key: cord-0284589-7700ywe5
authors: Stephenson, K.; Callaghan-Gillespie, M.; Maleta, K.; Nkhoma, M.; George, M.; Park, H. G.; Lee, R.; Humphries-Cuff, I.; Lacombe, R. J. S.; Wegner, D. R.; Canfield, R. L.; Brenna, J. T.; Manary, M. J.
title: Low linoleic acid foods with added DHA given to Malawian children with severe acute malnutrition improves cognition: a randomized, triple blinded, controlled clinical trial
date: 2021-09-12
journal: nan
DOI: 10.1101/2021.09.07.21263231
sha: ecfeab375c0a8acf24ef84cb4b4a311bddc05789
doc_id: 284589
cord_uid: 7700ywe5

Background: There is concern that the PUFA composition of ready-to-use therapeutic food (RUTF) for treatment of severe acute malnutrition (SAM) is suboptimal for neurocognitive recovery. Objective: We tested the hypothesis that RUTF made with reduced amounts of linoleic acid, achieved using high oleic (HO) peanuts, with or without added DHA, improves cognition when compared to standard RUTF (S-RUTF). Methods: A triple-blind, randomized, controlled clinical feeding trial was conducted among children with uncomplicated SAM in Malawi with 3 types of RUTF; DHA-HO- RUTF, HO-RUTF and S-RUTF. The primary outcomes, measured in a subset of subjects, were the Malawi Developmental Assessment Tool (MDAT) global and 4 domain (gross motor, fine motor, language and social) z-scores and a modified Willatts problem solving assessment (PSA) intention score for 3 standardized problems, measured 6 months and immediately after completing RUTF therapy, respectively. Plasma fatty acid content, anthropometry and eye tracking were secondary outcomes. Comparisons were made between the novel PUFA RUTFs and S-RUTF. Results: Among the 2565 SAM children enrolled, global MDAT z-score was -0.69 {+/-} 1.19 and -0.88 {+/-} 1.27 for children receiving DHA-HO-RUTF and S-RUTF, respectively (difference 0.19, 95% CI 0.01 to 0.38). The gross motor and social domains had higher z-scores among children receiving either DHA-HO-RUTF than S-RUTF. The PSA problem 3 scores did not differ by dietary group (Odds ratio 0.92, 95% CI 0.67 to 1.26 for DHA-HO-RUTF). After 4 weeks of treatment, plasma phospholipid EPA and a- linolenic acid were greater in children consuming DHA-HO-RUTF or HO-RUTF when compared to S-RUTF (for all 4 comparisons P values < 0.001), but only plasma DHA was greater in DHA-HO-RUTF than S-RUTF ( P <0.001). Conclusions: Treatment of uncomplicated SAM with DHA-HO-RUTF resulted in an improved MDAT score, conferring a cognitive benefit six months after completing diet therapy. This treatment should be explored in operational settings.

