Study design and overview

MISC is the first prospective cohort study in the UAE to comprehensively investigate maternal and infant factors in relation to child health outcomes, as well as early-life determinants of NCDs, through integration of sociodemographic, dietary, lifestyle, anthropometric and biological and cognitive data. At present, recruitment of study participants is completed; data collection and preliminary analyses are being conducted.

Ethical approval

The study was approved by the Research and Ethics Committee, University of Sharjah (REC/14/01/1505) and by Al Qassimi Clinical Research Centre Ethical Research Committee (REC Reference Number: 215 12015–03), by the Ministry of Health Ethical Research Committee (R02) and by Dubai Health Authority (DSREC-0/2016). The research ethics bodies focused on ensuring adherence to the regulatory requirements for the protection of pregnant women and babies as vulnerable research subjects, throughout all study procedures (consent documentation, data collection including blood withdrawal requirements, and data transfer). It was thus requested by Al Qassimi Clinical Research Centre to prepare a separate consent form to obtain the fathers’ approval on the behalf of the infants, since according to Emirati law, the father is the legal guardian. Moreover, special considerations for blood withdrawal by pediatric-certified phlebotomists were noted.

Study setting and population

Participants were recruited from antenatal clinics in three main public governmental hospitals, and seven Primary Health Care (PHC) clinics and Mother and Child Centers (MCH) in the Emirates of Sharjah, Dubai and Ajman. The research team visited the pre-specified clinics, screened eligibility criteria of pregnant women attending the clinics, informed them of the objectives, conduct and timeline of the study and invited them to participate. The study followed a convenient sampling and did not adopt a special advertisement for recruitment. Eligible women were asked, along with the fathers, to provide signed consent to participate in the study, and were given a copy of it. The participants were chosen according to the following criteria.

Inclusion criteria

• Pregnant women:
  • Emirati or Arab expatriate
  • Age: 19–40 years
  • Singleton pregnancy
  • Within the third trimester of pregnancy (27–42 weeks of gestation)
  • Free of chronic diseases (diabetes, hypertension, kidney disease, cancer, or other chronic diseases), autoimmune disorders, or infections with the human immunodeficiency virus, or hepatitis in preconception
  • Receiving antenatal care in any of the above-mentioned clinics and expected to give birth in a participating public hospital
  • Not planning to permanently leave the UAE during the timeline of the study.
• Infants:
  • Early term (37 weeks of gestation) or full term (between 39 weeks 0 days, and 40 weeks 6 days of gestation) live birth infants
  • Free of major illnesses (congenital diseases or any health condition that could deter breastfeeding, such as cleft lip or cleft palate), i.e. apparently healthy infants.

Exclusion criteria

• Pregnant women:
  • Multiple pregnancy (pregnant with two or more fetuses)
  • High risk pregnancy or pre-eclampsia
  • History of chronic diseases.
• Infants:
  • Preterm (<37 weeks of gestation) or late term (>42 weeks of gestation) infants
  • Congenital diseases or any health condition that could deter breastfeeding.

Bearing in mind that loss to follow-up might be “not at random” and lead to biased estimates [35], different strategies were put place to increase participation and retention in the cohort: sharing the biomedical results with the study participants at 7-month postpartum; providing gift vouchers and toiletries for attendance to each time-point, ensuring transportation coverage, if needed. In an attempt to increase retention, text messages will be sent to each participant at each time-point. Recruitment was completed during June 2016.

Sample size and power analysis

As mentioned above, one of the main purposes of the MISC is to explore the relationship between maternal characteristics and infant outcomes. When the assessed variables are both continuous, we often resort to Pearson’s correlation coefficient (r), enabling the assessment of a linear dependence (correlation) between the variables, as well as the strength and direction of this correlation. The absolute value of r ranges between 0 and 1, and its increasing magnitude indicates a stronger association. Taken the limited data on the above-mentioned associations, r = |0.2| was chosen, as it yields the largest sample size. An r<|0.2| indicates a weak correlation- usually considered too small to be clinically significant. Using the Power Analysis and Sample Size System (PASS) software version 11, the sample size needed to detect a correlation as small as |0.2|, at 80% power and 5% type I error, is determined at 193 (rounded-up at 200). Another way of assessing the relationship between maternal and infant-related factors is through regression analysis, where an outcome variable is regressed on a number of explanatory variables (predictors). Based on the rationale of considering 10–15 observations per predictor, the sample size of 200 would be sufficient to conduct regression analyses on up-to 15 predictors. Since it is not expected to exceed this number of independent variables per regression model, we considered the sample size of 200 as adequate. Considering that previous birth cohorts showed a retention rate of 85% [36], it was anticipated that around 15% of the participants will be lost after the first visit. Furthermore, we expected to lose an additional 10% due to preterm delivery and another 10% due to still birth. As such, the original sample size was inflated by 35% and the target sample size was determined at 270 pregnant women. We expect to retain around 65% of the participants at two years postpartum.

