Square Pegs & Round Holes: Complications Of Interpreting Secondary Analyses of Nutrition RCTs
In nutrition we depend on large clinical trials to answer important questions about the efficacy of nutrients. These studies are often designed to answer a single question – does vitamin D reduce fracture incidence? Do antioxidants prevent the progression of eye disease? When we try to answer questions other than what a study was designed to answer, it feels like we’re trying to put a square peg into a round hole. Sure, we can get it to fit, but does that mean it’s the right fit?
AREDS2 (Age-Related Eye Disease Study 2) was a follow up to the original AREDS trial, which showed that antioxidant supplementation decreased the progression of age-related macular degeneration (AMD) by 27%, a leading cause of blindness among Americans. AREDS2 measured whether specific changes to the original AREDS formula impacted the efficacy of supplementation, including the inclusion of either omega-3 fatty acids or lutein & zeaxanthin. The AREDS2 investigators found that the inclusion of lutein and zeaxanthin was beneficial for those who did not already consume large amounts of lutein and zeaxanthin in the diet and was an effective substitute for beta-carotene. This was an extremely important finding regarding eye health, and like AREDS1, AREDS2 was a landmark clinical trial in nutrition (Full disclosure: DSM Nutritional Products provided support for the AREDS2 study in the form of donated materials).
Yesterday saw the release of secondary analysis data from AREDS2 examining the role of lutein/zeaxanthin and omega-3 fatty acids in the prevention of cognitive decline, which, to cut to the punchline, found no effect of either supplement on cognitive function. But why? A few reasons:
1) There’s no true placebo group. Yes, some subjects did receive placebo versions of either lutein/zeaxanthin and EPA/DHA, but remember that all subjects were invited to participate in a secondary randomization to receive one of four different variations of the original AREDS formulation (which contained Vitamin C, Vitamin E, Zinc, Copper, and Beta-Carotene). Of the 4,203 participants, 3,036 (72.2%) were randomized to receive one of the 4 variations of the formula and 1,148 (27.3%) opted to take the commercially available AREDS formulation, leaving just 19 subjects (0.5%) who received no AREDS formula. In addition, 89% of subjects received a daily multivitamin (Centrum Silver). Since virtually everybody in the study was receiving supplemental nutrients, what’s our comparison here? The nutrients in the “placebo” are actively involved in cognitive function. Case in point: look at another large clinical trial from last year which showed that vitamin E supplementation reduced the progression of Alzheimer’s disease. If your “placebo” group is receiving 16x the recommended dietary allowance of a nutrient involved in maintaining cognitive function, how do you expect to see a benefit in your treatment group?
2) This wasn’t the right patient population to study, for a few reasons. First, these were fairly nutrition-conscious individuals: at baseline, the median DHA + EPA intake was 190-200 mg/day and the median lutein + zeaxanthin intake was 2553-2735 mcg/day, which is a far cry from the national averages of 90 mg/day of EPA + DHA and 1734 mcg/day of lutein + zeaxanthin seen among American adults. Second, look at statin use – an intervention that has been shown to decrease serum DHA concentrations – nearly 45% of their population was using statins at baseline. Third, what about genetics? Previous research has shown that the APOE4 genotype can impact the efficacy of supplemental DHA with respect to cognitive function, so it may have been worthwhile to see if a gene/diet interaction was present in this study. Fourth, this was not a population that suffered from cognitive impairment from the start, nor was there a cutoff for cognitive function present during recruitment. Yet, those with mild cognitive impairment or age-related cognitive impairment seem to be the ones most likely to benefit from supplemental omega-3 PUFAs. These were subjects selected based solely on the health of their eyes, not their brains.
3) Was this the right dose? It is widely believed that because EPA does not accumulate in the brain whereas DHA does, that DHA plays a more active role in brain health and the prevention of cognitive decline. In the AREDS2 analysis, 1000 mg/day of omega-3 PUFAs were given, comprising of 650 mg EPA and 350 mg DHA. A recent meta-analysis shows that intakes >580 mg DHA/day can improve cognitive functions such as episodic memory, which suggests that perhaps the amount of DHA given in AREDS2 was insufficient to provide a cognitive benefit.
Does all this mean that AREDS2 was a bad study? Absolutely not - AREDS2 was a landmark clinical trial that was designed for a very specific purpose, namely determining whether alterations to the AREDS formula impacted the progression of AMD. But does that make it the right trial to answer questions about nutrition and cognitive decline? When used for this purpose, AREDS2 is subject to common problems related to interpreting clinical trials: design issues, patient selection, and dosing concerns. It just doesn't fit quite right, much like putting that square peg into the round hole.
Chew EY, Clemons TE, Agron E, et al. Effect of omega-3 fatty acids, lutein/zeaxanthin, or other nutrient supplementation on cognitive function: the AREDS2 randomized clinical trial. JAMA 2015; 314(8): 791-801.
AREDS2 Research Group. The Age-related Eye Disease Study 2 (AREDS2): Study design and baseline characteristics (AREDS2 Report Number 1). Ophthalmology 2012; 119(11): 2282-2289.
AREDS2 Research Group. Secondary analyses of the effects of lutein/zeaxanthin on age-related macular degeneration progression: AREDS2 Report No. 3. JAMA Ophthalmol 2014; 132(2): 142-149.
Age-Related Eye Disease Study Research Group. A randomized, placebo-controlled, clinical trial of high-dose supplementation with vitamins C and E, beta carotene, and zinc for age-related macular degeneration and vision loss: AREDS Report No. 8. Arch Ophthalmol 2001; 119(10): 1417-1436.
Arterburn LM, Hall EB, Oken H. Distribution, interconversion, and dose-response of n-3 fatty acids in humans. Am J Clin Nutr 2006; 83(6 Suppl): 1467S-1476S.
Cederholm T, Salem N, Palmblad J. Omega-3 fatty acids in the prevention of cognitive decline in humans. Adv Nutr 2013; 4: 672-676.
Nozue T, Michishita I. Statin treatment alters serum n-3 to n-6 polyunsaturated fatty acids ratio in patients with dysplipidemia. Lipids Health Dis 2015; 14: 67. doi 10.1186/s12944-015-0066-6.
Quinn JF, Raman R, Thomas RG, et al. Docosahexaenoic acid supplementation and cognitive decline in Alzheimer Disease: a randomized trial. JAMA 2010; 304(17): 1903-9111.