Truth and the Internet

Everyone who travels by air likes to whine about the experience.  It's mostly about the TSA, but sometimes it's about the behavior of other passengers.  I wrote about that myself a couple of months ago.  Recently there was a Twitter-based account of the interaction between a fellow who works as a producer on an ABC network television show and a fellow passenger (Diane from 7A).  The story was later apparently outed as entirely fictional by its author.  Perspectives on that ranged from amusement to annoyance to musings on the assumptions people make about the veracity of what they read online.

This brought to mind the many conversations I have had with friends and relatives about our reading habits.  Not about reading online, or newspapers or magazines - but about our selection of books.  My reading of books is exclusively nonfiction, mostly history, and especially US political history.  I favor biographies, and especially those of important figures in American history, and even more particularly American presidents.  I find history fascinating.  However, I have learned that even the most scholarly works of historiography involve a prism.  I am looking at something through the eyes of another human being.  I am on my third biography of Lincoln, and three biographers see him in ways that are significantly different from each other.  Getting such varied perspectives affords me insight into the art of interpreting the historical record.

I find history much more interesting than fiction.  My sense is that, if I were reading a novel, and found myself becoming engrossed in the events of someone's life, I would stop every so often and think, oh, this didn't really happen (except in someone's imagination).  The implication inherent in that thought is that, because it didn't really happen, it doesn't matter.

So the story of "Diane from 7A" got me wondering about the way people look at such things.  Does it matter if the account of the interaction between two passengers on a plane as related via Twitter really happened? If you found it funny, is it any less funny if it didn't really happen?  My experience with funny stories people tell each other in conversation is that if something is truly amusing, they may want to know whether that really happened or was just made up, but not because it is any less funny if it's fictional.  No, it seems the reason they ask is so they will know whether the person telling the story has a great sense of humor.  After all, which is the person you want to be around: the person who has a few true funny stories (because, after all, most of us don't have funny things happen around us very often), or the person who takes ordinary experiences and weaves tales around them for everyone else's amusement?

When I read something from a conventional news outlet, whether in print or online, I expect it to be factual.  The reporters and editors have a responsibility to their readers to get the facts right and to report them as objectively as possible. But just about everyone else is, in my view, entitled to some literary license, or at least to an individualistic interpretation that may be strongly colored by an array of biases.

Tom Hulce as Mozart in the film Amadeus

When I was in college, I took three music courses.  Two of them were on specific periods (Classical and Romantic) and were taught by the same professor.  The term paper assignment was to select a composer and write about his life and works.  For the Classical period, I chose Mozart.  This was before the Internet, so I went to the big Free Library in downtown Philadelphia and read all of the biographies of Mozart I could find.  Some years later the film "Amadeus" appeared in theaters.  The story was related, at least in part, from the perspective of rival composer Antonio Salieri.  Even so, I found the portrayal of the title character jarring.  I reflected upon my readings from the decade before, when I wrote my term paper.  I looked at the film's protagonist and couldn't help thinking, "That's not the Mozart I knew."  But this was a movie, not a documentary produced for the BBC, and not a history-for-television-audiences such as Ken Burns might have produced.  The fellow who wrote the screenplay and the film's director are surely entitled to all kinds of license.  I thought they portrayed Mozart as a silly ass, and I didn't like that, but then I didn't really know Mozart.  I knew him only as his biographers depicted him, and traditionally biographies have been written by admirers.

As a lad, when I went to a public library and saw books neatly characterized as fiction or nonfiction, I thought life was just that simple.  Either it really happened, just as described, or it's the product of someone's imagination.  But life is not just that simple, and the Internet is certainly not.  My assumption nowadays is that everything I read online comes from someone who has considered himself at liberty to take license in telling a story.  Yes, I still expect "straight news" journalists to try to be factual and objective.  But they are human, and all humans have a point of view.  So I look for a variety of points of view and compare them, not assuming that any one is more reliable than others.

Sometimes when I look at things I am pretty sure they are fictional.  Some photos, for example, defy credulity, and - despite being something of a technophobe (or at least a techno-naïf) - I recognize that they've been altered.  It isn't always quite as obvious as the image to the right, but for anyone but a true babe in the woods, it should be plain to see that it's the product of an inventive mind.