1 0 high oleic peanuts. Increases in omega-3 PUFA were achieved by the addition of perilla oil. formulation was demonstrated to be acceptable in a pilot trial (17) . MDAT has been validated within the study context (28). We performed MDAT 1 8 6 assessments between five and seven months after a SAM outcome was reached to assess the are assessed by direct observation of the child (gross motor, fine motor, and language), while the global score and compared to normal population reference values to provide age-adjusted z- that domain and skipping to the subsequent domain (19) . Global z-score outliers were defined 1 9 4 after evaluation of the score distribution, and values < -5 were excluded. intervention RUTFs. The assessment was modified to accommodate testing a wider age range 1 9 8 than those used in prior studies and so contained three problems of increasing complexity (25) . The goal of the assessment was to judge intentionality in means-end task completion. All Each problem contained a barrier between the child and the goal, a toy of her or his 2 0 2 choosing. In problem 1, the toy was placed on a cloth and distanced from the child such that the 2 0 3 All rights reserved. No reuse allowed without permission. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this this version posted September 12, 2021. 1 1 cloth had to be pulled by the child to reach the toy. In problem 2, the toy was covered with an 2 0 4 opaque cloth such that the child had to remove the cover to obtain the toy. Problem 3 was an 2 0 5 amalgam of problems 1 and 2, wherein the toy was placed on a cloth and under a cover, such that 2 0 6 the child had to pull the apparatus toward them and uncover the toy to obtain it. Videos were coded using Behavioral Observation Research Interactive Software (29). Coders were trained to achieve >.90 agreement by kappa statistic for intention scores on a random sample of 20 test videos before coding for the trial. High reliability was maintained 2 1 0 throughout the coding period by double-coding >50% of videos followed by a joint resolution of behaviors on each subgoal. Problems 1 and 2 were scored 0-4, while problem 3 had two subgoals 2 1 5 (moving cloth and removing the cover) and thus was scored 0-8. "No score" was a possible outcome for all problems, and reasons for this result were recorded. Two eye-tracking tests were performed as secondary cognitive outcomes: a visual paired 2 1 8 comparison task, and the infant orienting with attention task. Identical procedures to those 2 1 9 previously described were used (20) . The visual paired task consisted of 4 trials of African faces. The infant orienting task measured saccadic reaction time in a standardized manner. The testing 2 2 1 sequence for paired comparison task and orienting attention task was randomized. All neurocognitive assessors underwent extensive training and were required to pass periodic evaluations before administering the tests and throughout the trial. Please see the study 2 2 4 protocol for further details. All rights reserved. No reuse allowed without permission. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this this version posted September 12, 2021. ; https://doi.org/10.1101/2021.09.07.21263231 doi: medRxiv preprint 1 2 Plasma sampling and analysis 2 2 7 The purpose of blood sampling was to characterize the concentrations of fatty acids in the study 2 2 8 groups, thus only a subset of children were sampled. Blood collection and analysis followed best 2 2 9 practices (30). After receiving 4 weeks of RUTF, blood was collected by venipuncture and Institute in Austin, TX. Plasma phospholipids were chosen as the preferred lipid pool because 2 3 3 they respond in a matter of weeks to changes in the diet compared to months for red blood cells, 2 3 4 but are stable to fatty acids from recent meals which are largely triacylglycerols. Red blood cells 2 3 5 PUFA also degrade at -20°C by peroxidation catalyzed by iron release from ruptured cells (30). For analyses, plasma phospholipids were isolated using an automated three-phase liquid- liquid extraction method (31). Samples were then transmethylated to generate fatty acid methyl Statistical analysis 2 4 6 Analysis were performed by a blinded investigator using a modified intention to treat 2 4 7 methodology wherein children who were discovered not to meet enrollment criteria were 2 4 8 excluded from analysis. Baseline characteristics were summarized as means ± SD, medians RUTF on MDAT global z-score and PSA intentions scores, respectively, is shown in Six key plasma PUFA concentrations were selected to describe the biochemical changes amounts of plasma ALA, as well as more EPA, while DHA was increased only in the DHA-HO- All rights reserved. No reuse allowed without permission. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this this version posted September 12, 2021. ; https://doi.org/10.1101/2021.09.07.21263231 doi: medRxiv preprint 1 6 DHA-HO-RUTF and 75% for S-RUTF. No adverse effects or negative preferences against 3 1 7 DHA-HO-RUTF were observed in this large trial. 