Prof. Jean-Pierre Habicht and Prof. Gretel H. Pelto, Division of Nutritional Sciences, Cornell University, Ithaca, NY, USA
“The purpose of efficacy trials is to provide evidence of impact. The scientific requirements for efficacy trials are clear and well accepted. Multiple micronutrient supplementation efficacy trials require a comparison of an intervention group that receives the supplementation with a control group that does not. The two groups must be statistically and epidemiologically comparable, a condition attained through randomization. Without information on changes over the intervention period among a group that did not receive intervention, one does not know the full impact of the supplementation. However, the impact of supplementation shown in efficacy trials is only the effect of additional (!) micronutrients given to the intervention group but not to the comparison group. Other than in randomized controlled trials (RCTs) testing drugs there is no comparable placebo group in supplementation trials: there is no control group that does not receive any micronutrients because every (!) participant has independently of a supplementation dietary micronutrient intakes.
Another feature of the multiple micronutrient supplementation trials is that they demonstrate that the populations in which they were conducted were deficient in at least some of the supplemented micronutrients. This is an important contribution because indicators of deficiency are inadequate for many micronutrients. In fact, efficacy trials are fundamental tools for demonstrating remediable deficiencies, and, when they are well designed, they can also identify those who benefited from the supplementation.
Trials to demonstrate public health benefit must show that the magnitude of benefit is adequate for policy considerations. In the case of supplementation trials the magnitude that is studied is the added benefit of the micronutrient intervention relative to the lack of supplementation. It is important to determine if the micronutrient interventions are worth the added costs and potential side-effects. The most common trade-off for decisions about public health actions are between the costs of planning, implementing, and sustaining programs compared to the positive returns in health and well-being. Assessing the positive returns requires information on the magnitude or ‘adequacy’ of the response. Unfortunately, at present there is no commonly accepted measure for determining adequacy of impact, unless one has all the benefits and costs at hand, in which case cost-benefit analyses can be used.
The impact of supplementation can be expressed in absolute values or effect sizes. However, neither the absolute impact nor the effect size can be used to ascertain the impact considered in relation to the potential to respond to the intervention. Potential to respond is a major source of biological heterogeneity of response across and within randomized controlled trials. Multiple factors affect the potential to respond to micronutrient interventions. Individuals in the population who are already replete do not respond to additional micronutrients. In a population with endemic under-nutrition there will always be some individuals who are not deficient and others whose deficiencies are so moderate that they will have only a marginal response at best. Clearly, these individuals dilute the impact of the intervention, but it may not be easy to identify them because measures for micronutrient deficiencies are inadequate. Others may not respond because of genetic influences on absorption and utilization. Also, individuals whose diets lack an essential component required for absorption or utilization do not have the potential to respond (e.g., fat is needed to absorb beta-carotene to improve vitamin A nutrition). Another source of heterogeneity within and across RCTs is behavioral heterogeneity in aspects of the implementation. It is essential to use efficacy trials to examine the extent to which the results may have been attenuated by problems in the pathway from introduction of the supplement in the intervention group to actual consumption (e.g., in the right amount, right frequency, for the right duration) by the intended subjects.
When the results of efficacy trials include findings that are counterintuitive or do not fit with expectations, such events typically lead to further research to identify the factors that explain them. In fact, this is a primary process in scientific work. When the results of meta-analyses of RCTs are used to derive policy implications, there should be a similar investigative process in the form of plausibility analysis. Plausibility analyses include assessing the biological intervention impact pathways by measuring the biological intermediary changes between the intervention at the level of the individual and the outcome. This assessment ascertains which of the plausible mechanisms explained the impact or lack of impact. This type of analysis forces one to consider more mechanisms of biological action than are implicit in the design of most RCTs.”
Based on: Habicht J-P. and Pelto G. H. Multiple micronutrient interventions are efficacious, but research on adequacy, plausibility, and implementation needs attention. Journal of Nutrition. Published online in December 2011.