Over at Communication Breakdown, my Scilogs-brother and science communicator par excellence Matt Shipman has brought out an interesting post, highlighting the problems in health research coverage by reporters as well as public information officers writing news releases. Matt exhorts these communicators to pay attention to three important concepts: context, limitations, and next steps.

I understand and accept the point about context: it is important and necessary to present a given set of observations/results in the context of existing work in the field, and how it corroborates or differs from what we already know. (This, incidentally, is what goes in the Introduction and Discussion sections of scientific research papers.) Matt also appeals for honesty, integrity and responsibility in these communications, and advocates the raising of questions, including difficult ones, regarding study limitations and next steps in prospect, before the results are presented for wider consumption.

That said (and accepted), I’d posit that the whole issue is perhaps even more complex. The way I see is that the news reporters (including those who write on science and health news), by trade, are seeking to provide the reading/listening public with what the public craves: not news, not merely information, but hope. Cynically put, hope makes good copy. And hope presented as a non-nuanced morsel makes very good copy.

That, however, is not the complex issue I referred to above. The complexity comes from the concept and understanding of nuance, not merely in the sense of subtlety, but an appreciation of the nature of science. That nuance is necessary – I’d even say, critical – when reporting science/health news; it is important, both for the reporters and the reading public, to understand that science, by nature, is a progressive, observational endeavor, which attempts to gradually build towards a larger picture by using smaller building blocks, like a jigsaw puzzle. Sometimes the puzzle piece at hand, which fits on one side, may not actually be the right fit when all sides are considered. Even worse, the larger picture may sometimes change; complex systems, such as body physiology, often have too many variables that cannot be controlled for, thereby making even the appearance of the picture fuzzy. This is the reason we have the scientific method, this is why we have hypothesis generation, the concepts of dependent and independent variables, tests, constants and controls, as well as rigorous testing of the hypothesis. Perhaps the most important power of science is the ability to reject or modify a failed hypothesis – one that is not supported by observational data.

And this is precisely why the nuance – or call it ‘balance’ if you will – is critical BOTH in reporting the results of scientific research, AND in understanding those results while reading. As Matt notes in his post:

“The reason the stories contradict each other is because the studies contradict each other,” Virginia Hughes wrote in a May 12 post at her blog, Only Human.
[…] she goes on to explain why it can be so frustrating to write about complex health research for an audience of readers that seem to be looking for health tips, not nuance-laden articles about the scientific process or incremental progress in our understanding of one facet of human health.

It is important to understand that the studies will sometimes contradict each other; the studies at two different times or places may not be conducted exactly identically, and not all external factors that may affect the outcome of such studies may be accounted, and adequately controlled, for. And it is important to understand and emphasize that THIS IS OKAY. It happens. This is how scientific knowledge progresses. When two similar studies show contradictory outcomes, trying to find the root cause for the apparent contradiction can also sometimes provide valuable clues about the overall picture.

To clarify, I am not advocating for the reading public to abandon all hope; that’d be cruel. I am, however, advocating for a modicum of understanding of nuance as a filter, through which the presented results and observations must pass. I am appealing for the application of nuance/balance exactly in the situation hypothesized by Virginia in her NatGeo Resveratrol story (linked above):

Take a look at those headlines again. I suspect a general reader is not coming away from those saying, “Gee whiz, look at the long and bumpy road to scientific progress!” They’re more likely to be saying, “When will those scientists get their act together?” Or worse, “Why do we keep dumping money into this capricious discipline?”

An understanding of science and appreciation of the scientific process can help alleviate this faulty (but popular, especially amongst many US politicians) line of thinking. For example, the right degree of nuance applied in the presentation and understanding of the health effects of Resveratrol over the years would perhaps have clarified the following points (indicated in the recent JAMA Internal Medicine paper referenced by Virginia in her post):

  1. Polyphenol Resveratrol (found in grapes, red wine, peanuts, chocolate, certain berries and plant roots) may indeed have potent anti-inflammatory and potentially beneficial effects, as seen in simple cell-culture based in vitro systems and well-controlled complex environments of animal models.
  2. Animal model studies have identified the possible mechanism by which Resveratrol may exert its beneficial action.
  3. The benefit seen in some clinical interventional trials may well be attributable to those known, as well as perhaps some hitherto-unknown, mechanisms.
  4. The apparently contradictory study published in JAMA Internal Medicine does not contradict any of those earlier observations. What instead it does call into question is the generalizability and population-wide applicability of those observations. And this is an important aspect of epidemiology.
  5. The 9-year study failed to find, in a specific community of a certain age and geographical origin, any association between Resveratrol metabolites in urine and inflammation, heart disease, cancer or overal mortality. What this observational study does not (in fact, cannot – by design) comment on are:
    • Possible effect and/or interaction of Resveratrol with co-morbidities in this population.
    • Cellular/tissue-level, local effects of Resveratrol vis-à-vis the appearance of its metabolites in urine.
    • Incidence of disease (e.g. cardiovascular or cancer) via mechanisms that fall outside of the known mechanistic gambit of Resveratrol.
    • Other possible factors pertinent to that geographical region, among other things, such as those pointed out by the authors in the Discussion: variability in exposure to Resveratrol, interindividual variation, and variability of host-gut microbiota.

To illustrate the critical need for nuance in the understanding of study results, I offer this example from the data in Table 2 of the article; the corresponding description says: Participants who died were older, more likely to be male, with lower education, lower BMI, physically inactive, with diabetic fasting plasma glucose and higher CRP, IL-6, and TNF concentrations, higher mean arterial blood pressure, and lower total, HDL, and LDL cholesterol levels. — From this statement, if one were to conclude that younger, educated, active, non-diabetic women with low inflammation and low blood pressure die less despite having higher total cholesterol outside the context of this study, that conclusion would be simplistic, erroneous, expansive and just as ridiculous as it sounds.

To be sure, my call for greater nuance also applies to the interpretation and presentation of the studies that have extolled in the past the virtues of Resveratrol and other such naturally-occurring substances with health-benefit potential. As I indicated above, it is to be an important filter.

I am staunchly of the opinion that this appreciation for nuance, which I associate with a spirit of enquiry, can come only from unfettered access to education – especially science education. Given the abysmal level of science literacy in the US at present, I think we, as scientists and science communicators, have an urgent responsibility to get that important message across to the news reporters, reading public, as well as policy makers.