oa Obtaining Reliable Evidence on the Determinants of NCDs in the UAE: The Abu Dhabi Cohort Study

Obtaining Reliable Evidence on the Determinants of NCDs in the UAE: The Abu Dhabi Cohort Study Cardiovascular disease and diabetes are extremely common in Abu Dhabi and throughout the Arab World. While cohort studies have made tremendous contributions to scientific knowledge of the epidemiology and determinants of cardiovascular disease, few have been done in Arab populations. To study the causes of these diseases and other diseases common to the Abu Dhabi, we are establishing a prospective cohort study (the Abu Dhabi Cohort Study) for epidemiologic research. In Phase I of the study, we will recruit 20,000 adult subjects from multiple sites including ambulatory care centers, universities, workplaces and blood collection centers. After providing informed consent, participants will complete a 30–45 minute survey, administered via tablet computer. The questionnaire will cover lifestyle, health habits (tobacco use, diet, physical activity), health status and other factors. Biological samples will be collected, including blood, urine and an oral wash. Physical measurements will also be carried out (blood pressure, grip strength, and anthropometric measures including weight, height, waist size and percent body fat). In Phase II, a subsample (approximately 2,000) of Phase I participants will be invited to return for non-invasive testing, which will include determinants of cardiovascular disease (carotid artery intima medial thickness and brachial artery flow-mediated vasodilation) and pulmonary function (spirometry). Study subjects will also provide consent for long-term follow-up for the adult lifespan to determine health outcomes and related health determinants through brief quarterly surveys and a longer annual questionnaire. Study design and progress is guided by an International Advisory Committee and a local Steering Committee. In order to finalize cohort study procedures, we carried out a pilot study on 500 participants in 2015. As this was the first time such a study was being done in the UAE, and following best practice from similar studies internationally and in the region, it was essential to do a pilot study first to establish both the study feasibility and to ensure that all protocols and participant materials were appropriate for the local population. The pilot study had five main objectives and the results in relation to each of these objectives are summarized below:

1. Development of participant materials (participant information leaflet, consent form and questionnaires.) The first stage of the pilot study (September to December 2014) was a series of focus groups with potential participants to develop and optimise the participant information material and study protocols. These were conducted at UAE University, Zayed University and NYUAD. Participants provided feedback on the study design, approaches to recruitment and the design of the participants’ materials including the participant information leaflet (PIL) and consent form; the questionnaire and all aspects of the assessment visit. Particular focus was given to aspects of the study which participants were unfamiliar with, such as genetic testing. In general, participants were happy with the study design and recruitment approaches, but some changes were made to the PIL in response to feedback (mainly making it shorter).

2. Assessment of participation rates for various recruitment strategies Making use of modifications suggested from the focus groups, the second stage of the pilot involved recruitment of 500 participants into the cohort study from January to May 2015, at facilities licensed by HAAD with authorisation for medical research (Zayed Military Hospital's Primary Health Care Clinic and clinic at Zayed Military City and the Abu Dhabi blood bank.) The response rate was 68% (522/769). The main reasons for not taking part were: Lack of time (60% of non-respondents)Not interested (20%) Not convinced (10%) Other (10%) Responders and non-responders were similar with respect to age and gender. No monetary or non-monetary gifts were given to participants.

3. Evaluation of baseline assessment visit. In each location, HAAD- licensed nurses carried out physical measurements including waist, hip, and neck circumferences, as well as height, weight, handgrip, blood pressure, and body composition (impedance). The nurse also assisted participants to complete a questionnaire using an iPad, and collected blood, an oral wash and urine samples. The assessment visit took most participants between 40 to 60 minutes. Completion rates for the various components of the study were approximately: Questionnaire – 70% Physical measurements – 80% Blood 95% Urine 80% Mouthwash 80% Based on the pilot study experience, one nurse and one research assistant are required to recruit, consent and collect materials from about 10 people daily.

4. Assessment of procedures for biological sample collection, transport and processing. Blood, urine and mouthwash samples were transferred to NYUAD laboratories; aliquots of which were stored at – 80°C for analyses. Samples will be analysed for a battery of routine and specialized biochemical markers from May 2015 to December 2015.

5. Assessment of acceptability of various parts of the study A brief post-visit questionnaire was sent to all participants by email/SMS to assess their understanding of key components of the consent (as well seeking opinions on the assessment visit and highlighting any areas for improvement). Of 330 questionnaires sent by email, 34 were completed (10% response rate). In response to the question ‘How would you answer if a close friend or family member were to ask you “should I participate in the Abu Dhabi Cohort study?”, 80% said they would definitely take part. Analyses of data collected in the pilot study Simple descriptive analyses of questionnaire and physical measurements data collected in the pilot study were carried out to help inform sample size calculations for the main study. Crude (not age-adjusted) prevalence for selected characteristics are shown below: Selected baseline characteristics Men Women Number of participants 325 150 Mean age 32 30 Self–reported Diabetes 8 9 Self-reported Hypertension 11 12 Self-reported Hypercholesterolemia 19 11 Overweight 34 25 Obese 23 18 Cigarette smoking 18 1 Midwakh smoking 17 1 Shisha smoking 13 2