Nulls: January-February 2021

Fish oil found to do little to prevent further cardiovascular problems after a heart attack^[1]

• What was studied? Whether adding fish oil (930 mg of EPA and 660 mg of DHA daily over two years) to standard care two to eight weeks after a heart attack could prevent further negative cardiovascular outcomes in people over 70 years old.
• Why study it? There’s reasons to suspect omega-3 supplementation could prevent cardiovascular disease, but its follow-up use after a heart attack hasn’t been well studied.
• What was(n’t) found? Supplementation did not significantly affect the composite primary outcome of death, another heart attack, stroke, unscheduled revascularization, or hospitalization due to heart failure. None of these individual outcomes look to have been affected, either.
• How null was it? Somewhat. On the one hand, this was a well-designed, long term trial with over a thousand participants. On the other hand, the trial wound up being underpowered due to negative outcomes being less frequent than expected. Thus, this trial is evidence against huge effects, but doesn’t necessarily rule out more moderately-sized benefits.
• Anything else? This study adds to other recent evidence that fish oil may increase the risk of new-onset atrial fibrillation, in that people in the fish oil group tended to have higher rates of atrial fibrillation compared to the control group. On the plus side, this study didn’t find any evidence of increased bleeding risk in the fish oil group.

Chlorella had little effect on metabolic measures in people with type 2 diabetes^[2]

• What was studied? Researchers investigated how 1,500 mg of daily chlorella for eight weeks affected glycemic control measures, blood pressure, food intake, and more in people with type 2 diabetes.
• Why study it? Chlorella is a single-celled green algae. It’s packed full of potentially beneficial micronutrients, and evidence in mice and human studies in people with nonalcoholic fatty liver disease suggests that it can help with glycemic control. However, it hasn’t been tested in people with type 2 diabetes before.
• What was(n’t) found? No effects were found for the primary outcome, HbA1c, or any other metric, including blood lipids, body weight, blood pressure, and more.
• How null was it? The study was designed to detect a 0.5% change in HbA1c, so the evidence is against there being clinically relevant effects for this metric. However, the population had relatively well-controlled diabetes and the study length was short. This leaves open the possibility that effects could be seen in longer trials with people with less well-controlled diabetes, perhaps at higher doses.

VITAL trial offshoot finds little evidence that vitamin D supplementation influences body composition^[3]

• What was studied? Researchers evaluated how 2,000 IU of daily vitamin D affected body composition over the course of two years.
• Why study it? Observational studies suggest that vitamin D is associated with body composition, but whether this link is causal or not is open to question.
• What was(n’t) found? There was no clear effect of vitamin D supplementation on any metric of body composition measured, including weight, waist circumference, and fat mass.
• How null was it? This was a secondary analysis of a smaller sub-study of the main VITAL trial, so these results shouldn’t be taken as definitive. Also, this substudy had an average participant age over 60, so the results may not extend to younger populations.
• Anything else? Subgroup analyses hinted that vitamin D supplementation could improve body fat percentage in people with a BMI of less than 25. However, the signal for this effect wasn’t very strong, and requires further follow-up.

Vitamin E supplementation does not seem to affect weight of waist circumference^[4]

• What was studied? Researchers meta-analyzed randomized controlled trials using vitamin E supplementation that measured body weight, BMI, or waist circumference.
• Why study it? There are a few mechanistic reasons to suspect vitamin E could have some effect on weight. For instance, one form of vitamin E, alpha-tocopherol, may boost adiponectin expression, which could ultimately lead to more fat being burned. More generally, vitamin E could reduce certain inflammatory markers linked to obesiy like IL-6 and TNF-alpha. However, evidence from clinical trials is mixed, suggesting that a meta-analysis could help clarify the situation.
• What was(n’t) found? No statistically significant effect on weight, BMI, or waist circumference was found.
• How null was it? Pretty null. The confidence intervals suggest that the evidence isn’t compatible with clinically significant changes in any of the parameters. Also, the heterogeneity was very low, although there was some risk of bias detected in key areas in some of the included studies.

A specific Ayurvedic herbal combination had no clear effect on metabolic markers in people with impaired glycemic control^[5]

• What was studied? Researchers evaluated whether Mohana Choorna, a 20-herb combination used in Ayurvedic medicine for type 2 diabetes, affected metabolic markers in people with impaired glucose tolerance.
• Why study it? This herbal combination is used to treat or prevent type 2 diabetes in Ayurvedic medicine, and there’s some animal and test tube evidence suggesting that certain herbs in the combination could exert the desired effect. However, there is little clinical evidence supporting this mixture.
• What was(n’t) found? Researchers concluded that 500 mg of Mohana Choorna taken three times per day for four weeks had no statistically significant effect on any metric of glycemic control, blood lipids, or blood pressure.
• How null was it? A power calculation wasn’t performed, making it difficult to interpret the null results found here. However, the study did use a crossover design involving 22 participants, which is a decent sample size.
• Anything else? There was some evidence suggesting that Mohana Choorna may raise fasting insulin and inflammatory gene expression in fat tissue, raising possible concerns about the safety of the herbal mixture. However, this data was only suggestive, and requires further follow-up.

