Trial of Thiamine Supplementation in Cambodia

Improving Estimates of the Global Burden of Thiamine Deficiency Disorders and Approaches to Their Control: Trial of Thiamine Supplementation in Cambodia

Sackler Institute for Nutrition Science, Bill and Melinda Gates Foundation, Helen Keller International, NCHADS - Ministry of Health of Cambodia, Ministry of Planning, Cambodia, South Australian Health and Medical Research Institute, Institut de Recherche pour le Developpement, University of Oregon

Beriberi is a potentially fatal disease caused by vitamin B1 (thiamine) deficiency that still occurs in Southeast Asia despite near eradication elsewhere. Mothers with a diet low in thiamine produce thiamine-poor milk, putting their infants at a high risk of developing thiamine deficiency and beriberi. There is also a growing body of evidence suggesting thiamine deficiency not severe enough to cause clinical symptoms may negatively effect cognitive development and functioning of the infant. Since human milk should be the sole source of nutrition for babies during the first six months, maternal thiamine intake must be improved to combat this disease. The investigators' recent study of thiamine-fortified fish sauce in Cambodia showed that fortification could increase maternal and infant thiamine status'. However, centrally produced fish sauce may not reach the poorest communities who make their own fish sauce, and fish sauce is not consumed in all regions where we find thiamine deficiency. Salt, by contrast, is a common condiment in most regions of the world and has proven to be a successful global fortification vehicle for iodine. Suboptimal maternal thiamine intake puts exclusively breastfed infants at risk of low thiamine status, impaired cognitive development, and infantile beriberi, which can be fatal. Thiamine fortification of salt is a potentially low-cost and sustainable means of combating suboptimal thiamine status; however knowledge gaps must be filled before thiamine fortification can proceed. In this study, mothers will consume thiamine supplements in order to model the thiamine dose required to optimize human milk thiamine concentrations for the prevention of beriberi. Other thiamine biomarkers will be assessed, and usual salt intake will be measured. Finally, the investigators will assess the effects of early-life thiamine exposure on infant neuro-cognitive development.

Opaque capsules containing varying amounts of thiamine hydrochloride and cellulose filler. All thiamine is delivered as thiamine hydrochloride, calculated using a 1.271 correction factor (ratio of molecular weights of thiamine hydrochloride and thiamine).

To estimate the dose on the dose response curve where additional maternal intake of thiamine (oral dose) no longer meaningfully increases human milk total thiamine concentration at 24 weeks postpartum.

To estimate the dose on the dose response curve where additional maternal intake of thiamine (oral dose) no longer meaningfully increases infant thiamine diphosphate concentrations (ThDP) 24 weeks postnatally, and assess whether this depends on the presence/absence of a genetic hemoglobin disorder.

To estimate the dose on the dose response curve where additional maternal intake of thiamine (oral dose) no longer meaningfully increases human milk total thiamine concentration at 4 and 12 weeks postpartum.

To estimate the dose on the dose response curve where additional maternal intake of thiamine (oral dose) no longer meaningfully increases infant transketolase activity at 24 weeks postnatally, and assess whether this depends on the presence/absence of a genetic hemoglobin disorder.

To estimate the dose on the dose response curve where additional maternal intake of thiamine (oral dose) no longer meaningfully increases maternal ThDP at 24 weeks postpartum, and assess whether this depends on the presence/absence of a genetic hemoglobin disorder.

To estimate the dose on the dose response curve where additional maternal intake of thiamine (oral dose) no longer meaningfully increases maternal ETKac at 24 weeks postpartum, and assess whether this depends on the presence/absence of a genetic hemoglobin disorder.

To test for differences between the 4 randomized groups on human milk total thiamine at 4, 12, and 24 weeks postpartum.

To test for differences between the 4 randomized groups on maternal ThDP at 24 weeks postpartum, and assess whether this depends on the presence/absence of genetic hemoglobin disorder.

To test for differences between the 4 randomized groups on maternal ETKac at 24 weeks postpartum and assess whether this depends on the presence/absence of a genetic hemoglobin disorder.

To estimate usual household salt intake from mean fortnightly salt disappearance (weight lost, in g).

To estimate salt intake among a subset of 100 lactating women, their male partners (if applicable), and their children 24-59 months (if applicable) using observed weighed salt intake records.

To estimate sodium intake using 24 hr urinary sodium concentrations among a subset of 100 lactating women.

