived stress, positive and negative emotion, self-efficacy, social support, sleep quality, mindfulness, and general mental and physical health will be used to assess potential pathways through which interventions may exert influence on primary outcomes. Potential mediating effects of psychological and physical health domains will be assessed using appropriate mediation (process) analysis statistical models.

Primary efficacy analysis will contrast total days of illness and area-under-curve global severity in each intervention group with corresponding values in the control group. Unadjusted contrasts will be done by t-test, using variable transformation if skewness requires. Adjusted analyses will be based on zero-inflated regression models taking into account the episodic and variable nature of ARI illness. Zero-inflated regression models incorporate a logistic sub-model to account for people without ARI illness, and a linear sub-model to account for the variability in continuous measures of severity and duration. These will employ conservative estimates with standard error inflated using Huber/White maximum likelihood estimation. Co-variates to be controlled for in this model will include: age, sex, body mass index, smoking status, comorbidity, highest level of education achieved, neuroticism, conscientiousness, general physical health and general mental health. The null hypotheses of "no effect on primary outcome" will be rejected in favor of alternative hypothesis of "some effect" if p-values are ≤0.025. Confidence intervals will be constructed for all outcome variables, allowing estimation of effect size. Zero-inflated regression models for primary efficacy analyses will be done using M-Plus version 6.1or similar software.

Influence of interventions on secondary outcomes will be assessed using ANOVA-based multivariate regression models using SAS software. Adjustment for multiple comparisons will be incorporated, and interpretation will be cautious. In general, the investigators' will want to see relationships with p<0.01 in order to justify tentative null hypothesis rejection. Pre-planned secondary efficacy analyses will include effects of interventions on: 1) absenteeism, 2) health care utilization, 3) general physical health (SF-12), 4) general mental health (SF-12); 5) perceived stress (PSS-10), 6) self-efficacy (MSES, ESE), 7) sleep quality (PSQI), 8) body weight (BMI), 9) blood pressure and 10) pro-inflammatory cytokines. One-sided testing will be based on the underlying hypotheses that the behavioral trainings lead to improved physical and mental health, sleep quality and self-efficacy, and to reductions in absenteeism, health care utilization, stress, body weight, blood pressure and proinflammatory cytokines.

Apart from flu shots, hand-washing, smoke avoidance, and maintenance of general health, there are no known safe and effective prevention strategies for acute viral respiratory infection. Interventions to be tested are low risk, and more likely to confer benefit than harm. Influenza vaccination (flu shots) will be provided free of charge to all participants, and could be considered a benefit. Participants will not be required to forego any proven or accepted preventions or treatments, and in fact will be encouraged to seek appropriate medical attention for any condition, including respiratory infection, and to continue their regular health care practices. Antibiotics, anti-influenza antivirals, and other ARI treatments will not be disallowed, but instead will be tracked as secondary outcomes. In general, participants will be asked to continue the study interventions they are assigned, and to forego the study interventions they are not assigned. However, they will be informed that they remain free to make their own health care, behavioral and life style decisions. In particular, exercise will not be discouraged for participants in the meditation or wait-list control group, as exercise is known to confer some health benefit. However, the investigators' will ask participants in the exercise and control group to refrain from starting a meditation program, and that all participants refrain from starting any purportedly immune-modifying supplements (eg, echinacea, ginseng, zinc, elderberry, etc.), unless advised to do so by their physicians. If any participants do choose to actively begin an intervention they were not assigned (i.e. a person randomized to exercise chooses to take the mindfulness meditation course instead), the investigators' will ask them to continue in the study, and will use that "cross-over" information in secondary analyses using the principles outlined in Section 4.3 of the protocol.

This proposed study was submitted to and approved by the Human Subjects Committee of the Health Sciences Institutional Review Board of the University of Wisconsin - Madison (UW HS-IRB), and will be in compliance with HIPAA and all other federally mandated human subjects regulations. All members of the research team will complete the protection of human subjects' tutorial required by the UW HS-IRB's Human Subjects Committee prior to the study's initiation. All named UW Co-Investigators, Consultants, and current Project Personnel have in fact already completed this tutorial. All study personnel, including research assistants, will be trained in confidentiality, informed consent procedures, and other aspects of human subject protection.

A Data and Safety Monitoring Committee (DSMC) will oversee human subjects recruitment and monitoring, and will function independently of the principal investigator and co-investigators. The investigators' expect that this committee will function similarly to that established for the first MEPARI trial, and will choose to meet once or twice yearly to monitor recruitment and retention, and to provide oversight and guidance in relation to safety, bioethics, and other human subjects concerns that may arise.

Acute infection from influenza and other respiratory viruses leads to much human suffering and loss of economic productivity. The investigators' own evidence suggests that training in either meditation or exercise may lead to substantial reductions in ARI disease burden and work absenteeism. In addition to testing whether the investigators' findings are replicable in a larger sample with refined methodology, this proposed comparative effectiveness translational research will investigate mechanisms of action and provide initial estimates of cost-effectiveness. If positive findings are confirmed, this line of research could have direct and immediate impact on public and private health-related policies and clinical practice, as well as on scientific understanding of respiratory infection.

A copy of the full protocol (including references) is available upon request.