Developmental Origins of Mental Health Disorders

Common mental disorders (CMD) such as Depression, contributes significantly to the global burden of disease. Fetal exposure to adverse intrauterine environment mediated by life course factors can enhance risk of non-communicable disease in later life. Maternal micronutrients such as Vitamin B12 and folate play an important role in early fetal development through their effect on one carbon metabolism.

Vitamin B12 deficiency is common in Indian women; however guidelines recommend only iron, folic acid supplementation during pregnancy. This study aims to investigate effects of maternal B12, folate during pregnancy on mental health and neurocognitive outcomes in offspring during adulthood.

The Pune Maternal Nutrition Study(PMNS) birth-cohort(n=762) was established in 1993 at the Diabetes Unit of KEM Hospital Pune with well characterized serial data and archived biological samples. The subjects of the cohort are now in age range of 20-22 years and this provides an opportunity to examine the proposed objectives.

Primary objectives:

To examine the specific association between maternal vitaminB12, folate, homocysteine levels(independently and in interaction with each other) at 18 & 28 weeks of gestation and neurocognitive outcomes and risk for CMD at around 21 years of age, controlling for effect of other nutrients (calorie, protein and fat intake, and levels of vitamin C,D, ferritin).

Hypothesis: That lower maternal vitamin B12-higher folate, and higher homocysteine levels at 18 and 28 weeks of gestation will be associated with lower cognitive functioning scores on neuropsychological tests and higher prevalence of CMD in men and women aged 20-22 yrs.

This nutrient pattern (lowB12, high folate, high homocysteine) in the mother will be associated with:

• Lower cortical thickness, smaller total brain, hippocampal volumes and altered hippocampal-dorsolateral prefrontal cortex functional connectivity.
• Disturbed temperament and character dimensions in internalizing and externalizing spectrum.

In addition the mediating and modifying effects of other life course factors like childhood growth and nutrition, education, socioeconomic status, early life stress, lifestyle on risk for CMD will be examined.

Secondary objectives:

• To test the causality of association between maternal VitaminB12, folate, homocysteine and neurocognitive outcomes by Mendelian Randomization analysis(triangulation and instrumental variable analysis) using maternal genetic determinants of vitaminB12 (TCN2rs1801198, FUT2rs492602) and homocysteine(MTHFRrs1801133).
• To assess whether the above associations are mediated by levels of biomarkers of neurodevelopment(serum Brain Derived neurotrophic factor BDNF, Insulin like growth factor IGF-1 levels) measured at 6,12&18 years of age in the cohort.
• To examine whether later life nutritional supplementation (Vitamin B12 with or without multiple micronutrients or placebo) can modulate the effect of maternal B12 status on neurocognitive outcomes in young adulthood.

Methods Study design Subjects from the longitudinal prospective cohort of the PMNS (n=690) will be recruited for cross sectional assessment. The time required for assessments will be approximately 2-3 hours (excluding neuroimaging). Female members of the cohort who are currently pregnant at time of assessment would be excluded from neuroimaging, however they would be screened for CMD.

Procedures:

The cohort participants will be approached by the field workers who have been following them for last many years. Subjects will be provided the study information sheet and an opportunity to discuss their doubts and concerns regarding the study with the investigator. Subjects giving informed consent will be recruited.

Research staff (one psychiatrist and one clinical psychologist) will be trained and monitored in the administration of the assessment of the tools and inter-rater reliability will be periodically examined. The subjects will be examined at the Diabetes unit at KEM Hospital where the following assessments' will be performed.

• General socio-demographics and a structured interview:

Semi structured proforma will be used to record current demographic details, and socio-economic status (assessed on the Standard of Living Index which provides information on occupation, residential living arrangement, educational status, possessions), Life style risk factors (Physical activity, stress - Cohen's perceived stress scale, substance use, high risk behaviors, early life stress - World health organisation Adverse Childhood Experiences International Questionnaire). Family history of mental illness will be assessed on the Family interview for genetic studies.

• Physical examination A general physical examination would be performed to assess, height, weight, Body mass index BMI, signs of any nutrient deficiencies, resting pulse rate and blood pressure.
• Neurocognitive assessment A Standardized battery of neuropsychological tests will be administered to assess the domains of executive functions, working memory, intelligence, verbal memory. The tests will take 60 minutes to administer

