Non-communicable diseases in the southwest of Iran: profile and baseline data from the Shahrekord PERSIAN Cohort Study

The Bakhtiari ethnic group mainly lives in Chaharmahal and Bakhtiari (CH&B) province in Iran and has an estimated population of 1.25 million. It is a subgroup of the Iranian Lurs, and the genetic background of Bakhtiari people is different from other Lur populations [19]. They speak the Bakhtiari dialect. The Bakhtiari have maintained their bloodlines primarily intact over the centuries, mainly marrying within their tribe. Other notable differences with other Iranian ethnic groups include their culture and social and local customs (mourning and weddings), and dietary habits (tiri bread, mountain vegetables, animal oil consumption, traditional dairy products), type of employment (animal farming, herding, agriculture, hunting), clothing and apparel (local dress), and different environmental exposures, such as exposure to sunlight at high altitudes. CH&B province covers an area of 16.421 km² and is situated in the southwest of Iran, north of the Zagros Mountains, which have the highest average elevation above sea level in Iran; the Shahrekord region is known as the “roof of Iran”. Despite its relatively small area (1% of the total area of Iran), CH&B holds 10% of the country’s water resources. Because of its mountainous nature where moist Mediterranean air converges, this province has relatively abundant rainfall [18, 19]. Because of the rare ethnic groups (Bakhtiari), Fars, and Turk living in this region, this cohort is unique in Iran and worldwide [18].

Participants were required to be aged between 35 and 70 years at the time of recruitment, having lived in the specified area for at least one year, having completed and signed the informed consent, and having Iranian nationality (i.e., having an Iranian birth certificate and a national identification number).

Both gender (male and female) and pregnant women were included. The 35-70 years age range was chosen because people in this age group are more likely to have well-established behaviors and lifestyles, are active enough to participate in a cohort study and have a sufficiently high risk of developing NCDs during the study [13, 18]. People unable to undertake the required questions and measures (e.g., due to disability or mental disorders) were not eligible for the study.

The numbers of invited and recruited SCS participants, the area names in these districts, and the region map are presented on the SCS protocol and website. The formula used for calculating the study's sample size was described previously in the SCS protocol [18]. The implementation, feasibility, and sampling processes of the SCS were performed in the pre-pilot phase of the study from 22 Nov 2015 to 10 Sept 2016. In this phase, the data collection process and biological sample storage capacity were evaluated. Aspects of the regulation of ownership, preservation, and storage of data were also finalized. The pre-pilot phase also allowed i) to evaluate the participants' response rate and the recruitment and training of interviewers, ii) to determine the frequency of participant follow-up contacts, iii) to check the validity of measurements, and iv) to implement adequate procedures for bio-sampling and quality control of the collected data. The study protocols were revised accordingly to improve the validity and reliability of the questionnaires and the acceptability of the data collection techniques [13, 18].

The pilot phase was conducted from 6 Oct 2016 to 20 Dec 2016 to further evaluate the study protocol's main strengths and weaknesses and participant recruitment process. A total of 1000 participants aged 35–70 years were recruited for this phase. After receiving confirmation from the quality control team, the main stage, enrolling all participants, started with the pilot phase on 6 Oct 2016. Checklists about several study processes (invitation process and evaluation of response rate, questionnaires, sampling and measurements, assessment of ascertainment methods, establishment of follow-up data processes, testing the validity of disease outcomes) were completed. The quality assurance team issued a certificate of study commencement after the infrastructure and procedures were in place.

The multistage sampling method (stratified proportional cluster sampling) was applied to recruit participants in the SCS in the pilot and main phase. According to the national census statistics, 70% of the CH&B province population lives in urban areas, and 30% live in rural areas, so the urban and rural strata accounted for 70% and 30% of the study sample size respectively [18]. Sampling in urban areas was also carried out using the cluster sampling method. Each of the four regions of Shahrekord County (the capital city of CH&B) was considered an eligible cluster. Then a specific geographical region of Shahrekord was randomly chosen as the cohort cluster. The population covered by each urban healthcare center (cluster) was used as sampling weights for recruiting participants from the corresponding (n=77030) health care centers to constitute the final study sample for the urban area (n= 7034). For rural areas, the sampling process was as follows: i) nine CH&B counties were considered as eligible clusters, ii) the Ardal county was randomly selected as the cohort cluster, and iii) three of the five rural clusters of the Ardal county were randomly selected for inclusion in the SCS. A total of 3041 participants from these rural areas were recruited based on census-collected information on healthcare coverage provided by the health centers and health houses. The first contact with prospective SCS participants was made through an invitation by phone to eligible people, and this process continued until the required sample size was met. Several initiatives were taken to increase participant enrolment and satisfaction, such as inviting people to visit the center, friendly staff and interviewers, and the provision of a suitable space for participants to undertake data collection. All medical examinations and tests were free of charge, and the results of the tests were provided to the participants. Less than one percent of the people contacted declined to participate in the study, and only 20 participants dropped out. Living areas and sex- and age-specific proportions of the participants included in the SCS were cross-checked against and found by the national population figures provided by the national census conducted by the Statistical Center of Iran. The distribution of the main socio-demographic characteristics of SCS participants is shown in Table 2.

