Sleep disturbance will be assessed using the Sleep Disturbance Scale for Children (SDSC), consisting of 26 items, rated on a 5-point Likert scale. The total score ranges from 26-130 with higher scores indicating higher levels of sleep disturbance. Besides the total score, 6 subscales representing more specific sleep disturbance will be calculated, i.e., difficulty in initiating and maintaining sleep (DIMS); sleep breathing disorders (SBD); disorders of arousal (DoA); sleep-wake transition disorders (SWTD); disorders of excessive somnolence (DOES); sleep hyperhidrosis (SH). The SDSC will be completed by participants' parents within 72 hours prior to the at clinic visits at months 0, 2, 4, and 6.

Levels of physical activity over the past week will be assessed using the Physical Activity Questionnaire for Older Children (PAQ-C). The PAQ-C consists of 10-items, with items 1-9 being rated on 5-point Likert scales, scored from 1-5. The overall activity summary score is represented by the mean of items 1-9, thus ranging from 1 to 5, with higher scores indicating higher levels of physical activity. The PAQ-C will be completed by participants within 72 hours prior to the at clinic visits at months 0, 2, 4, and 6.

Mental health disorder symptoms will be assessed via the Development and Well-Being Assessment (DAWBA), an online questionnaire/interview package indicating likely DSM-5 psychiatric diagnoses. Participants and parents will complete the online questionnaires within 72 hours prior to Visit 1 and Visit 4. The DAWBA covers Separation anxiety, Specific phobia, Social phobia, Panic disorder / agoraphobia, Post-traumatic stress disorder, Obsessive compulsive disorder, Generalised anxiety disorder, Body dysmorphic disorder, Disruptive mood dysregulation disorder, Major depression, ADHD / hyperkinesis, Oppositional defiant disorder, Conduct disorder, Eating disorders, including anorexia, bulimia and binge eating, Autism spectrum disorders, Tic disorders, including Tourette syndrome, Bipolar disorders. Diagnostic predictions are given at probability bands, i.e., (1) less than 0.1% (--), (2) around 0.5% (-), (3) around 3% (+/-), (4) around 15% (+), (5) around 50% (++), (6) above 70% (+++).

Cognitive assessment will be conducted using the Cambridge Neuropsychological Test Automated Battery (CANTAB), a cognitive assessment research software, administered via a touchscreen device. The battery for participants with a diagnosis of Autism consists of measures examining executive functions, planning, episodic memory, and processing speed. CANTAB is language independent and requires no prior familiarity with computers. It will be administered at clinic during Visit 1 and Visit 4.

Nutritional assessment is based on an adapted version of the Nutritional Assessment for Children and Adults with ASD (NACA-ASD). The NACA-ASD was designed to assist families and researchers in making a rough estimate of the quality of diet and nutritional supplementation of a person with Autism. Participants will be asked to estimate the number of servings of different food groups they have eaten over the past 24 hours. Assessment will be conducted at clinic during all visits.

Heart Rate in Beats per Minute (BPM) will be assessed via the noninvasive Vivalink wearable electrocardiogram (ECG) patch (ECG sampling 128 Hz; Heart rate range 40-300 BPM). Participants will be asked to wear the sensor for 5 days in the first 2 weeks following the baseline visit (i.e., 10 days total). Subsequently, they will be asked to wear the sensor for 72 hours every other week over the study course.

Respiratory rate will be assessed via the noninvasive Vivalink wearable electrocardiogram (ECG) patch. It is collected in Breaths per Minute (BrPM) with a range of 5 to 25 BrPM. Participants will be asked to wear the sensor for 5 days in the first 2 weeks following the baseline visit (i.e., 10 days total). Subsequently, they will be asked to wear the sensor for 72 hours every other week over the study course.

Infinium Global Screening Array-24 BeadChip with DNA from buccal swab at visit 1 will be utilised. Bead array content is selected for imputation accuracy at minor allele frequencies of >1% across Genomes Project populations. Clinical research content includes variants with established disease associations, relevant pharmacogenomics markers, and curated exonic content based on databases. After intensity calling, sample quality control includes removing samples with genotyping call rate <98% and SNPs with a call rate <98% or Hardy-Weinberg equilibrium p-value <5 × 10-6. For imputation, array-derived genotypes were prephased using SHAPEITv2 and imputation was carried out using IMPUTE2 software with 1000 Genomes phase 3 reference panel. SNPs with minor allele frequency <.10 and an IMPUTE2 'info' matric <.9 were excluded to ensure max. confidence in imputation quality. Each SNP will be coded 0,1,2; associations with environmental factors and mental health will be examined.

DNA methylation will be assessed using the Illumina Infinium MethylationEPIC v2.0 BeadChip array from DNA extracted from buccal swabs collected at each visit. The EPIC array is a genome-wide methylation screening tool that targets over 935,000 CpG sites in the most biologically significant regions of the human methylome. Infinium 850K data will be processed using the Bioconductor package minfi in R (version 3.4.2). Beta-mixture quantile (BMIQ) normalization will be used to remove array biases and correct for probe design. DNA methyaltion will be a continous measures with values between 0 and 1. Robust regression models using limma will be run to assess associations between DNA methylation and the measured environmental factors, genotype and mental health outcomes together with changes in DNA methylation over time.

Hormone levels in urine will be assessed using ELISAs and Mass spectrometry from samples taken at each visit. These methods will give ug/ml measures of the different hormones assessed. Hormone levels will then be compared to the measured environmental parameters, DNA methylation, metabolome, and mental health outcomes.

Metabolic profiles of urine and faecal samples will be measured by nuclear magnetic resonance (NMR) spectroscopy (700 MHz Bruker spectrometer equipped with a cryoprobe and autosampler) and liquid-chromatography-mass spectrometry (LCMS) (Waters Premier LC system hyphenated to a Waters Synapt Q-TOF). These approaches will be applied in an untargeted manner to measure a broad range of biochemical classes (e.g., amino acids, sugars, organic acids, vitamins, aromatic compounds). Multivariate metabolic profiles will be processed and analysed using bespoke pipelines in Matlab and R. Statistical approaches used will include, but will not be limited to, principal components analysis, projection to latent structures, self-organising maps, DIABLO, and random forests. These methods will be employed to highlight metabolite features and data structures related to mental health, environmental parameters, DNA methylation and genotype data.