Maternal iodine supplementation during pregnancy may not affect child’s growth or neurodevelopment^[6]

• What was studied? This meta-analysis was designed to evaluate whether clinical trials involving women supplementing iodine during pregnancy found any effect on the growth or neurodevelopment of their children.
• Why study it? Iodine is necessary for proper growth and brain development, and maternal supplementation of iodine increases during pregnancy. This has led the American Academy of Pediatrics and the Endocrine Society to recommend that women who are or plan to be pregnant supplement with 150 micrograms of iodine daily. However, randomized trial results exploring iodine’s efficacy have been mixed.
• What was(n’t) found? Meta-analyzing five trials found that maternal iodine supplementation had no clear effect on childrens’ growth or neurodevelopment up to age 2.
• How null was it? There’s a lot of room for more data for three reasons. First, the mothers in the studies tended to be mildly iodine deficient at best, suggesting that benefit could be seen in more iodine-deficient populations. Second, the iodine supplementation tended to start rather late in pregnancy, leaving room open for the possibility of early supplementation having a stronger effect. Finally, there weren’t very many studies exploring the issue, leaving a lot of room for uncertainty.

Creatine may not help much with short-term recovery from exercised-induced muscle damage^[7]

• What was studied? Researchers meta-analyzed randomized controlled trials to determine whether creatine supplementation can help muscles recover from exercise-induced muscle damage.
• Why study it? There are theoretical reasons for thinking that creatine may help with muscle recovery in the short term. The clinical trial results, however, have been mixed, and this evidence has not been meta-analyzed previously.
• What was(n’t) found? Creatine supplementation had no clear effect on muscle strength, soreness, or muscle damage markers up to 96 hours post-exercise. There was one exception: creatine kinase was lower 48 hours post-exercise in people taking creatine, but this could have just been a statistical blip.
• How null was it? While the quality of the studies covered tended to be good, there weren’t very many studies and they were all small: only 13 studies were found, including a total of just 278 participants. Also, these studies had different dosing regimens, different populations ranging from untrained to athlete, and different methods for inducing muscle damage. Thus, averaging all these studies together may not mean much. Larger studies looking at specific populations would be useful before ruling out creatine’s short-term effects on exercise-induced muscle damage entirely.

Grape powder failed to blunt after-meal spikes in blood glucose and lipids in people with obesity^[8]

• What was studied? Researchers evaluated whether 46 grams of grape powder (equivalent to 252 grams of fresh grapes) served at breakfast affected post-meal blood glucose and triglyceride levels after high-fat, high-sugar breakfasts and lunch in people with obesity (BMI between 30 and 40).
• Why study it? Big spikes in blood sugar and triglycerides after a meal both predict cardiovascular disease risk. Polyphenols found in grapes have been found to mitigate these spikes. However, most of the research has focused on single-meal responses after breakfast. The authors aimed to extend this timeframe by monitoring blood triglycerides and glucose after both breakfast and lunch.
• What was(n’t) found? There were no clear differences between the placebo and grape powder groups in glucose, nor in any other parameter measured.
• How null was it? Pretty null, for this dose. The study was powered to detect a pretty modest difference of about 3.2 mg/dL in blood glucose between the two groups after lunch, so this is decent evidence that there may not be much of an effect. However, there was also no difference in plasma antioxidant between the groups. This is surprising, since the grape polyphenols should have boosted this parameter. This finding strongly suggests that the dose used in this study was too low, and underscores the need for further studies that use higher doses of polyphenols.

A combination of four promising supplements had little effect on hand osteoarthritis pain^[9]

• What was studied? This study was designed to determine whether a combination of 250 mg of Boswellia serrata extract, 100 mg of pine bark extract, 1,500 mg of methylsulfonylmethane (MSM), and 168 mg of curcumin daily for 12 weeks could improve hand pain in people with hand osteoarthritis.
• Why study it? These four supplements were found to be promising from a systematic review conducted by the authors of this trial. However, their systematic review also raised some concerns about past study quality and most of the evidence focused on knee osteoarthritis, so they decided to run their own trial to determine if a combination of these promising supplements would work for hand arthritis.
• What was(n’t) found? No clear reduction of hand pain (the primary outcome) in the treatment group compared to placebo was found.
• How null was it? This study was done completely over the internet, so important metrics like compliance were self-reported, which could have skewed the results. Also, the lowest doses with some evidence of efficacy were chosen for some of the supplements to reduce pill burden and complexity, raising the possibility of higher doses yielding better results.

Mendelian randomization study finds that vitamin D doesn’t play much of a role in a host of diseases^[10]

• What was studied? The authors intended to determine the effects vitamin D levels have on a large range of diseases and biomarkers through a Mendelian randomization study.
• Why study it? Vitamin D receptors are spread throughout most tissues in the body, suggesting that vitamin D could affect a wide range of disease states. Furthermore, observational studies have found strong associations between vitamin D, disease risk, and a variety of biomarkers. However, whether these associations are causal is still unclear.
• What was(n’t) found? There was no evidence that vitamin had a causal effect on most biomarkers and disease states studied, ranging from obesity to cognitive function, cardiovascular disease, and cancer.
• How null was it? Wide confidence intervals for some metrics leave room for more evidence, and it’s possible that the effects of interventional trials could differ from those seen in Mendelian randomization studies.
• Anything else? One notable exception to the lack of effect was evidence suggesting a causal role of vitamin D in multiple sclerosis.

Vitamin C and zinc failed to clearly speed up recovery for people with COVID-19^[11]

• What was studied? Researchers evaluated whether 8 grams of vitamin C, 50 mg of zinc gluconate, or both, helped reduce the number of days it took for COVID-19 symptoms to halve in outpatients.
• Why study it? Both zinc and vitamin C may have some utility in mitigating respiratory tract infection symptoms and improving recovery. However, their efficacy in the context of COVID-19 specifically is unclear.
• What was(n’t) found? The trial was stopped early due to a low probability of finding an effect.
• How null was it? The trial participants knew what they were receiving and there was no placebo control. This, combined with the fact that symptoms were self-reported, could have influenced outcomes. Also, the dosing of zinc may have been suboptimal.