Study protocol

Assessments for this cohort were planned over six time-points, extending from late pregnancy till two years postpartum. The postpartum time-points were chosen in parallel with the immunization schedule in the UAE, when mothers are expected to visit the PHCs/MCHs to vaccinate their infants. The cohort assessments and timeline are detailed in Table 1. Data were collected via interviewer-administered questionnaires, anthropometric measurements, dietary records, bio-specimens (blood and breast milk) analyses and review of hospital electronic medical records. Data collection tools and methods are detailed below.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Data collection and study timeline.

https://doi.org/10.1371/journal.pone.0198278.t001

Questionnaires

Multi-component questionnaires used for this study were adapted from the MINA cohort study [25]. These are content-validated, culturally-adapted and pilot-tested questionnaires that were revised for use in this study among Emirati pregnant women. The Arabic version of the questionnaires were administered by trained research nutritionists.

Maternal sociodemographic characteristics and lifestyle: sociodemographic factors include age, nationality, place of residence, occupation, education, parity, and income. Lifestyle practices include physical activity, cigarette and shisha (narghile/water pipe) smoking. Physical activity was assessed using the Pregnancy Physical Activity Questionnaire (PPAQ) [37] during the third trimester of pregnancy, and the short International Physical Activity Questionnaire (IPAQ) [38] during postpartum. Metabolic equivalents (METs) were computed, and physical activity levels were classified as low, moderate or high-intensity. Maternal cigarette smoking dose will be assessed according to Bachir and Chaaya [39], maternal shisha smoking dose will be assessed according to Tamim et al. [40], These questions were extended with one question inquiring about other smokers living in the same house.

Maternal dietary intake: maternal intake was evaluated using a culture-specific food frequency questionnaire (FFQ) and multiple-pass 24-hour dietary recalls. The used FFQ is a semi-quantitative questionnaire consisting of 89 items, grouped into 24 food categories comprehensively representing the Gulf and Middle Eastern cuisine. The 24-hour recalls were assessed using the United States Department of Agriculture’s Multiple Pass Food Recall (MPR), which attenuates recall bias [41,42]. For the FFQs and 24-hour recalls, participants were assisted with the reference portions of the two-dimensional food portion visual (Millen and Morgan, Nutrition Consulting Enterprises, Framingham, Massachusetts, United States), as well as supplementary visual aids about portion sizes of common items in the traditional Gulf and Middle Eastern cuisine meals [43], to help estimating ingested quantities. Daily energy and nutrient intakes will be calculated using the food composition tables provided by the NUTRIONIST PRO^™ diet analysis software (Axxya Systems LLC., USA, version 5.1.0, 2014, First Fata Bank, Nutritionist Pro, San Bruno, CA) and the food composition table of Middle Eastern foods for local and traditional dishes [44]. Supplement use were assessed via questions on the type, dose and frequency of used vitamins, minerals, herbal or other products.

Infant feeding practices and dietary intake: breastfeeding and complementary feeding practices will be assessed based on WHO indicators [45], and categorized as exclusive breastfeeding, predominant breastfeeding, or complementary feeding. In order to assess the infant’s complementary feeding, mothers will be asked about the timing of introduction and type of solid, semi-solid and soft foods. Minimum dietary diversity, minimum meal frequency, minimum acceptable diet, and consumption of iron-rich or iron-fortified foods will be assessed.

Infant cognitive development: assessed using the 12 and 24-month Ages and Stages Questionnaire (ASQ). The ASQ system is originally a parent-completed questionnaire but may also be completed by a professional interacting with the infant and the parent. The ASQ is a developmental screening tool, widely used in pediatric populations. It exhibits high measures of sensitivity, specificity, and reliability [46]. Total scores will be computed for communication, gross motor, fine motor, problem solving and personal-social areas, and assessed based on the cut-off values suggested by Squires and Bricker [47]. The Arabic versions provided by the publisher will be used.