3 1 8 This trial demonstrated a clear discrepancy between anthropometric recovery and 3 1 9 cognitive recovery in SAM. Children fed S-RUTF show inferior cognitive performance when suggests that a frameshift in thinking and approach may be warranted with respect to SAM treatment and recovery. The focus on cognitive recovery must rise to equal importance as 3 2 4 anthropometric recovery to enable the child to thrive. The distribution of MDAT scores in that improvements in treatment may avert lifelong disabilities. The increase seen in MDAT by 3 2 7 the inclusion of less LA and more DHA in RUTF is encouraging and actionable. Differences SAM insult. Our results report a functional benefit that extends the importance of our previous The MDAT was developed in a rural Malawian setting, is fully validated, reliable, and predictive of later intellectual performance. This is why MDAT was chosen was as a primary 3 3 4 outcome, and why Malawi was chosen as the study location. MDAT has revealed cognitive 3 3 5 deficits in acutely malnourished populations in Malawi and Burkina Faso. In a previous study, 3 3 6 MDAT domain scores were reported in 150 Malawian SAM children upon discharge from the hospital (21). All 4 domain scores were remarkably lower among hospitalized SAM children 3 3 8 than in our population; differences of -1.7 z-scores seen in gross motor, -1.0 z-scores in fine 3 3 9 All rights reserved. No reuse allowed without permission. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this this version posted September 12, 2021. ; https://doi.org/10.1101/2021.09.07.21263231 doi: medRxiv preprint 1 7 motor, -0.7 in language, and -3.0 z-scores seen in social. The MUAC of these hospitalized 3 4 0 children was 11.5 cm, similar to that in our population, although the clinical status due to 3 4 1 infectious complications was likely worse. These findings lead us to speculate that significant 3 4 2 improvements in MDAT occur after discharge, especially in the gross motor and social domains, 3 4 3 and that maximizing this "cognitive recovery" should be a primary goal in malnutrition programs 3 4 4 The PSA scores did not differ between food groups. PSA problem 3 has been used to 3 4 5 demonstrate cognitive differences among healthy infants aged 9-12 mo receiving DHA enriched 3 4 6 infant formulas (25) . In this study, half of the healthy infants achieved a perfect score on the PSA on all 3 attempts. This is in contrast to our results, wherein only 15% of children achieved a 3 4 8 perfect score on any attempt. This suggests the presence of a substantial cognitive deficit at the 3 4 9 time of anthropometric recovery. Our PSA was conducted using this very same problem and 3 5 0 protocol, and with a sample size increased three-fold, as previously reported, so it is unlikely the explore its timing relative to anthropometry and the potential utility of repeated testing. 3 5 4 Two previous, smaller studies in SAM children were done using blood measures of fatty 3 5 5 acid content (18, 35) . Our findings are consonant with these; increases in DHA were only seen 3 5 6 when fish oil was added to the diet, and reduction in dietary LA resulted in greater EPA content. The food formulations of HO-RUTF and DHA-HO-RUTF were achieved by Project Peanut Butter by ingredient changes, without alterations in the mixing or packaging processes. HO peanuts can be purchased in the major peanut production markets worldwide. DHA is now 3 6 0 encapsulated and available from the same ingredient producers as the micronutrient premix. to limiting oxidative degradation. Non-genetically modified, high oleic vegetable oils are 3 6 3 available at a cost of about 10% more than traditional vegetable oils. They were developed for 3 6 4 purposes of increasing the shelf-life of the oil and have been shown to have a reduced risk of 3 6 5 heart disease in adults. 3 6 6 Our study has multiple limitations. The study population did not habitually consume fish, 3 6 7 and the positive effect of DHA-HO-RUTF in a fish consuming population might differ. However, most children recovering from SAM worldwide consume little else but RUTF, and introducing bias. Re-analysis of the sample enrolled outside of these stock-outs did not show We estimate that during treatment, SAM children were consuming about 240g RUTF/d. For the DHA-HO-RUTF group, DHA intake averaged 173 mg/d or about 0.24%w/w DHA. The Importantly, the effective intervention was to supply RUTF with DHA and with limited LA All rights reserved. No reuse allowed without permission. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this this version posted September 12, 2021. ; 1 9 DPA does not support neurocognitive function similarly to DHA, thus our results may point to This study is the first to provide direct evidence that reduction in LA and addition of All rights reserved. No reuse allowed without permission. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. All rights reserved. No reuse allowed without permission. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this this version posted September 12, 2021. ; https://doi.org/10.1101/2021.09.07.21263231 doi: medRxiv preprint All rights reserved. No reuse allowed without permission. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this this version posted September 12, 2021. ; https://doi.org/10.1101/2021.09.07.21263231 doi: medRxiv preprint preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. All rights reserved. No reuse allowed without permission. preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this this version posted September 12, 2021. ; https://doi.org/10.1101/2021.09.07.21263231 doi: medRxiv preprint Percentages of children in each intervention group with each score are shown. Overall, children in each study food had similar intention scores in all three problems; the differences between food groups were not significant. The lowest and highest scores possible in each problem were the most common results among children with scores. Children with "No Score" were unable to engage in the task. 