• Raven's standard progressive matrices (SPM): This is a measure of intelligence which can be used in adults. Test contains 60 items in 5 sets each containing 12 items. Number of correct responses are recorded. Test produces a single raw score and percentile rank
• Color trail test: This is a test of executive functions. It involves of 2 parts A & B where subjects have to connect numbered circles from 1 to 25 followed by connecting numbered circles in alternating colors (yellow and pink). Time to complete task and number of errors are recorded. Longer time indicates poor performance.
• Digit span test: This is a test of working memory. Increasing sequence of digits is presented which the subject has to recall forwards and then backwards. The longest sequence of words correctly repeated is recorded.
• Block design: This is a test for visuo-spatial ability. Task is to replicate designs shown on cards using red and white blocks. Time taken for completing test is recorded. Longer time indicates poor performance.
• Word recall: WHO/University of California Los Angeles(UCLA) version of Rey's auditory verbal learning test measures verbal learning and memory. Test involves immediate and delayed recall of list A of 15 nouns with a interference presented with list B. Words used have been translated into local language. Number of words correctly recalled, are recorded.
• Psychological assessments:

Temperament and Character Dimensions: This would be assessed on the short Temperament and Character inventory - R. This is a 140 item self-report questionnaire that yields scores on 4 temperament dimensions (Harm avoidance, reward dependence, novelty seeking, persistence) and 3 character dimensions (self-directedness, cooperativeness and self-transcendence).

• Mental Health assessment -diagnostic All subjects will be assessed on a structured clinical interview (Mini Neuropsychiatric interview plus (MINI) 6.0; which yields syndromal psychiatric diagnosis based on the Diagnostic and statistical manual (DSM) 5. CMD of interest assessed would be depression and anxiety disorders. All subjects meeting criteria on the MINI will be reassessed by comprehensive clinical mental status examination by 2 qualified psychiatrists to confirm the diagnosis. Any subject found to have syndromal psychiatric diagnosis would be offered treatment as per standard clinical protocols.
• Magnetic resonance imaging MRI of brain for structural sequences and resting state fMRI would be performed on a 3Tesla MRI scanner. The MRI scans would be performed on subset of cohort selected based on their maternal vitaminB12 levels using a nested case control study design. 146 members have a maternal vitaminB12 level at 18 weeks of <103pmol/l. Age and gender matched Control group would be selected from remaining members of cohort whose maternal vitaminB12 levels are >175pmol/l (n=175). Accounting for attrition an estimated number of 200 subjects (100 in each group) will be scanned. The following MR sequences would be performed - Structural 3D T1 weighted MPRAGE (magnetization prepared rapid gradient echo), fMRI with 2D gradient-echo EPI sequence.
• Lab assessments:

Assessments for serum BDNF, IGF-1 will be performed on archived serum samples of cohort subjects collected at 6,12,18 years of age using commercially available ELISA kits. The analysis will be performed in the lab at Diabetes Unit, KEM Hospital which participates in the United Kingdom National External Quality Assessment Service (UKNEQAS).

Power estimation & Statistical analysis:

There are 690 members of the cohort available for assessments. Assuming a 90% participation rate (from experience of previous studies on the cohort) we estimate that 621 subjects would be recruited for the study. Using a test at 5% significance level, the study will have 80% power to detect and association of 0.11 standard deviation of continuous outcome (eg. neurocognitive test score) per standard deviation of continuous exposure (eg. maternal vitaminB12 level) Data on pre-pregnancy maternal size(height, skin fold thickness, head circumference) nutrition status during pregnancy(nutrition intake, circulating micronutrients, physical activity) newborn size(birth weight, length, head circumference) childhood nutrition, growth and development (BMI, body fat measurements, skin fold thickness) and cardiometabolic risk factors (insulin resistance, glycemic status, adiposity) assessed previously on the cohort are available and will be extracted.

Regression modeling technique would be primarily used to test the proposed hypothesis. More specifically, the neuro-cognitive outcomes would be modeled as a function of the exposure(s) after adjusting for the effect of the mediator and the confounding variables keeping in consideration the features of the data (non-linearity, multivariate response etc.). A similar technique would be adopted to assess the form and the strength of the relationship between Magnetic Resonance Imaging outcomes and outcomes of CMD with the exposure(s) respectively. Appropriate use of the linear models or generalized linear model would be made based on the nature of the outcomes. The effect sizes along with the confidence intervals and appropriate diagrams to visualize the effects/relationships would be performed.

Open Clinica software would be used for database design and data entry while the entire data cleaning and analysis and visualization would be done using the R statistical software.

Mendelian randomization analysis (triangulation and instrumental variable analysis) using maternal genetic determinants of vitamin B12 (TCN2,FUT2) and homocysteine (MTHFR)genes will be performed to examine causal role of vitaminB12 and one carbon metabolism on neurocognitive outcomes. Genetic analysis for vitaminB12 (TCN2rs1801198, FUT2rs492602) and homocysteine (MTHFRrs1801133) genes have been performed on 526 mothers and children of the cohort and are available.

Inclusion criteria

• male or female adult member of the PMNS cohort
• Availability of data on maternal vitamin B12 and folate levels during 18&28 weeks of pregnancy for the subject
• Subject should have completed longitudinal assessments at 6, 12 and 18 years of age

Exclusion criteria

• Pregnancy
• Contraindications for MRI such as metallic orthopedic or dental implants, claustrophobia.