Nulls: July-August 2020

L-arginine gets an “L” for severe asthma^[1]...

• What was studied? How L-arginine, dosed at 0.05 grams per kilogram ideal bodyweight, for 12 weeks affected “moderate” asthma exacerbations (defined as a composite of several measures) in people with severe asthma.
• Why study it? Several lines of evidence suggest some patients with severe asthma may have problems with L-arginine metabolism, leading to problems generating nitric oxide and worse asthma. Based on this evidence, the authors hypothesized that supplemental L-arginine would be especially effective in people with severe asthma and low or moderate amounts of exhaled nitric oxide.
• What was(n’t) found? There was no significant effect of L-arginine supplementation in the overall cohort on asthma exacerbation.
• How null was it? While the study was well powered, there’s plenty of room for more evidence for a couple of reasons. One is that, while well powered, the study was still relatively small and done out of a single center, leaving room for replication. Also, the authors discovered some biomarkers during the course of the study that may help identify people who would benefit from L-arginine supplementation. So, while this study didn’t find an effect, it provided researchers with a lot of information that could help identify people whose severe asthma may be helped by supplementation in the future.

...and so does vitamin D supplementation for severe asthma exacerbations in children^[2]

• What was studied? Children aged 6-16 with baseline vitamin D levels below 30 mg/dL with asthma who had at least one severe asthma exacerbation episode in the past year were randomized to receive 4,000 IU of vitamin D₃ or placebo daily for 48 weeks.
• Why study it? There are a lot of mechanisms that suggest that vitamin D supplementation may help with asthma. Furthermore, a previous meta-analysis using individual participant data found that supplementation could help lower asthma exacerbation risk. However, children with asthma didn’t benefit from supplementation in that meta-analysis, which encouraged the researchers to look more closely at whether children could benefit.
• What was(n’t) found? Vitamin D supplementation had no clear effect on either the primary outcome (time to exacerbation) or main secondary outcome (viral-induced severe asthma exacerbation) was found.
• How null was it? The rate of asthma exacerbation in both groups was lower than expected, and the study wasn’t powered to detect such a small difference. However, the trial was well-designed and was stopped early due to a prespecified futility definition. Overall, it seems that vitamin D supplementation had no effect on asthma exacerbations in at-risk children with low vitamin D levels.

Vitamin D intake didn’t have a clear impact on depression risk and symptoms^[3]

• What was studied? The impact of 2,000 IU daily vitamin D₃ supplementation on the risk of depression and depressive symptoms over the course of roughly 5 years.
• Why study it? Almost all the studies looking at vitamin D’s impact on preventing depression have been null. However, they’ve been small and short term. This is the largest, longest study looking at vitamin D supplementation’s ability to prevent depression.
• What was(n’t) found? No statistically significant impact on the risk of depression or depressive symptoms was found.
• How null was it? This study was branched off from the famous VITAL study, whose aim was to take a look at omega-3 fatty acid’s and vitamin D’s effects on cardiovascular and cancer outcomes. Thus, while the study was large and long, the study’s primary purpose wasn’t to test depression prevention. Also, given that baseline vitamin D levels for the population were adequate (~30 ng/mL), these results may not extend to people with vitamin D deficiency. Finally, the primary way depression was evaluated was through self-reporting, which may have introduced inaccuracies.

Vitamin D supplementation does not have a large effect on breast cancer risk^[4]

• What was studied? How vitamin D with or without calcium affects breast cancer risk, studied through a trial-sequential analysis of randomized controlled trials.
• Why study it? Vitamin D levels have been found to be correlated with cancer risk for many types of cancers and populations. The specific link between breast cancer risk and vitamin D has been bolstered by observations that risk goes down as women spend more time in the sun. However, randomized controlled trials involving vitamin D supplementation have yielded inconclusive results concerning breast cancer risk, raising the question of whether the breast cancer-vitamin D link is causal.
• What was(n’t) found? Vitamin D had no clear influence on breast cancer risk in trials ranging from 1-12 years in length using doses close to 4,000 IU daily (average daily dose across trials: ~1,000 IU). Adding calcium to vitamin D supplementation also had no clear effect one way or the other.
• How null was it? The authors’ trial sequential analysis concluded that it’s highly unlikely that more evidence will show that vitamin D reduces breast cancer risk by at least 30%. However, the authors point out that the included studies may not have lasted long enough to see an effect. Also, they cite evidence suggesting that benefit could be seen at higher doses (at least 2,000 IU daily) in people who are deficient in vitamin D (less than 30 nmol/L or 12 ng/mL). Taken together, future studies with larger doses in vitamin D-deficient populations over long periods of time could still see an effect. However, it’s unlikely that lower doses of vitamin D reduce breast cancer risk much in women whose baseline vitamin D levels are close to normal.

