The electronic database search yielded 1611 results, of which 534 references were reviewed. A total of 12 articles were included for full text review as the remaining 522 looked at different aspects of DS. Eleven studies fulfilled the eligibility criteria (20-30) and one study was excluded as it demonstrated the difference in the biochemical markers rather than reporting the developmental outcome. (19) One study had mixed population with five children with DS (20). Study characteristics of the included studies are presented in Tables 1 and 2.
Study characteristics: Six studies were carried out in North America (20-25,30) while two each in France (21-22) , and the UK (23,24). Majority of the studies (8/11) were conducted in the 1980s (20,24-30), and only three trials were conducted after the year 2000 (21-23). Most of the studies from the 1980s were q-RCTs (7/11), (19-24,28,29) while remaining were RCTs (21-23). A total of 683 participants with a male preponderance (male to female ratio of 1.2:1) were present. The age range of included participants was between newborns to 17 years.
Intervention: Intervention type, dosage, and mode of administration is outlined in Table 1. The most common intervention used was Harrel’s formula (A compound comprising of Vitamin A palmitate 15,000 IU (reduced to 5000 U for months 5 through 8), Vitamin D (cholecalciferol) 300 IU, Thiamine mononitrate 300 mg, Riboflavin 200 mg, Niacinamide 750 mg, Calcium pantothenate 490 mg, Pyridoxine hydrochloride 350 mg, Cobalamin 1,000 µg, Folic acid 400 µg, Vitamin C (ascorbic acid) 1,500 mg, Vitamin E 600 IU, Magnesium (oxide) 300 mg, Calcium (carbonate) 400 mg, Zinc (oxide) 30 mg, Manganese (gluconate) 3 mg, Copper (gluconate) 1.75 mg, Iron (ferrous fumarate) 7.5 mg, Calcium phosphate 37.5 mg, Potassium Iodide 0.15 mg) which was used in three studies (20,25,26), while three studies used Folinic acid (Calcium leucovorin) (21-23). The remaining studies employed pyridoxine (27), other compounding of antioxidants (24,28), and other multivitamins (29-30). All studies compared the intervention with the placebo. (20-30)
Quality assessment of included studies: The overall quality of included studies and the ROB was variable with qualitative deficits (Figure 2 and 3). The reviewers observed that the three studies with low risk of bias (21,22,23) were carried out in 2008 and 2010s, with up to date methods, with the exception of the study by Lonsdale et al. showing low risk of bias and carried out in 1986(29). However, the majority of the studies in the review demonstrate high risk of bias. (19) Key concerns for ROB included potential selection bias in the form of lack of random sequence generation (n=5) (19,24-27), allocation concealment (n=5) (19,23,24,28,29), blinding of outcome assessors (n=5) (20,22,24,25,26), and selective reporting (n=2) (19-20).
Effect of intervention: In terms of outcomes reported and their measurement scales, study design, source population, participant’s age, and anti-oxidants used, quantitative analyses could not be performed as the trials were heterogeneous. Twelve different domains have been assessed and fourteen outcome measures were used (Table 2 matrix) with the outcomes summarized in the Table 3. Global Developmental Quotient (GDQ) using Griffith’s Mental Development Scale was the most common outcome measure employed in three studies (21-23). Additionally, a high proportion of studies were assessed to have a high risk of bias (7/11), thus the meta-analysis was carried out using these three studies.
A meta-analysis was applied to compare changes in the effect sizes (post and pre-intervention) between the intervention and control groups. The forest plot including the 3 studies (Figure 4) does not demonstrate a statistically significant improvement. Of the 11 included studies, eight studies did not show any statistically significant clinical benefit of using antioxidants (22-25,27-29). Harrel et al. showed improvement in 3/5 children with DS, however, this study had a mixed population of children with intellectual impairment (19). The second study by Blehaut et al. in the intention to treat analysis did not show any increased benefit of using leucovorin versus placebo, however, on the as per protocol analysis, there was demonstration of improvement in the developmental age (leucovorin=53.1% and placebo=44.4%, p value= 0.031) (20). The third study was by Lonsdale et al. where 5/22 showed increased IQ with the use of Thiamine Tetra hydrofurfuryl Disulfide (28).
Intelligence quotient (IQ) was the outcome used in three studies employing Stanford Binet (20,25-26), while two studies used Wechsler Intelligence Scale for Children (WISC) (28,29). Other outcomes include Language skills (23,26,28,30) Visuomotor integration (25,28) Social Maturity (27,30) articulation proficiency (26), academic achievement (26), learning aptitude (26), behavior rating (26), psycholinguistic skills (26). To report language skills, three different tools were used, such that MacArthur communicative development inventory (M-CDI) has been used by Ellis et al. (23), Peabody Picture Vocabulary Test (PPVT) by Smith and Bennet et al. (26,28), Receptive Expressive Emergent Language Scale (REELS) by Pueschel et al (30). The difference between the intervention and control group was not statistically significant for any of these developmental outcomes. (23-28,30)
Adverse events were observed in nine patients which include nocturnal enuresis (25), gastrointestinal symptoms (20,22-24), skin manifestations (23-24), infantile spasms (21,26), and sudden death (21).