Acute malnutrition recovery energy requirements based on mid-upper arm circumference : secondary analysis of feeding program data from 5 countries, combined protocol for acute malnutrition study (ComPAS) stage 1

World Health Organization, the World Food Programme, the United Nations System Standing Committee on Nutrition and the United Nations Children's Fund . Community-based management of severe acute malnutrition

Tissue levels of polyunsaturated fatty acids during early human development

The relationship between age and the fatty acid composition of cerebral cortex and erythrocytes in human subjects

Brain docosahexaenoic acid uptake and metabolism

Genetic variants of the FADS1 FADS2 gene cluster as related to fatty acid metabolism

Desaturase and elongase-limiting endogenous long-chain polunsaturated fatty acid biosynthesis

Nutritional Neuroscience as Mainstream of Psychiatry: The Evidence-Based Treatment Guidelines for Using Omega-3 Fatty Acids as a New Treatment for Psychiatric Disorders in Children and Adolescents

Efficacy of omega-3 highly unsaturated fatty acids in the treatment of depression

High-Oleic Readyto-Use Therapeutic Food Maintains Docosahexaenoic Acid Status in Severe Malnutrition

Statistical power analyses using G*Power 3.1: tests for correlation and regression analyses

doi: den Heuvel M

Developmental and behavioural problems in children with severe acute malnutrition in Malawi: A cross-sectional study

A meta analysis of nutrition interventions on mental development of children under two in low and middle income countries

Early development in children with moderate acute malnutrition: A cross-sectional study in

Influence of long-chain polyunsaturated fatty acids on infant cognitive function

Three Randomized Controlled Trials of Early Long-Chain Polyunsaturated Fatty Acid Supplementation on Means-End Problem Solving in 9-Month-Olds

Antibiotics as Part of the Management of Severe Acute Malnutrition

All rights reserved. No reuse allowed without permission

preprint (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity

Ready-to-use therapeutic food with elevated n-3 polyunsaturated fatty acid content, with or without fish oil, to treat severe acute malnutrition: a randomized controlled trial

Docosahexaenoic and arachidonic acid concentrations in human breast milk worldwide

Polyunsaturated fatty acid biosynthesis pathway and genetics. implications for interindividual variability in prothrombotic, inflammatory conditions such as COVID-19(, bigstar, bigstar bigstar)

FADS1 and FADS2 Polymorphisms Modulate Fatty Acid Metabolism and Dietary Impact on Health

Tissue-specific impact of FADS cluster variants on FADS1 and FADS2 gene expression

Can prenatal N-3 fatty acid deficiency be completely reversed after birth? Effects on retinal and brain biochemistry and visual function in rhesus monkeys

Plus-minus values are means ± SD. ART, antiretroviral therapy

docosahexaenoic acid added to ready-to-use therapeutic food made with high-oleic acid peanuts; HAZ, height-for-age z-score

RUTF with high oleic acid peanuts; IQR

MUAC, mid-upper arm circumference

WHZ, weight-for-height z-score

Programmatic Outcomes Recovered