To test for differences between the 0 & 10 mg randomized groups on Composite Mullen and the 5 subscales of the Mullen at 24 and 52 weeks postnatally. Score notes for Mullen Scales of Early Learning Gross Motor Scale: Raw score range: 0-36 (higher scores reflect better performance) Visual Reception Scale: Raw score range: 0-50 (higher scores reflect better performance) Fine Motor Scale: Raw score range: 0-49 (higher scores reflect better performance) Receptive Language Scale: Raw score range: 0-48 (higher scores reflect better performance) Expressive Language Scale: Raw score range: 0-50 (higher scores reflect better performance) All raw scores are converted to age-adjusted T-scores (provided by manual). An optional Early Learning Composite Score can be calculated by summing the T-scores of all but the Gross Motor Scale.

To test for differences between the 0 & 10 mg randomized groups on Visual Paired Comparison Novelty Score and the attention and processing speed subscales at 24 and 52 weeks postnatally. Score notes for Visual Paired Comparison task Scale ranges: An overall novelty score is calculated that averages across all face and pattern trials, and is a percentage reflecting the percent of time infants' looked at the novel item during the test phase, calculated as the duration looking to the novel item divided by the sum of the duration of looking to the novel item plus the duration looking to the familiar item. Sub-scales: Two sub-scales are calculated: the novelty score averaged across face trials, and the novelty score averaged across pattern trials. The total scale composite averages the novelty score across the two sub-scales. For all scales, he range is 0-100%, and higher novelty scores indicate better outcomes.

To test for differences between the 0 & 10 mg randomized groups on the Language Preference Task Score at 24 weeks postnatally.

To test for differences between 0 & 10 mg randomized groups on oculomotor scores at 24 weeks postnatally. Score notes for the Oculomotor Test: The proposed oculomotor test is not a single, published instrument, but instead, combines several standard screening items. Strabismus: absent or present; score: 0-1 Nystagmus: absent or present; score: 0-1 Amblyopia: absent or present; score 0-1 Saccade: guided saccade: absent or present; score 0-1 scanning: absent or present; score 0-1 predicted saccade: absent or present; score 0-1 Smooth pursuit: overall pursuit: absent or present; score 0-1 horizontal pursuit: score 0-1 vertical pursuit: score 0-1 Total oculomotor score (optional): sum of all scored exam items: range 0-9

To determine the effect of inflammation, as measured by C-reactive protein (CRP) and α-1-acid-glycoprotein (AGP) on maternal ThDP at 2 and 24 weeks postpartum, and infant ThDP at 24 weeks postnatal.

Inclusion Criteria: Mothers of a newborn who: are aged 18 - 45 years had a recent normal pregnancy (i.e. no known chronic conditions, no preeclampsia, gestational diabetes etc), and the singleton infant was born without complications (e.g. low birth weight (<2.5 kg), tongue tie, cleft palate) are intending to exclusively breastfeed for six months reside in Kampong Thom province, Cambodia, and are not planning to move in the next six months are willing to consume one capsule daily from 2 weeks through to 24 weeks postpartum are willing for her entire household consume only salt provided by the study team are willing for the following biological samples to be collected: a maternal venous blood sample and human milk sample at 2 weeks postpartum, a human milk sample at 4 and 12 weeks postpartum, and maternal and infant blood samples and a human milk sample at 24 weeks postpartum. Exclusion Criteria: Mothers of a newborn who: are currently taking or has taken thiamine-containing supplements over the past 4 months are currently participating in nutrition programs beyond normal care

Gallant J, Chan K, Green TJ, Wieringa FT, Leemaqz S, Ngik R, Measelle JR, Baldwin DA, Borath M, Sophonneary P, Yelland LN, Hampel D, Shahab-Ferdows S, Allen LH, Jones KS, Koulman A, Parkington DA, Meadows SR, Kroeun H, Whitfield KC. Low-dose thiamine supplementation of lactating Cambodian mothers improves human milk thiamine concentrations: a randomized controlled trial. Am J Clin Nutr. 2021 Jul 1;114(1):90-100. doi: 10.1093/ajcn/nqab052.

Whitfield KC, Kroeun H, Green T, Wieringa FT, Borath M, Sophonneary P, Measelle JR, Baldwin D, Yelland LN, Leemaqz S, Chan K, Gallant J. Thiamine dose response in human milk with supplementation among lactating women in Cambodia: study protocol for a double-blind, four-parallel arm randomised controlled trial. BMJ Open. 2019 Jul 9;9(7):e029255. doi: 10.1136/bmjopen-2019-029255.