Anthropometry

Maternal anthropometric assessment: height, weight, waist circumference and blood pressure were measured following standard techniques. Pre-pregnancy and post-partum weights were evaluated according to WHO classification criteria for BMI [48]. The rate of pregnancy weight gain was assessed according to the Institute of Medicine BMI criteria for pregnant women [49]. Waist circumference will be classified according to the International Diabetes Federation’s cut-off for abdominal obesity in females of Middle Eastern origin (≥80 cm) [50]. Percent body fat will be measured by bioelectrical impedance (Tanita WB 100A digital scale); values >32% will be classified as high [51]. Blood pressure will be evaluated according to the National Heart, Lung, and Blood Institute of the National Institute of Health criteria [52]; values of systolic (SBP) >140 mmHg or diastolic blood pressure (DBP) >90 mmHg indicate hypertension.

Gestational Diabetes Mellitus (GDM): GDM was diagnosed according to the National Institute for Health and Care Excellence (NICE) Diabetes in Pregnancy [53] guidelines. Blood glucose was measured and analyzed at the PHC/MCHs. All participants did a 2-hour 75 g Oral Glucose Tolerance Test (OGTT) to test for GDM. The latter was defined as a 2-hour plasma glucose level ≥7.8 mmol/L (140 mg/dl). The results were obtained from the mothers’ medical record.

Infant anthropometric assessment: measurements of head circumference, length, weight and mid-upper arm circumference (MUAC) will be performed following standard techniques by nurses at the hospitals, PHCs/MCHs. Weight and length measurements will be used to assess stunting, wasting, underweight, overweight, and obesity, following WHO criteria on child growth standards [54]. The MUAC will be used to assess the infant’s nutritional status, following WHO and UNICEF criteria [55]; values <11.5 cm indicate severe under-nutrition (wasting), 11.5–12.5 cm indicate moderate under-nutrition, and >12.5 cm indicate that the infant is well nourished.

Biochemical assessments

Blood profile: maternal and infant biochemical assessments include fasting serum lipid profile (triglyceride, total cholesterol, high density lipoprotein (HDL) and low density lipoprotein (LDL)-cholesterol), glucose, insulin, high molecular weight (HMW) adiponectin, high-sensitivity C-reactive protein (hs-CRP), Tumor necrosis factor (TNF)-α and interleukin (IL)-6 and vitamin D. Following the delivery, maternal blood samples (10 mL) were collected by certified phlebotomists in appropriate test tubes, centrifuged and temporarily stored, within 30 minutes of collection, in hospital freezers (-80°C). Stored samples were transported monthly in an icebox from the hospitals to the University of Sharjah, Sharjah Institute for Medical Research (SIMR) Laboratories, for storage (-80°C) and analysis. During postpartum, at the 6^th month-visit, maternal (10 ml) and infant (5 mL) blood samples will be collected, processed and aliquoted in PHCs/MCHs’ laboratories and transported directly to the SIMR laboratories in an icebox for storage and analysis.

Breast milk profile: the assessment includes proximate analysis, fatty acid, oligosaccharides, HMW adiponectin, hs-CRP, TNF-α, IL-6 and insulin and vitamin D. Breast milk samples (10 mL) will be collected, upon delivery and during 6 months postpartum, using commercially available electrical breast pumps by the mother or the nurse, and placed in sterile polyproylene tubes, which ensures the collection of fore, mid, and hind milk within each sample. The samples will be directly transported to the SIMR laboratories for storage (-80°C) and analysis.

Perinatal information

Perinatal information was obtained from the mother’s medical records at the hospital in which the delivery took place. These include occurrence of complications and GDM during pregnancy and delivery, delivery method, gestational age, date of delivery, sex, weight, length, and head circumference of infants.

Quality assurance procedures

Several quality control measures are undertaken to ensure validity and reliability of collected data. Prior to data collection, collectors followed extensive training on study procedures, standardization of data collection and minimization of interviewer bias, and were provided with a comprehensive operation manual. This training was done in collaboration with experts from the Department of Nutrition and Food Sciences at the American University of Beirut (AUB). Manuals describing data collection procedures were developed and administered to participating centers. Data collection instruments were identical in all sites and were calibrated. Quality checks are routinely conducted for entered data by trained research assistants and one statistician. Finally, since birth cohorts often face planning and implementation challenges [56,57], the MISC research team sought to document encountered challenges and addresses them in a systematic manner. The challenges were noted through weekly meetings with the research assistants, and discussions with the participants and the nurses in the PHCs/MCHs, and responding strategies were planned based on researchers’ experiences and implemented after discussions within the research team, consultations with key scholars and support by relevant literature.

