Approval was obtained from the Human Research Ethics Committee (HREC) of UNSW Sydney, Australia (HC15586) as the lead site. Details of individual site ethics approvals from participating sites are available upon request.
The SIMPAQ was iteratively developed between April 2014 and May 2016 by a multidisciplinary, international working group with both clinical and research expertise (including psychiatrists, psychologists, physiotherapists, exercise physiologists, and epidemiologists) regarding physical health care interventions for people living with mental illness. The first meeting was held in Padua, Italy, in April, 2014, to identify the common challenges experienced when assessing physical activity among people with mental illness. At a subsequent meeting in July, 2015, held at the Institute of Psychiatry, Psychology and Neuroscience (UK) in London (UK), consensus agreement on the wording of the questions that constitute the SIMPAQ was obtained.
Participating Research Sites
In addition to disseminating information about the project via the international workgroup, an editorial was published in 2016 describing the proposed validation process that helped to identify additional study sites . All study material and administration protocols were available from the project website (www.simpaq.org) when recruitment commenced in May 2016. All sites were required to nominate a site coordinator and sign an authorship agreement document. Along with study material, site coordinators received a briefing from investigators SR and PBW and were also in regular contact with the study coordinator RM. Eligibility criteria for potential sites included willingness to recruit patients meeting the inclusion criteria outlined below and availability of a site coordinator with expertise in either mental health or physical activity research.
Translation was conducted according to the Principles of Good Practice for the Translation and Cultural Adaptation Process for Patient-Reported Outcomes (PRO) Measures . This process involved ten steps including 1) preparation, 2) forward translation, 3) reconciliation, 4) back translation, 5) back translation review, 6) harmonization, 7) cognitive debriefing, 8) review of cognitive debriefing results and finalization, 9) proof reading, 10) publication on SIMPAQ website.
All participants were required to provide written informed consent and be willing to wear an accelerometer for seven days. Eligibility criteria also included: i) aged between 18 and 65 years, ii) a current inpatient or outpatient of one of the treatment facilities identified as a SIMPAQ validation study site and iii) met DSM-5 or ICD-10 criteria for any mental disorder, excluding eating disorders.
Participants were approached by a researcher nominated by the site coordinator who was not involved in the direct care of the patient. The researcher obtained written informed consent. Data was collected from each participant during two face-to-face sessions, at least seven days apart. Researchers involved in data collection included either mental health or exercise professionals.
Demographic and descriptive information was collected including assessment of symptoms and cognitive ability . Participants completed the SIMPAQ (Time 1) and were given a tri-axial accelerometer (Actigraph GT3x or GT3x+ (both models contain the same accelerometer and processing method)) along with standardised instructions for wearing the device.
Participants completed the SIMPAQ (Time 2) covering the period of accelerometer wear time.
Participant demographics and descriptive information
A standardised form was used to obtain demographic and descriptive information including: age, sex, treatment setting (inpatient or other), years of completed education, previous 7-day employment status (yes or no), previous 7-day tobacco smoking status (yes or no), body mass index (derived from measures of height [m] and weight [kg]).
Each country in which a site acquired SIMPAQ data was assigned an income status (either high income or other) based on World Bank classification (www.worldbank.org).
Psychiatric diagnoses that applied to individual participants based on medical records were recorded. It was recognized that participants may meet criteria for more than one psychiatric diagnosis, and all diagnoses that applied to each participant were recorded. The standardised form asked researchers to tick yes or no for the following diagnostic categories based on clinical diagnoses; schizophrenia spectrum disorders, bipolar disorder, depressive disorders, anxiety disorders, obsessive-compulsive disorders, substance-related & addictive disorders, neurocognitive disorders and other disorders. We identified individuals who were assigned a single diagnostic category, and those with psychiatric co-morbidity.
Physical health conditions
The presence or absence (yes or no) of the following physical health conditions at the time of assessment were also recorded by the researcher; diabetes, high cholesterol, high blood pressure, stroke and chronic pain based on self-report and medical records.
Researchers were asked to indicate whether participants were currently prescribed the following classes of psychotropic medication (yes or no): antidepressant, antipsychotic, or mood stabilising medications.
Symptom severity – DSM-5 Self Rated Level 1 Cross Cutting Symptom Measure
The 23-item DSM-5 Self-rated Level 1 Cross-cutting Symptom Measure  was used to assess symptom severity. This measure consists of 23 questions that assess 13 psychiatric domains, including depression, anger, mania, anxiety, somatic symptoms, suicidal ideation, psychosis, sleep problems, memory, repetitive thoughts and behaviours, dissociation, personality functioning, and substance use . Each question asks about how much (or how often) the individual has been bothered by the specific symptom during the past two weeks and is rated on a 5-point scale (0=none or not at all; 1=slight or rare, less than a day or two; 2=mild or several days; 3=moderate or more than half the days; and 4=severe or nearly every day). We summed the total scores across these domains and dichotomized the scores around the median (20); lower symptom severity was defined as scores <21; higher symptom severity was defined as scores >=21.