"Debunking" websites like Snopes.com are a wonderful resource.  I use them all the time, taking advantage of the time and effort they put into investigating Internet-based stories and letting the rest of us know whether they have found truth or fiction.  On social media, people are always posting things that are then promptly outed as "hoaxes" by their friends who have done a little checking.  This sort of skepticism is a good thing, I suppose.  And it seems to have become the rule, rather than the exception, at least in my social media circles.  A recent posting about the death of a minor celebrity was immediately followed by a series of skeptical postings noting that all the links that turned up in a Google search seemed to stem from one source, which then was the subject of a discussion of its reliability (as might well be appropriate for an Internet source that mainly reports on celebrities).

Some of the stories I see are inspiring.  Some are funny.  Mostly I don't care whether they are truth or fiction, because their effect on my mood at the moment is independent of their grounding in reality.  It just doesn't matter whether Diane ever really sat in seat 7A.

It does matter to me whether there was anything Abraham Lincoln might have done to prevent the Civil War.  And it matters to me what he really believed, at the core of his being, about African slavery, and why he thought certain infringements upon civil liberties were justifiable in time of war.  And I guess maybe that's why I have such a strong preference for nonfiction: I want to spend my time reading about things that really matter.

The Truth About Trans Fat

A few weeks ago the Food & Drug Administration announced the opening of a comment period on its regulatory proposal to ban trans fats from being added to foods sold by industry (processed foods) and restaurants.

The latest available data suggest that the current average consumption of trans fats is about one gram per day (according to the FDA, as reported by CNN last month).  Given that experts on nutrition recommend we get no more than 1% of our daily caloric intake from trans fat, it appears the average American is already well below that.  (A gram of fat yields about 9 calories, so that would be 1% of a 900 calorie diet, far less than what most people must consume to meet their needs for energy from food calories.)

So it would seem the elimination of trans fats is unlikely to make much difference in the public health, because consumption will change little.  Certain foods that we still eat, though, contain notable amounts of these substances, and so if your diet includes significant numbers of, say, doughnuts, your trans fat intake might be a good bit higher than average.

Let's back up a little to look at how trans fats became part of our diet.

Several decades back the experts told us animal fat was bad for our health.  And we responded by switching to vegetable oils.  So the first thing to understand is the difference between a fat and an oil.  In simplest terms, these substances are all built on molecules that are esters of glycerol (glycerides).  If such a substance is liquid at room temperature, we call it an oil; if it's solid, we call it a fat.  In nature, most of the glyceride-based substances from animals are fats, while most from vegetables are oils.  There are well known exceptions.  Coconut oil, for example, is solid at room temperature but liquifies slightly above that.

On the molecular level when you look at the bonding of carbon atoms to each other and to hydrogen atoms in the hydrocarbon chain that is the backbone of a glyceride, oils have fewer hydrogen atoms.  The electrons that are part of the carbon atom, instead of participating in bonds with hydrogen atoms, form double bonds between the carbon atoms.

If we take an oil, which has fewer hydrogen atoms, and add some, we get a hydrogenated oil, which then behaves more like a fat (solid at room temperature). This is called a "partially hydrogenated oil," and you've probably seen phrases like that in the lists of ingredients on packages of food.  In nature, the hydrogen atoms tend to be on the same side of the carbon chain, a configuration called "cis" in organic chemistry.  When we add hydrogen atoms artificially, they tend to go on opposite sides of the chain, and chemists call that "trans."  There is very little "trans" in naturally occurring animal fat.

So is trans fat bad for us because it's different from naturally occurring cis fat?

We used to think just the opposite.  Those of us who are old enough to remember when margarine was first introduced will recall that it was thought to be much better for our health than butter.  Margarine (partially hydrogenated vegetable oil) is trans fat.  Butter is naturally occurring animal (cis) fat.

As the decades passed, there was intriguing evidence from population studies that trans fats might actually be worse than cis fats.  They seemed to be associated with an increased incidence of colon cancer.  And while we thought margarine was more "heart healthy" than butter, trans fats seem to be associated with worse, rather than better, findings when we have our blood tested for lipids. Specifically, trans fats are associated with higher levels of low-density lipoprotein (LDL) - the so-called "bad cholesterol."