Data collection in SCS, including the definition of NCDs, used the PERSIAN cohort protocols [13, 18]. In brief, the following NCDs were considered: heart diseases (myocardial infarction, ischemic heart disease), hypertension, endocrine conditions (diabetes mellitus, hypo/hyperthyroidism), neurological diseases (stroke/transient ischemic attack, epilepsy, Parkinson’s disease, migraine, chronic headache, Alzheimer’s disease/dementia, multiple sclerosis), chronic kidney disease, musculoskeletal diseases (arthritis, osteoporosis), respiratory diseases (chronic obstructive pulmonary disease, asthma), gastrointestinal conditions (peptic ulcer, Chron’s disease, ulcerative colitis, fatty liver diseases), cancers, mental and psychological disorders (depression, anxiety). Diabetes was diagnosed as fasting blood sugar (FBS) ≥ 6.99 mmol/L (126 mg/dL) or receipt of blood glucose-lowering treatment. In addition, individuals who were under treatment for physician-diagnosed diabetes were considered to have diabetes. Hypertension was defined as systolic blood pressure ≥140 mmHg or diastolic blood pressure ≥90 mmHg or a prior diagnosis of hypertension or being on antihypertensive medication. Other chronic diseases were self-reported (participant’s response to the question; ‘Has a doctor ever told you that you have any of health problems or presence/absence of NCDs?’), medical documents, clinical history, interview and physician diagnosis [13, 18, 20]. Anthropometric variables including height (in centimeters), weight (in kilograms), waist and hip circumferences (in centimeters) were measured using US National Institutes of Health protocols [13, 18].

The height of participants was measured using a Seca 206 stadiometer and their weight was measured using a Seca analog scale. A standard tape meter was used to measure the participants’ wrist, hip, and waist circumference. Body mass index (BMI, kg/m²) was calculated. Physical activity information was collected using the general questionnaire, and self-reported daily activities were converted to metabolic equivalent rates (METs). Blood pressure and pulse rate measures were obtained using a standard barometer (Richter Japan). Tobacco smoking (including tobacco type), drug use (including type of illicit drug), and alcohol use in the last thirty days and one year, respectively. Alcohol use was assessed as standard drinks (one standard drink =100ml wine, 285ml beer or 30ml spirit/toddy/arrack). A consumption of at least six and four standard drinks for men and women, respectively, on at least one occasion during the last 30 days, was considered harmful alcohol use. Based on the laboratory kits used (Pars Azmoun), hypercholesterolemia was defined as having a total cholesterol level of ≥240mg/dl or currently using lipid-lowering drugs. Hypertriglyceridemia was defined as having an overall triglyceride level of ≥150mg/dl or currently using lipid-lowering drugs. According to the American Association for Clinical Chemistry, the average value for urine pH is 6.0, but it can range from 4.5 to 8.0. Urine pH under 5.0 is acidic, and urine pH higher than 8.0 is alkaline or basic. Socio-economic status was defined based on education and wealth index. Education was defined in 5 levels: no schooling (<1 year of primary school), primary school (1-5 years), middle school (6-8 years), high school (9-12 years), and university (>12 years). The wealth index was calculated using multiple correspondence analysis (MCA) on household assets and divided into five quintiles [13, 18, 21]. The tools were used for baseline data collection from 2015 to 2019 and are described in Table 1. In addition to the extensive standard PERSIAN questionnaires, additional questionnaires unique to SCS were also completed, including General Health [22], WHO Quality of Life-BREF [23], Chronic Stressors and Coping Strategies [24], WHO MONICA and the ROSE Angina questionnaire [25], Social Capital [26], Screening Tool for Joint Pain and Musculoskeletal Diseases [27], Health Literacy [28], Oxford Happiness [29], and Oswestry Low Back Pain Disability [30], following SCS protocols [18] in CH&B (Table 1) and additional data unique to SCS were also completed including body composition variables included total body water, body fat mass and percentage, and muscle thickness, which were measured using a body composition analyzer (Tanita, Japan). A spirometry test (pulmonary function test) was performed using the Spirometer device (Spirolab [MIR, Italy)]. An electrocardiography test (ECG) was carried out using an electrocardiogram device (Cardiax®, USA). Although the validity of these questionnaires was assessed in previous studies [13‐15, 18, 23, 24], SCS's experts further assessed their face validity and approved their use in the pre-pilot phase of the study.

