UC Davis

Choline is an essential micronutrient that has been positively associated with child growth and development in human and animal studies. Suboptimal choline intake is thought to be common in low and middle income countries (LMICs) because of the relative expense of animal source foods, the major sources of choline. Few studies have examined the role of choline for growth and development in LMICs (reviewed in Chapter 1). This dissertation aimed to help fill this gap using data from the Mazira Project randomized controlled trial, in which 660 children age 6-9 months from rural Malawi were randomized to receive either 1 egg per day or a nonintervention control for 6 months. Despite large improvements in growth from a similar egg intervention trial in Ecuador, which were mediated by improvements in plasma choline, there were few improvements to growth or development with the provision of 1 egg per day in the Mazira Project. This dissertation includes three secondary analyses that explore the role of plasma choline in the Mazira Project. In the first analysis (Chapter 2), we examined the effect of the egg intervention on plasma choline and several of its metabolites. Plasma choline, betaine, dimethylglycine (DMG), trimethylamine N-oxide (TMAO), and docosahexaenoic acid (DHA) were measured from blood samples collected at enrollment and 6 month follow up. Semi-quantitative lab analyses were conducted in 200 children per group using UPLC-MS/MS, and quantitative lab analyses were measured in a subsample of 60 children using LC-MS/MS. Mean plasma choline values using the quantitative data (n=60) were ~17.0 µmol at enrollment and decreased to ~14.0 µmol at 6 month follow up. Plasma choline, betaine, DMG and DHA were not significantly different by group at 6 month follow up in minimally or fully adjusted linear regression models using the semi-quantitative data (n=200 per group). However, plasma TMAO was significantly higher (26% [SD 7%, 48%]) in the intervention group at 6 month follow up. In the second analysis (Chapter 3), we investigated the association between plasma choline and child development. Plasma choline, betaine, DMG, and TMAO were measured as described in Chapter 2. Additionally, measures of child development were collected at enrollment and 6 month follow up. Fine motor, gross motor, language, and personal social normed z-scores were calculated using the Malawi Developmental Assessment Tool (MDAT). Visual attention was measured by response times in an Infant Orienting with Attention (IOWA) task and peak look lengths in a visual paired comparison (VPC) task. Memory was assessed by novelty preference in a VPC task and the number of actions recalled during an elicited imitation task (collected at 6 month follow up only). Plasma choline was not associated with most measures of development in fully adjusted generalized linear models, except a negative association with MDAT fine motor normed z-score (-0.13 SD [95% CI -0.22, -0.04]) and a positive association with IOWA response time (8.84 ms [1.66, 16.03]), both suggesting poorer development with higher plasma choline concentration. Associations with the related metabolites (betaine, DMG, and TMAO) were null, except positive associations of plasma TMAO with peak look length and total actions recalled score. In the third analysis (Chapter 4), we examined the association between plasma choline and child growth in the Mazira Project. Child length, weight, and head circumference were measured by trained anthropometrists at enrollment, 3 month follow up, and 6 month follow up, then converted to z-scores (length-for-age (LAZ), weight-for-age (WAZ), weight-for-length (WLZ), and head circumference-for-age (HCAZ)) using WHO Growth Standards. Dichotomous outcomes were calculated using cutoffs of -2 SD. Conditional outcomes were calculated by regressing data from the 6 month follow up on data from the previous time points. The association of plasma choline with continuous and dichotomous growth outcomes was assessed using generalized linear models. Conditional growth outcomes were included in linear regression models, with baseline plasma choline as a predictor. Plasma choline was positively associated with WLZ in a minimally adjusted model, but this was not significant after further adjustment for covariates. Additionally, plasma choline was weakly and negatively associated with LAZ in a fully adjusted model (-0.09 [-0.17, -0.01]). There were no significant associations of betaine, DMG, or TMAO with child growth indicators. Together, the analyses presented in this dissertation provide a partial explanation for the null results in the Mazira Project. The egg intervention was hypothesized to work in part by improving plasma choline status, which would then lead to improved growth and development. However, the intervention did not significantly improve plasma choline, and plasma choline was not associated with most measures of growth or development. These analyses also suggest that the relationship between choline, development and growth may be context specific, perhaps depending on children’s background diet as well as socioeconomic and health factors. Finally, this dissertation highlights the need for sensitive and specific biomarkers of choline status, and their application in the study of young children in LMICs. At this time, choline intake at the Adequate Intake level (AI for 7-12 months: 150 mg/d; 1-3 years: 200 mg/d) should be recommended for young children.