Nulls: May-June 2020

Vitamin D supplemention didn’t help high-risk infants’ allergy risk ^[1]

• What was studied? Infants at high genetic risk for allergies but who had adequate vitamin D levels were supplemented with 400 IU vitamin D or placebo daily for their first six months of life. This study followed up with those infants to explore how allergen sensitivity and other allergy-related outcomes were affected at up to 2.5 years of age.
• Why study it? While vitamin D levels in children have been associated with higher rates of allergies and vitamin D plays a major role in immune development, maternal supplementation during pregnancy hasn’t affected children’s allergy rates. However, supplementation just after birth hasn’t been well explored.
• What was(n’t) found? There was no evidence to suggest that vitamin D supplementation up to 6 months of age influenced any allergy-related outcome up to 2 years afterwards.
• How null was it? This was a follow-up to a study designed to test immunological outcomes, not allergy rates. Furthermore, there was a lot of uncertainty in the outcomes measured. Thus, there’s a lot more room for more evidence.

Sugar may not make you smarter^[2]

• What was studied? How two isoenergetic foods with different glycemic indices (GIs) impact cognition through a randomized crossover trial.
• Why study it? Glucose is the primary source of energy for the brain, so giving it more fuel may boost cognition in theory. However, previous studies exploring this issue have been inconsistent and sometimes flawed.
• What was(n’t) found? There was no clear effect of GI on a wide range of cognitive tests measuring recall and information processing up to 120 minutes after consuming either food.
• How null was it? Not very. The study controlled for blinding and confounding better than previous studies exploring this issue, but only used two foods that differed only in what sweetener was used (sucrose for the higher GI condition and isomaltulose for the lower one). However, the measured GI of both foods turned out to be relatively low and not that different from each other (44 versus 33). Future research using a wider range of GIs would be useful to verify this finding.

Probiotics don’t clearly help cognition in people with Alzheimer’s disease^[3]

• What was studied? Randomized controlled trials that explored how probiotic supplementation could affect cognition in people with Alzheimer’s disease.
• Why study it? Animal studies suggest that the gut-brain axis plays a role in preventing brain cells from damage and that the gut microbiome could influence brain health.
• What was(n’t) found? Three studies involving 161 people with Alzheimer’s disease supplementing with Lactobacillus and Bifidobacterium strains were meta-analyzed. No statistically significant effect was found on cognition.
• How null was it? Not very, since there were only a few studies, the overall sample size was small, and the authors rated the evidence as very low certainty.
• Anything else? There was some evidence that supplementation helped with blood lipids, insulin resistance, and oxidative stress.

Resveratrol had little effect on insulin sensitivity^[4]

• What was studied? The effect of 6 months of 150 mg resveratrol daily on insulin sensitivity in people with BMIs ranging from 27-35 and without diabetes.
• Why study it? Resveratrol activates sirtuin 1, which could help protect against the development of metabolic issues in theory. Also, animal studies have found a protective effect for insulin sensitivity, while human studies have been inconsistent.
• What was(n’t) found? Resveratrol had no statistically significant effect on the primary outcome of insulin sensitivity, nor on cholesterol levels.
• How null was it? The study had a large enough sample size to detect major changes in its primary outcome of insulin sensitivity. However, it measured this outcome using a repeated oral glucose tolerance test rather than the gold standard hyperinsulinemic euglycemic clamp.
• Anything else? The study found a statistically significant drop in HbA1c of 0.2% in the resveratrol group, but this was a secondary outcome amongst a host of other items measured, so could be due to chance. And even if it’s not, it’s a pretty small drop.

Walnuts don’t seem to affect blood pressure^[5]

• What was studied? The effect of walnut consumption on blood pressure through meta-analyzing randomized controlled trials
• Why study it? Walnuts are rich in compounds that could lower blood pressure in theory, but trials examining their blood pressure-lowering effects have been equivocal.
• What was(n’t) found? There was no statistically significant effect on either diastolic or systolic blood pressure.
• How null was it? Quite null. The data were at best consistent with clinically insignificant changes in blood pressure. Also, the total number of participants was quite large (around 1,800), there was no evidence of publication bias, and the studies were of moderate or good quality for the most part. However, the individual studies that went into the meta-analysis had pretty small sample sizes and didn’t always involve a hypertensive population, leaving open the possibility that longer trials with larger sample sizes of people with hypertension could find some effect.

HMB picks up an L for strength and body comp outcomes^[6]

• What was studied? Trials examining β-hydroxy-β-methylbutyrate (HMB’s) effect on strength and body composition in trained and untrained adults under 50.
• Why study it? While there’s reason to suspect HMB could boost muscle mass and strength, trials have had conflicting results, suggesting the need to pool them.
• What was(n’t) found? There was no clear evidence that HMB improved strength, fat-free mass, or fat mass.
• How null was it? While the meta-analysis does not support the use of HMB for body composition or strength gains in younger people, it involved studies with small sample sizes, many of which had an unclear risk of bias for blinding. This suggests higher quality, larger studies could change the conclusion. Also, while the effect on fat-free mass was statistically insignificant, the 95% confidence interval suggests that the data was mostly compatible with an improvement.

Carb mouth rinses didn’t help boxers’ reaction time^[7]

• What was studied? The effect of a 6% dextrose mouth rinse on a measure of male amateur boxer’s reaction time.
• Why study it? There’s some evidence to support using carbohydrate mouth rinses in sports, but it’s utility for boxers’ reaction times hasn’t been well studied.
• What was(n’t) found? There was no clear effect on reaction times.
• How null was it? Not very. The study only involved eight boxers, measured a lot of outcomes, and used a questionable statistical method for analyzing the data.

Beetroot didn’t boost repeated sprint performance^[8]

• What was studied? The effects of nitrate-rich beetroot juice (containing 0.40-0.45 grams of nitrate) on repeated sprinting in male team sports athletes who compete at college level.
• Why study it? The research to date on nitrate’s effects on repeated sprint performance have mostly focused on people who participate in solo athletics (e.g., cycling) rather than team athletes, and the results have been inconclusive.
• What was(n’t) found? There was no difference between sprint times when the athletes consumed either nitrate-rich or nitrate-depleted beetroot juice around 3 hours before performing repeated bouts of sprinting.
• How null was it? It’s unclear, since no power analysis was done, making it harder to tell if this crossover trial was more likely a false or true negative.

The sun may be setting on vitamin D supporting bone health in the general population^[9]

• What was studied? Whether vitamin D levels have a causal role in fracture risk through Mendelian randomization.
• Why study it? There’s a lot of observational evidence suggesting that low vitamin D levels correlate with fracture risk, but randomized controlled trials examining vitamin D supplementation haven’t found a clear benefit.
• What was(n’t) found? While vitamin D levels were found to be correlated with fracture risk, the Medelian analysis did not support a causal relationship between vitamin D levels and fracture risk.
• How null was it? Somewhat. On the one hand, this was a large study and some care was taken to choose genes that would only affect vitamin D levels. On the other hand, the chosen genes could only explain around 2% of the variation in vitamin D levels, which is relatively low compared to other Mendelian randomization studies. This leaves open the possibility that the genes that were chosen for this analysis may not have been sufficient.
• Anything else? The study was done in a white, Norwegian population; the results may not carry over to other populations. The results may also not apply to people with osteoporosis or severe vitamin D deficiency.