Randomized Controlled Clinical Trial of Pediatric Pneumococcus and Hepatitis A Vaccinations With or Without a High-Dose Oral Vitamin A Supplement

Abstract

Previous studies have shown that high-dose vitamin supplements can improve vaccine-induced immune responses and pathogen protection in the context of vitamin deficiencies. To further elucidate the influence of vitamin supplements on immune responses toward pediatric vaccines, we performed a randomized controlled clinical trial (PCVIT) of 20 healthy children 1-4 years of age in Memphis, Tennessee. Study participants received a booster vaccine for pneumococcus and a primary vaccine for hepatitis A virus with or without a high-dose, oral, liquid supplement of 10,000 IU retinyl palmitate. We found that the children enrolled in PCVIT had higher baseline vitamin levels than previously described older children and adults living in Memphis. Only one child in PCVIT had a serum retinol level of less than 0.3 µg/mL. The children frequently consumed milk and baby foods that were likely vitamin-fortified, providing an explanation for the relatively high vitamin levels. Most children in PCVIT responded well to pneumococcus and hepatitis A vaccines by pathogen-specific antibody upregulation. The one child with a serum retinol level below 0.3 µg/mL did not receive a vitamin supplement and exhibited the lowest fold-change in antibody responses toward pneumococcal serotypes. A correlation matrix encompassing demographics, vitamin levels, vaccine-induced immune responses, C-reactive protein, and total serum immunoglobulin isotypes, including IgG2 and IgA, identified variables associated with vaccination outcomes. Perhaps because children were predominantly retinol-sufficient at baseline, the high-dose vitamin A supplement exhibited no benefit to vaccine-induced immune responses. In fact, when vitamin supplemented and vitamin unsupplemented groups were compared among participants with the highest baseline retinol levels, there was a trend toward weaker vaccine-induced immune responses in the vitamin supplemented group. Results encourage the performance of larger clinical studies before high-dose vitamin supplements are recommended for populations that are otherwise vitamin-replete.

Keywords: hepatitis A vaccine; pneumococcus vaccine; retinol; retinol binding protein.

Figure 1

A comparison of baseline vitamin levels between participants in a previously described FLUVIT study and participants in PCVIT. The previously described FLUVIT study evaluated 44 children 2–8 years of age, enrolled during the years 2016–2017 [28]. Baseline vitamin levels are shown for comparisons between the FLUVIT and PCVIT study data. Each symbol represents a different study participant, with medians shown. There was a measurement for all 20 PCVIT study participants for RBP and vitamin D, and 19/20 participants for retinol. Two-toned symbols represent the three children 2–4 years of age in the PCVIT study; these values always fell at or below the group median. Mann–Whitney tests were performed to compare FLUVIT with PCVIT groups. The one significant p value is shown.

Figure 2

Antibody responses toward pneumococcal serotypes represented in the vaccine following Prevnar vaccinations with or without high-dose vitamin A supplements in a predominantly retinol-sufficient pediatric population. Graphs are labeled by the particular pneumococcal serotype (termed ‘serotype’) toward which antibodies responded. Each symbol represents a different participant. The x-axis identifies test (vitamin supplemented, +) and control (unsupplemented, −) participant groups. The y-axis identifies differences (fold-change, Δ) in specific antibody levels toward a given pneumococcal serotype represented in the vaccine, before and after vaccination. Participants were further grouped according to baseline vitamin levels (cut-off = 0.4 µg/mL retinol). Medians are shown for Δ antibody levels in each group. Only one participant (open diamond) had a baseline retinol level below 0.3 µg/mL. Among children with the higher retinol levels at baseline, median pneumococcus-specific antibody fold-change (Δ) levels trended lower in the supplemented group. In this figure, 19A was the only pneumococcal serotype represented in the vaccine for which all specific antibody values fell within assay range. The cut-off value for the limit of detection was used when pneumococcus-specific antibody values were out of range.

Figure 3

Correlation matrix analyses. In this correlation matrix, circle sizes represent the absolute values of the corresponding correlation coefficients, and circle colors indicate the +/− values of the correlation coefficients. The arrangement of parameters was determined using hierarchical clustering (‘hclust’). Two vaccine-induced antibody responses were included. These were specific for pneumococcal serotype 19A, for which no values were above or below the assay’s limit of detection, and specific for pneumococcal serotype 19F. Non-significant (p > 0.05) coefficients were not displayed.