Lycopene doesn’t seem to have an impact on cardiovascular risk markers^[5]

• What was studied? The effect of lycopene intake on blood pressure and blood lipids was explored by meta-analyzing randomized controlled trials.
• Why study it? Lycopene is an antioxidant that could reduce cardiovascular disease risk through several possible mechanisms. Also, tomatoes are rich in lycopene and tomato intake has been correlated with better cardiovascular risk marker outcomes.
• What was(n’t) found? Lycopene supplementation had no effect on any blood pressure or blood lipid markers.
• How null was it? The included studies used a lot of methods from supplementing lycopene, from pills to paste to diet, making it hard to draw conclusions about lycopene in general. More focused research is needed to confirm these null results.

Omega-3s don’t help with overall cognition in older people without dementia (but may help memory)^[6]

• What was studied? The effect of omega-3 polyunsaturated fatty acid supplementation on cognition in older adults without dementia (but possibly with mild cognitive impairment) was studied through a meta-analysis or randomized controlled trials.
• Why study it? Animal and observational studies have found evidence of an association between omega-3 intake and cognition, but clinical trials have had mixed results.
• What was(n’t) found? No effect on global cognition scores was found.
• How null was it? So-so. On the one hand, the 95% confidence interval suggests that the data are mostly compatible with a tiny effect at best. Also, the studies tended to be high quality. On the other hand, there weren’t many trials that went into the meta-analysis for global cognition, and the doses were relatively low (median EPA daily dose: 480 mg; median DHA daily dose: 800 mg). While these results aren’t promising, there’s more room for more trials in this population, especially using higher doses over longer time periods.
• Anything else? Analysis of specific cognitive domains found a small effect of omega-3 supplementation on memory.

High-flavanol chocolate had no effect on cognition over low-flavanol chocolate in healthy older adults^[7]

• What was studied? The impact of commercially available high-flavanol dark chocolate (50g daily containing 410 mg flavanols) on measures of cognition in cognitively healthy, educated people ages 65-75, compared to low-flavanol dark chocolate (50g daily, containing 86 mg flavanols) over the course of eight weeks.
• Why study it? Observational studies have found a correlation between dark chocolate intake and cognitive function in older adults. However, the randomized controlled trials examining the issue have been mixed and mostly used cocoa drinks or supplements. There isn’t much evidence on the effects of commercially available dark chocolate on cognition. The low-flavanol dark chocolate was prepared for the purpose of this study as a control.
• What was(n’t) found? No difference between groups was seen for measures of task-switching ability, verbal fluency, and visual attention.
• How null was it? Not very, for a few reasons. The first is that both groups showed mild, but statistically nonsignificant, improvements over eight weeks. Whether this was due to a learning effect, the efficacy of the low-dose of flavanols the control group received, or is a statistical artifact is unclear. In addition, the lack of difference between the groups could be due to a ceiling effect: the participants were cognitively healthy and highly educated, so they may not be representative of the general population. Finally, the study was powered to detect a 20% difference between groups, so even if these results are a true negative, the study wasn’t powered to detect smaller, but still noticeable, differences.

Both 4- and 6-hour feeding windows help lose weight and improve heart health, but one isn’t better than the other^[8]

• What was studied? People with obesity were randomized to a 4-hour (3pm-7pm) or 6-hour (1pm-7pm) eating window, or a control group, to see the effects of time-restricted feeding on weight loss and cardiovascular health markers for eight weeks.
• Why study it? Time-restricted feeding has been found to be effective for helping people lose weight and improve cardiovascular health markers. However, it’s unclear if shorter eating windows will provide a better benefit compared to larger windows.
• What was(n’t) found? Both time-restricted feeding groups lost similar amounts of weight, had a lower calorie intake, and had some improved glycemic control markers compared to the control group. However, there weren’t any major differences between the two groups.
• How null was it? A little. This was a relatively small study (58 total participants), which was only powered to detect weight change. This study provides preliminary evidence that 4- and 6-hour feeding windows don’t differ much, but should be followed up with a study with a larger sample size over longer time periods.
• Anything else? There was a mild benefit for lean mass in the 4-hour feeding group, although this could be a statistical blip, since the study wasn’t well-powered to detect secondary outcomes and multiple comparisons were made. It should also be emphasized that the null here is in the difference between the 4- and 6-hour eating windows. This study did find that time-restricted feeding worked compared to the usual-diet control group!