In the pre-pilot phase of the SCS, 100 participants completed the questionnaires. The questionnaire items had coefficients of Cronbach's alpha ranging from 82% to 91%, so they were considered reliable. A complete description of the data dictionary can be found in the http://​persiancohort.​com/​wp-content/​uploads/​2020/​09/​PERSIAN-Cohort-Data-Dictionary.​pdf.

All phases of the study and data collection were monitored by a quality control team, including clinicians, a laboratory specialist, two statisticians, and an epidemiologist, under the supervision of the principal investigators.

The follow-up process aimed to register new cases of common NCDs and their outcomes, including death, cause of death, hospital admissions, and update information on exposures. The SCS focuses primarily on the most common NCDs, including cardiovascular diseases, cancers, and the main endocrine, digestive, hepatic, renal, psychiatric, and respiratory disorders, defined using the International Classification of Diseases 10th version (ICD-10). The annual follow-up of the SCS began in October 2017 and included questionnaires, medical examinations, and linkage with other databases (death, cancer registry). The study follow-up is done annually through telephone calls and links with health databases to identify disease outcomes. More specifically, the follow-up of participants is performed in two forms: an active form including phone interviews and face-to-face interviews (when outcomes occur), and an inactive form, including self-reports. Identification of outcomes is made through automatic notifications received from the healthcare system and linkage with other health databases such as the National Disease and Health Outcome Registry Systems [13, 18]. The follow-up is carried out by a team of trained staff under the supervision of an experienced epidemiologist. Three internal medicine physicians and a professional epidemiologist do outcome assessments, including the cause of death identification. During phone call follow-ups, if an interviewer cannot gather the requested information, additional phone calls are made during the three consecutive weeks (up to 5 to 6 rings). Further attempts to collect incomplete data are made during home visits and face-to-face interviews. In rural areas, the data collection process is conducted in local health care units (Health Houses) by health care staff and an experienced epidemiologist. Participants who experience an outcome are invited to undergo an in-person examination. Additional information about the participants' health history is obtained from the Hospital Information System (HIS) and the integrated electronic health system. When a death occurs, the SCS team visits the participant's home and completes an autopsy form. A schematic representation of the phases of the study and the data collection and follow-up process is shown in Figure 1.

‘Multimorbidity’ was defined as the co-existence of two or more chronic diseases in the cohort, and we further categorized participants into groups defined as having 0, 1, 2, or 3 and more comorbidities; the following common NCDs were considered: myocardial infarction, ischemic heart disease, hypertension, diabetes mellitus, hypo/hyperthyroidism, stroke, epilepsy, chronic headache, chronic kidney disease, arthritis, osteoporosis, chronic obstructive pulmonary diseases, fatty liver, cancers, asthma, gastrointestinal conditions, peptic ulcer, Crohn’s disease, ulcerative colitis, multiple sclerosis, chronic depression and psychological disorders). ‘Multi-risk factors’ was defined as having two or more of the following risk factors among obesity, hypercholesterolemia, hypertriglyceridemia, low physical activity, opium use, tobacco use, alcohol use, hookah use, high fasting blood sugar, high systolic blood pressure and high diastolic blood pressure. Logistic regression models were used to calculate crude and adjusted odds ratios (OR) and 95% confidence intervals (95% CI) for the association of socio-demographic and biochemical variables, as well as the multi-risk factors variable (independent variables) with the multimorbidity variable (either ≥2, or using the 4 categories) as the outcome. Statistical tests were two-sided, and P values less than 0.05 were considered statistically significant. Data analysis was performed using Stata software (Stata Corp. 2020. Statistical Software: Release 16. College Station, TX: Stata Corp LP).