Statistical analysis

Collected data are entered and analyzed using the Statistical Package for the Social Sciences (SPSS) software version 21.0. Six categories of variables are planned to be generated: maternal factors, pregnancy complications, birth outcomes, dietary intake, feeding practices and infant factors. The normality of data for all variables will be examined using Shapiro-Wilk test. For normally distributed continuous variables, means and standard deviations will be presented, and Student’s t test will be used. For skewed data, medians and interquartile ranges will be generated and Mann-Whitney U test will be performed. For categorical variables, frequencies will be presented, and Pearson Chi-square and Fisher’s Exact tests will be used, as appropriate. The associations between the variables will be examined using multiple linear regressions when the outcome of interest is continuous, and logistic regressions when the outcome considered is categorical. As a first step, the outcome variables will be cross-tabulated against potential confounders; those with p-values <0.2 will be entered into the model. A p-value <0.05 indicates statistical significance.

Finally, a comparative analysis is planned for baseline sociodemographic characteristics and exposures of participants who remained in the cohort and those who were lost to follow-up. If both groups are found to be similar, it is highly plausible that participants who finished the study and based on whom the conclusions will be generated are representative of the population who enrolled in the cohort. If significant differences are found with respect to main assessed factors, minimum requirements for addressing attrition will be reported [58] and alternative analytical approaches will be used to handle missing values of participants who were lost to follow-up [59].

Participants’ characteristics

420 pregnant women in their third trimester were invited to participate, and 266 accepted to enroll (63.3%). As reasons for non-enrolment, our record showed that 36% (55/154) were not interested, 27% (42/154) declined due to the long duration of the cohort, and 20% (31/154) refused to enrol due to husband disapproval, while the remaining 17% (26/154) did not enrol because of blood collection. Ten mothers, who initially agreed to enroll, had to be excluded, since they had to leave before completing the baseline questionnaire due to time constraints. They did not differ in their characteristics from the other participants. Only 256 mothers completed the baseline questionnaire. Their baseline characteristics are shown in Tables 2 and 3. The mean age of the participants was 30.5±6.0 years, more than half of them were non-Emirati and the majority were university graduates and not employed. The majority of our starting sample was multiparous (76.6%) and up to 60% were either overweight or obese before they were pregnant. All except 7 were non-smokers (97.2%); however, 105 (40.5%) were subject to passive smoking. The prevalence of GDM was 19.2%. Between inclusion and delivery, 48 women were lost to follow-up, resulting in 208 women-infant pairs (dropout rate: 18.7%). The mean gestational age at delivery was 38.7±1.5weeks. One-third of the infants were delivered by caesarean section, and about half of the newborns were boys. Finally, 5.3% and 5.7% of pregnancies resulted in low birthweight and macrosomia, respectively (Table 4).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Baseline sociodemographic characteristics of study participants.

https://doi.org/10.1371/journal.pone.0198278.t002

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Baseline clinical characteristics and practices during pregnancy of study participants.

https://doi.org/10.1371/journal.pone.0198278.t003

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 4. Baseline delivery and infant characteristics.

https://doi.org/10.1371/journal.pone.0198278.t004

Challenges and responding strategies

First, we underestimated the number of required staff and we faced a high turnover of trained research assistants that could not be easily replaced. This resulted in supplementary training sessions for newly recruited agents, time delays, and additional costs. Furthermore, data collection used interviewer-administered questionnaires, which is resource-intensive, in terms of human capital, cost and time. Second, the recruitment was limited by the participants’ concerns of study duration or lack of interest. Also, some mothers were reluctant to enroll before obtaining their husband’s approval. To develop and maintain confidence and familiarity between the participants and research team, we established a nurse-liaison in each participating center, where a key nurse acted as a facilitator between involved parties (research assistants, PHCs/MCHs administration, and participants). Third, data collection at delivery was extremely challenging. Same-day discharge and shortened post-delivery hospital stays offered limited time to collect information and samples from mothers and neonates at the hospital. I response, we adopted a daily tracking of participants approaching their due dates. Despite this, some women were missed at delivery or discharge. Participant retention was also critical. Common reasons for attrition were the inability to locate participants usually due to disconnected communication, lack of interest of some participants and husbands’ disapprovals. Several actions were taken to enhance retention, such as offering incentives (e.g. gift vouchers, infant toiletries…), obtaining other contact detail information for participants who consistently did not answer phone calls, and maintaining communication with the participants through phone calls or text messages. Moreover, we highlight the importance of discussing and negotiating the research details with the husbands in order to enhance their understanding of the study’s significance, and their approval to their wives’ participation.

