A primer on making sense of nutritional headlines
We see them so often that we tend to blend them out: “Forget dieting, this pill will help you lose weight!” Or: “These food additives could be changing your gut microbes.” Day in, day out, we are bombarded by health and nutrition related media headlines, ads, and social media messages. And as outrageous as some may seem, others accurately present the findings of exemplary research. Only, with so much information overload, it can be difficult to tell apart the good science, and the good reporting of it, from the bad.
The good news is that a little knowledge can go a long ways in improving your ability to tell quality research from quackery. By understanding the different types of studies and how they work, you’ll be better prepared to dissect the next nutritional headline you come across and decide for yourself whether to take it seriously or simply keep calm and carry on reading.
In vitro extrapolations
The media are rife with reports that take observations made on single cells in the lab and extrapolate them to human beings. But the fact that a compound – be it caffeine, alcohol, a vitamin, or any other molecule – affects cells one way when administered outside the body by no means proves that ingesting it will have the same effect. In vitro studies play a crucial role in research, as they are relatively inexpensive, yet can reveal biochemical mechanisms behind health and disease and help scientists which hypotheses to pursue further. But if the report you are reading is entirely based on the outcome of an in vitro study, its implications for human nutrition should be taken with a grain of salt.
Even when properly controlled and carefully executed, animal experiments in nutrition do not provide conclusive evidence for humans. While they let researchers test hypotheses in living systems, the number of factors that have to be controlled for and the many variables that can easily be overlooked mean that nutritional headlines based on animal experiments could be pointing in the right direction. But more often than not, their outcomes are not borne out in human trials.
On average, some human populations live longer, healthier lives than others. Part of this can be ascribed to cultural factors, in particular: diet. Observing populations in so-called ecologic studies – i.e. with no outside intervention – is one way that researchers seek to infer the faults and merits of different culinary habits. Basing nutritional claims on ecological studies alone, however, has limited conclusive power, as they do not allow researchers to disentangle causation from correlation – in other words, is the population healthy because of its diet, or because of other unmeasured factors?
In case-control studies, researchers compare a group of people with a specific health outcome, such as obesity (the case group), with a similar group of people lacking it (the control group). By reconstructing the participants’ past habits using questionnaires, they attempt to determine the factors that led to the outcome. These studies are largely limited by the fallibility of the participants’ memory.
Whereas case-control studies look back in time, cohort studies follow initially healthy participants over prolonged periods of time. Researchers collect data from the participants at regular intervals to identify the factors that eventually cause certain members of the cohort to contract the disease under study. Though expensive, long-term cohort studies are considered to be the most reliable “non-experimental” study type.
This takes us to an experimental study type that is considered to be the “gold standard” of nutritional research: double blind placebo controlled randomized trials. Two groups of individuals receive either the compound under study or a placebo. Neither the participants nor the researchers tasked with evaluating the results know who belongs to which group. This form of study comes as close a possible to directly linking dietary compounds to health outcomes. But in order to be conclusive, trials must be carried out on sufficiently large populations over sufficiently long durations.
Studying studies (of studies)
For every positive health claim associated with a compound, say caffeine, there seems to be a negative one as well. So where does the truth lie? To find out, researchers combine the outcomes of multiple studies in so-called meta-analyses, in which studies are evaluated in terms of their design, their conclusions, and the strength of their conclusions. Sometimes, even different meta-analyses on the same topic offer contrasting results, in which case they themselves can then become the object of a meta-analysis.
The next level?
With almost everyone connected to the Internet through their mobile devices, the fourth industrial revolution has opened new avenues for massive, yet affordable participatory research studies. By tracking diet and exercise, smartphone apps could potentially provide unprecedented amounts of information for clinical nutrition studies. And with image capture technology to estimate the nutritional content of your meals just around the corner, it could become as simple as snapping a shot of your plate every time you sit down and eat.
In the meantime, be sure to download the openfood app [link], with which you will be able to scan the barcode of over 14,000 products sold in Switzerland to learn more about your groceries than you’ll find on the label. By contributing, you will be helping us grow and refine our database of products, which, by the way, is completely open, free, and, most importantly, accessible to mobile phone application programmers for development.