Reducing entrée portion size doesn’t affect dessert intake afterwards^[9]

• What was studied? How reducing a lunch entrée portion by up to 25% affects ad libitum (eating as much as you want) dessert intake.
• Why study it? The effects of reduced portion sizes on future ad libitum food intake isn’t clear, especially when it comes to food intake right after a meal. Since dessert often follows a meal, the authors aimed to explore how lunch size impacts dessert intake right after the meal.
• What was(n’t) found? Reducing lunch portions by up to 10-25% had no effect on how much dessert the participants consumed afterwards.
• How null was it? Not very. The results of this study probably aren’t easily generalizable, since the meal (macaroni and cheese) and dessert (ice cream) was fixed and also generally highly palatable. Also, the participants were told about the purpose of the study, which may have affected their behavior.

Magnesium supplementation doesn’t influence weight much^[10]

• What was studied? How magnesium supplementation affects weight, waist size, body fat percentage, and BMI was studied through a meta-analysis of randomized controlled trials.
• Why study it? Magnesium is involved in many energy-producing processes in the body, and observational studies have found a correlation between serum magnesium levels and obesity. However, clinical trials looking at magnesium supplementation’s effects on measures related to obesity haven’t been too promising.
• What was(n’t) found? There was no clear effect on any obesity-related measure.
• How null was it? The study quality of the examined was mostly low or middling, and many of the included studies primary purpose wasn’t to study obesity-related outcomes. More high-quality evidence focusing on magnesium supplementation’s effect on obesity-related metrics would be useful.
• Anything else? Subgroup analysis found that magnesium supplementation did decrease waist circumference only in people with obesity. However, this result should be taken with a grain of salt since lots of subgroup analyses were done, and other obesity-related metrics didn’t show a strong effect.

Vitamin D supplementation doesn’t affect obesity metrics in healthy populations^[11]

• What was studied? The effect of vitamin D supplementation on obesity-related measures in healthy people was studied through meta-analyzing randomized controlled trials.
• Why study it? Low vitamin D levels are correlated with obesity outcomes, but randomized controlled trials looking into whether supplementation could help haven’t been too encouraging. However, there haven’t been many international meta-analyses examining this topic.
• What was(n’t) found? Vitamin D supplementation had no statistically significant effect on BMI, waist circumference, or waist-to-hip ratio.
• How null was it? While the overall sample size for meta-analysis was large, the individual studies were on the small side, opening up the possibility that larger trials could provide more information. Also, subgroup analysis found positive results for studies done in Asian countries, in women, and in studies lasting longer than six months. So, while these results don’t provide evidence for much of a benefit overall, there’s a possibility that vitamin D supplementation could influence obesity-related metrics in certain populations. More research is needed to confirm this hypothesis, though.

Resveratrol seems to have little effect on non-alcoholic fatty liver disease^[12]

• What was studied? Randomized controlled trials examining resveratrol supplementation on non-alcoholic fatty liver disease were meta-analyzed.
• Why study it? Resveratrol has been shown to slow down fat synthesis in test tube and animal studies, suggesting it could help with fat accumulation in the liver. However, the evidence coming from controlled trials has been mixed.
• What was(n’t) found? Resveratrol supplementation ranging from 500 to 3000 mg daily had no clear effect on most outcomes related to non-fatty acid liver disease, including blood lipids, blood pressure, obesity-related parameters, and one liver enzyme blood level (AST). However, there was a borderline statistically significant effect on another liver enzyme (ALT).
• How null was it? While the studies were mostly high quality, there also weren’t many of them and they had small sample sizes and heterogenous inclusion criteria, so there’s room for more evidence.

Collagen peptide’s efficacy for self-reported knee pain gets kneecapped ^[13]

• What was studied? The effect of 10 grams of daily collagen peptide derived from cowhide for 12 weeks on knee pain in 200 physically active people over 50 years old who had self-reported knee pain, but no diagnosed joint conditions (e.g. rheumatoid or osteoarthritis), and weren’t on other joint health supplements.
• Why study it? Collagen peptide has mostly been studied in athletic people and people with diagnosed osteoarthritis. It hasn’t been as well studied for general self-reported knee pain.
• What was(n’t) found? Collagen peptide didn’t help with knee pain or with any joint biomarkers that were measured.
• How null was it? The study had a large enough dropout rate so that the sample size wasn’t as large as the researchers hoped it would be, making the study underpowered.
• Anything else? The population should be kept in mind when interpreting these findings. Participants were active people without diagnosed joint conditions who were preparing for a multi-day exercise event. This is probably not representative of the population who would likely take collagen peptide on their own.