Cognitive functioning – Montreal Cognitive Assessment (MoCA)
The Montreal Cognitive Assessment (MoCA) is a brief screening tool used to assess cognitive functioning . The MoCA assesses multiple cognitive domains including attention and concentration, executive functioning, memory, language, visuo-constructional skills, conceptual thinking, calculation and orientation. Scores ranged from 0-30 with scores of 26 or higher considered within normal range. Given that many psychiatric syndromes are associated with cognitive impairment (e.g. schizophrenia), we did not exclude participants scoring less than 26. Results are reported for those with scores above and below this threshold.
Simple Physical Activity Questionnaire
The 5-item SIMPAQ required people being interviewed to account for time spent in bed overnight (box 1), time sedentary, including napping (box 2), time spent walking (box 3), time spent exercising (box 4) and time engaged in incidental activity (box 5), averaged over the past seven-day period (see Figure 1). The sum of the hours recorded in the five SIMPAQ boxes should add to approximately 24-hours, providing interviewers with an opportunity to clarify with participants if significant under or over-reporting has occurred (e.g. <18 hours or >30 hours of estimated time). For an estimate of total self-reported moderate-vigorous physical activity (MVPA) time, time spent walking (box 3) and exercising (box 4) were combined to provide total MVPA (hours per week).
Percentage of 24-hour period accounted for by SIMPAQ items
The SIMPAQ was designed to capture activity over a representative 24-hour period from the previous 7-days. By summing Boxes 1 through 5, the total hours accounted for should equal approximately 24. To evaluate how well this was achieved in the current study, we calculated the fraction of time accounted for by using the following formula:
sedentary time (box 2) + walking time (box 3) + exercise time (box 4) + incidental activity time (box 5)
24 – time in bed (box 1)
Accelerometer – Actigraph GT3/x
Participants were asked to wear a tri-axial accelerometer (Actigraph GT3x or GT3x+; ActiGraph LLC, Fort Walton Beach, FL) on the right hip during waking hours for a period of seven consecutive days to objectively assess physical activity. Accelerometers record raw acceleration data that is converted into objective activity measures such as step counts. Participants were shown how to wear the device on the right hip using either a belt clip or elastic waist band. After the seven-day period participants returned the device and again completed the SIMPAQ for comparison with Session 1 data. Prior to Actigraph devices being issued to participants they were initialised using the online portal. Each participant was setup in CentrePoint and sex, age and weight were entered and the device allocated to the subject. Accelerometry data were retrieved from the device using CentrePoint, a secure online portal designed and distributed by Actigraph specifically for multi-site study co-ordination. ActiLife v6.13.3 software was used to extract data from CentrePoint and derive variables to be used in the calculation of validity between accelerometry data and SIMPAQ items. Participant data were included for analysis if at least eight hours of valid wear time were available for at least four days. Non wear time was defined as at least 60 minutes of consecutive zeroes, allowing for spike level of 100 counts per minute . We followed Freedson et al.  to classify activity intensity using cutpoints for time spent in sedentary (<100 counts/min), light (100–2019 counts/min), moderate (2020–5998 counts/ min), and vigorous intensity (>5999 counts/min) activity .
Data analysis and cleaning
Non-parametric Spearman correlation coefficients were calculated as the primary measure of agreement between assessment time points (Session 1 and Session 2) (test-retest reliability), and between the SIMPAQ data and accelerometer counts (evidence of validity). Agreement between the SIMPAQ and accelerometer data was also assessed through Bland-Altman mean-difference plots with 95% limits of agreement. Intraclass correlation coefficients (ICC) along with 95% confidence intervals were also calculated. Analyses were conducted both with all valid data, and excluding outliers defined as those with SIMPAQ values that were greater or less than 2.5 SD from the mean for that item. Results are reported for the entire sample with available data and stratified by cognitive function as assessed by the MoCA and psychiatric symptom severity derived from the DSM – Cross-cutting tool. The sample were also stratified according to specific diagnoses, and those with psychiatric comorbidity. Income status, treatment setting, sex, age, body mass index (BMI) and smoking status data were analysed separately. Data were analysed using SPSS v24.
Test re-test reliability was determined using Spearman Rho correlation coefficients between SIMPAQ items at Session 1 and Session 2. Given that the SIMPAQ asks responders to report activity from the previous seven-day period, and the potential for hospital admission to impact physical activity levels, only data from outpatients were utilised for reliability calculations.
To provide evidence for the validity of the SIMPAQ questionnaire, Spearman correlation coefficients were calculated for MVPA as assessed by the SIMPAQ (box 3 + box 4) and MVPA as recorded by the accelerometer, and for sedentary time (SIMPAQ box 2) against the accelerometer.