My regular readers know that I am fascinated by the notion of truth in science. As knowledge advances, we are convinced that we are getting inexorably closer to the truth.  We believe that what we "know" now is always more likely to be the truth, or closer to the truth, than what we "knew" before.  We believe we are steadily getting better at discovering scientific truth, that our scientific methods are constantly bringing us closer to a complete and accurate understanding of scientific phenomena, including anything having to do with the life sciences - and therefore with health and nutrition.

We may be right about this.  I'd certainly like to think so.  But I believe it is important to bear in mind that evidence isn't better just because it's newer.  Any time we have new evidence, we should regard it with a certain level of skepticism - not because we are clinging to our "old" truths, but because evidence should always be examined closely, and its quality and meaning judged carefully.

So let's imagine how we might go about answering the question of whether trans fats are bad.  We'd start with a very large number of people willing to be assigned at random to any of several groups whose consumption of trans fats would be carefully controlled over a period of many years.  We'd assign them to getting, say, 1%, 3%, 5% (and so on) of their daily caloric consumption from trans fats. And then we'd follow them for two or three decades, closely monitoring health outcomes we think might be affected by consumption of trans fats.

As you might surmise, no one has done a study like that.  Instead, we have epidemiologic evidence.  We ask people to tell us about their dietary habits, we hope that their recall and reporting are correct, and we compare them with other people whose diets (as they recall and report them) are different but who are otherwise similar in ways we think might influence their health.

As you likely realize, the second kind of study is much easier to do, but it is highly vulnerable to what scientists call "confounding."  Many other factors, dietary and otherwise, may affect people's health, and trying to identify them all and make adjustments in our statistical analysis, is tremendously challenging.

Thus a study done prospectively, in which the use of large numbers of subjects and the process of randomization minimize confounders, is much more reliable. We still have the question of whether subjects will do what they're told, but we can ask them to keep an accurate diary and see for ourselves whether they followed instructions.  A diary maintained contemporaneously tends to be much more reliable than recall and self-reporting.

Trans fats are bad.  That is the message people have been hearing for quite a while now.  And some laws have already been enacted.  Restaurants have been prohibited from using trans fats in New York City, for example.  Thus our consumption of trans fats has already declined very substantially.

So we can predict that, if the FDA declares that trans fats are no longer "generally recognized as safe" and prohibits their use in commercially prepared foods, the impact on the public health may be modest at best.  Might it affect an individual who eats a lot of doughnuts much more?  Maybe.

Doughnuts and some other baked goods are made with trans fats to give them a lighter texture.  Without trans fats, doughnuts would likely seem more oily. Partially hydrogenated oils stay in the "matrix" of the food and do not exude out of it.  If doughnuts wind up being greasier because of a new FDA regulation, some people won't like it.  I suspect the folks who make doughnuts will figure out a way to solve that problem, but I'm no food chemist, so I really don't know.

Let us keep in mind what we're talking about here.  This is a proposed government regulation that would say, "Thou shalt not" use trans fats - unless you can convince the FDA that doughnuts made with trans fats really are safe.

What level of evidence should we expect for such an edict?  There are some things for which we very reasonably believe we don't need evidence from a prospective, randomized, controlled trial.  No one would seriously suggest, to offer my favorite example, that we need to do a study to find out whether it really is safer to jump out of an airplane with a parachute than without one.

When it comes to public health and the use of case control studies to try to figure something out, however, there are so many sources of error, bias, and confounding, that I have to wonder whether what we "know" today is clearly much closer to the truth than what we "knew" when the food scientists first created margarine to save us from animal fat.  Just a quick review of the dizzying array of opinions on whether polyunsaturated versus monounsaturated vegetable oils are best for us will give you an idea of how challenging it is to be certain about things in this realm.

I don't have a problem with the government telling us what is good for us or outlawing what it thinks is bad for us.

Ha!  I probably caught at least a few of you not paying close attention, reading that sentence and thinking I really meant that.  The fact is, I do have a problem with that, especially when the evidence for what is good or bad for us is somewhat less than compelling.  And this, I believe, is an instance in which it is quite reasonable to doubt the evidence and to ask whether what we "know" today is scientific truth.