Blood collection, especially from infants, was our biggest anticipated challenge. It was sometimes surrounded by technical difficulties and parents’ refusals. When needed, specialized pediatric nurses were hired to perform blood withdrawal. Finally, maintaining the quality of collected data is an ongoing challenge, requiring a rigorous information system, periodic data tracking and quality check, and unanimous adaptation of pre-specified data collection and management protocols (Table 5).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 5. The MISC: Challenges and responding strategies.

https://doi.org/10.1371/journal.pone.0198278.t005

Strength and limitations

The MISC is a longitudinal study integrated within the healthcare system, rather than operating in the idealized experimental environment, consequently allowing the generation of “real life” data [68] in an ethical and practical fashion. The study entails comprehensive and systematic prospective quantitative assessments of mothers and children. It provides an all-inclusive analysis of the determinants of pregnancy, birth, mothers and children’s hard and proxy outcomes, without recurring to participant recall, and allows the suggestion of a clear temporal sequence between exposures and outcomes. Another strength of this study is the adoption of culturally-adapted data collection tools. Furthermore, the interviewer-administered questionnaires used in this study are expected to provide consistency and clarity in the administration and understanding of questions, unlike self- administered questionnaires.

The MISC aims to explore new perspectives in “fetal programming” and management of NCDs. Specifically, the MISC is first to explore the effect of early-life exposures on the cognitive development of young Arab children and is expected to enrich existing international evidence [69–71]. In addition, the MISC addresses numerous novel biomarkers, and pioneers in exploring these biomarkers specifically in children, by correlating maternal milk and blood biomarkers with infant blood biomarkers and immediate outcomes, in people with Arab ethnicity. Accordingly, results from the MISC complement the international evidence on the association between maternal and infant biomarkers and their effects on infant physiology [72–74]. In particular, the MISC will evaluate the specific short-term effects of lactation on maternal body composition (adiposity), as well as maternal and child cardio-metabolic homeostasis, and anti- and pro-inflammatory biomarkers. If lactation is found to exert beneficial metabolic and cognitive effects, then the translation of the findings from the MISC may have a significant public health impact. Lactation promotion is expected to be a low-cost, feasible strategy for the prevention of NCDs and promotion of intergenerational health [75,76]. The findings of MISC may be applicable to other country members of the Gulf Cooperation Council, exhibiting similar cultural, social and economic characteristics, and are similarly struggling to contain and mitigate the NCDs epidemic. Finally, the MISC design and measurements are consistent with those of other regional and international birth cohorts. This is expected to facilitate comparisons across-studies and improve the power to identify predictors of health and disease among infants.

As for the limitations, first, since mothers were not randomly assigned to different exposure groups, unequal distribution of confounders between-groups of comparison, and inherent dissimilarities relating to the mothers’ choice of specific behaviors might lead for possible biases. Therefore, we collected data relating to numerous potential confounders, and we planned to control for them through statistical analyses. This shall minimize bias from reverse causation or unmeasured confounding and might provide a robust alternative to the randomization of mothers which is unlikely to be feasible or ethical. Second, we could not collect information about non-enrollees because they did not consent to participate in the study. It would have been interesting to compare their characteristics with the participants to identify potential biases in recruitment. However, we do not expect to have major differences regarding their characteristics that could influence our results. Future cohorts could benefit from our experience in this regard, and not oversee this limitation. Furthermore, differential loss to follow-up could introduce bias. Subsequently, we planned for a comparative analysis of baseline and last-observation characteristics between women (and their children, where applicable) who attrited and those who completed the cohort. Finally, the exclusion of non-Arab women limits our ability to generalize our results to the population residing in the UAE. However, we only included Arab women to avoid any ethnic variability in the evaluation of adiponectin, as well as its relationship with other biomarkers [77–80]. We prioritized this issue over the possibility of generalizing our findings, taken the lack of available evidence related to adiponectin.

Finally, the implementation of this cohort was subject to many logistical, methodological and sociocultural challenges that required tailored remedial strategies. Many of these challenges are common to birth cohorts [56,81,82]. At this stage, we recommend that the complex ethical issues involved in conducting a birth cohort, must be carefully considered. Moreover, the necessary start-up time, staff, and costs must not be underestimated and funding of extra costs relating to logistical complexity and added expenses must be accounted for. Moreover, it is recommended that future cohort studies actively include the father as a participant, and examine the paternal factors that contribute to NCDs in infants. This is suggested to strengthen the study scientifically and logistically.