Study design and study population

This study is part of The ORIGINS Project (ORIGINS). ^{Reference Silva, Hagemann and Davis38} The ORIGINS Project commenced in January 2017 and is a decade-long collaborative initiative between the Joondalup Health Campus and the Telethon Kids Institute to establish a Western Australian birth cohort of 10,000 families enrolled in pregnancy over a 5-year period, creating a major research platform with an extensive biobank and databank. ORIGINS aims to collect detailed information about the influence of the early environment’s influences on a broad range of non-communicable diseases through the collection of biological samples (such as blood, urine, stool, cheek swabs, and breastmilk), online questionnaires, and routine collected data. ^{Reference Hagemann, Colvin, Silva and Prescott39}

This study focused on a subset of 458 maternal participants in their third trimester of pregnancy, and their newborns. The eligibility criteria for participants (pregnant women) in this study were as follows: (1) participant who consented in The ORIGINS Project; (2) singleton pregnancy; (3) birthing at Joondalup Health Campus; (4) completed online questionnaire during their third trimester of pregnancy; (5) newborns ≥37 weeks’ gestation who obtained a PEA POD measurement. With all inclusion criteria met, data from a total of 458 participants were analysed. Approval from the Ramsay Health Care WA|SA Human Research Ethics Committee (RHC WA|SA HREC) and previously the Human Research Ethics Committee of Joondalup Health Campus (JHC HREC) was obtained prior to commencing this study.

Maternal and newborn information

Birth measurements including APGAR score, newborn length, birth weight, and head circumference were obtained by the attending midwife and stored on the JHC Meditec hospital patient information system, along with information on other birth outcomes, ethnicity, and antenatal risk factors. Socio-Economic Indexes for Areas (SEIFA)/Index of Relative Socio-Economic Advantage and Disadvantage (IRSAD) scores were calculated based on the Australia Bureau of Statistics data cubes which uses participants Postal Area Code. ⁴⁰ Maternal pre-pregnancy BMI was not a routinely collected measure; however, this was available on 54% of this study population.

Data collection

The Mediterranean Diet index (MDI) consisted of 13 questions, which corresponded to receiving a positive score if patients consumed (1) high ratio of monounsaturated fats : saturated fats, (2) high olive oil intake, (3) high vegetable intake, (4) high fruit intake, (5) low intake of meat and meat products, (6) low intake of saturated fats, (7) low intake of sweetened carbonated beverages, (8) high legume intake, (9) high fish or seafood intake, (10) less sweets/pastry intake, (11) high intake of nuts, (12) higher consumption of white meat: to red meat ratio, and (13) high sofrito (tomato/onion based) sauces used when cooking. In general, the medians of the sample, specific for sex, were used as cut-off points ^{Reference Costacou, Bamia, Ferrari, Riboli, Trichopoulos and Trichopoulou41} and grams per day were used as the intake measurement. ^{Reference Trichopoulou, Kouris-Blazos and Wahlqvist33} A participant received a point if intake was over the sample median for a protective component (vegetables, fruits, etc.) and below the median for non-protective components (dairy products, meat, etc.). A low-moderate alcohol intake has been associated with higher MDA in the vast majority of population studies; however, as alcohol consumption in pregnancy is associated with increased risk of miscarriage, stillbirth, and premature birth, ^{Reference O’Leary42} abstinence from alcohol is recommended in this population hence excluded from this questionnaire. Calculating MDA in this study was based on a similar format to the multicentre PREDIMED trial which used the 14-item MD assessment tool. ^{Reference Martínez-González, García-Arellano and Toledo43} MDA was categorised into three categories based on the overall score from 13 questions (as alcohol intake was excluded from the maternal questionnaires): with low (score 0–4), medium (score 5–7), and high (score 8–13) MDA.

The PEA POD is an ADP system which uses whole-body densitometry to determine body composition in infants weighing between 1 and 8 kg. The PEA POD has been widely used in research as it is easy to use, provides immediate results, is non-invasive, and has been validated for both term and preterm infants against deuterium dilution. ^{Reference Ellis, Yao, Shypailo, Urlando, Wong and Heird44} To determine body composition, the infant is first undressed, and their mass is measured using an integrated scale. The mass and volume calibrations take into account irremovable items from the neonate (two hospital identification bracelets and an umbilical cord clip). The percentage body fat is then computed by a software to the PEA POD system (version 3.0.1) which estimates body composition by whole-body densitometry using a two-compartment model – fat and fat-free compartments. Within the volume calculation, corrections are applied for air within the lungs and in close proximity to the participants skin, which behaves isothermally. The infants percentage body fat is then calculated from body density assuming the density of fat to be 0.9007 kg/L. Age- and gender-specific densities of free-fat mass (FFM) are computed based on the data of Fomon et al. ^{Reference Fomon, Haschke, Ziegler and Nelson45} Once body volume is known and whole-body density is calculated, fat mass (FM) and FFM are estimated. ^{Reference Villar, Puglia and Fenton46} Infants length was measured from heel to crown using an infant length board. Data on the infant were then recorded and incorporated into PEA POD measurements, including birth weight, length, gender, ethnicity, and gestational age using participants date of birth. The PEA POD is routinely calibrated by trained ORIGINS Project staff and is housed in a temperature-controlled room. PEA POD measurements were taken on infants within 7 days post-birth and results were uploaded into The ORIGINS Project Databank. The PEA POD results were based on the estimates of Fomon et al., which uses a number of datasets and assumptions to describe a ‘typical’ boy weighing 3.5 kg and a girl weighing 3.35 kg. ^{Reference Fomon, Haschke, Ziegler and Nelson45}

Data analysis

Data were analysed between maternal MDA (high, moderate, and low adherence categories) in the thirdd trimester, and maternal characteristics, birthing characteristics, and neonatal outcomes. Neonatal outcomes included infants’ weight, length, head circumference, and PEA POD body fat percentage measured ≤7 days post-birth.

Data are presented as frequencies and percentages for categorical data and were analysed using chi square or Fisher’s exact test (when expected cell counts <5). For continuous data, means and standard deviations are reported and associations with MDA analysed using linear regression. For the linear regression models, the continuous variable was the outcome and MDA entered in the model as a categorical predictor. Comparisons were made between each MDA group using mean differences and their corresponding 95% confidence interval. Firstly, the regression analysis was performed using univariable (unadjusted) regression and later adjusted for pre-pregnancy BMI, infant sex, ethnicity, and gestational diabetes. There is a substantial drop in sample size in the adjusted models due to incomplete data for BMI; however, as BMI and infant sex are important confounders, we have chosen to still present these results. All data were analysed using Stata 16.1 (StataCorp, College Station, TX). Statistical significance was considered p < 0.05.