Assorted Links

The Baltimore Shipyard Study

In a comment on my last post, Sean Estey described a study of Baltimore shipyard workers, some of whom handled radioactive materials. The ones exposed to more radiation were healthier than those exposed to less. The difference in death rate was huge: 25%. This is so large and consistent with other data I doubt it is due to a confounding.

You can read more about this study here and here. If one quarter of all deaths are due to suboptimal stimulation of repair systems, that’s extraordinary news. The study was finished around 1990. The plausibility of such a large benefit should have led to experiments. The observation that people in mountain states (such as Colorado) have less cancer than those in gulf states (such as Alabama) as well as greater radiation exposure suggested to John Cameron, a professor of toxicology, an experiment in which some gulf state residents are exposed to enough radiation to bring their total exposure up to what mountain state residents receive. This has yet to be done.

In a paper about the effects of low-dose radiation, the authors say we should ignore the Baltimore study because of “the healthy worker” effect — the possibility that persons in one exposure group were healthier than those in another exposure group because workers are healthier than non-workers (and fitness for work may have differed between the exposure groups in the Baltimore study). They give three examples to illustrate the healthy worker effect. In these examples, a group in which everyone has a particular job were healthier than the general public, which includes many people without a job. In their examples, the median effect of being in the full-employment group (in which everyone has a job) is a 10% decrease in mortality compared to the general-public group (in which some people don’t have a job because of disability). That should give a good idea of the maximum size of the healthy worker effect — when something is explicitly varied, that’s what happens. The Baltimore study compares person with job to person with job, not person with job to person without job. This suggests that in the Baltimore study, the healthy worker effect was smaller than the effect in the examples, meaning smaller than a 10% reduction. Such an effect cannot explain a 25% reduction.

A comment by Alrenous on my earlier post linked to a 2007 study of people in Taiwan whose apartment building was accidentally contaminated with radioactive materials. By the time of data collection, they had gotten far less cancer (3% of what would have been expected) than the general Taiwan population. A healthy worker effect cannot explain this. Again, the reduction is so great it is unlikely to be due to confounding.

If I could buy something to put under my bed that would expose me to the level of radiation received by people in Colorado, I would.

Will Like vs. Might Love vs. Might Hate

What to watch? Entertainment Weekly has a feature called Critical Mass: Ratings of 7 critics are averaged. Those averages are the critical response that most interests me. Rotten Tomatoes also computes averages over critics. It uses a 0-100 scale. In recent months, my favorite movie was Gran Torino, which rated 80 at Rotten Tomatoes (quite good). Slumdog Millionaire, which I also liked, got a 94 (very high).

Is an average the best way to summarize several reviews? People vary a lot in their likes and dislikes — what if I’m looking for a movie I might like a lot? Then the maximum (best) review might be a better summary measure; if the maximum is high, it means that someone liked the movie a lot. A score of 94 means that almost every critic liked Slumdog Millionaire, but the more common score of 80 is ambiguous: Were most critics a bit lukewarm or was wild enthusiasm mixed with dislike? Given that we have an enormous choice of movies — especially on Rotten Tomatoes — I might want to find five movies that someone was wildly enthusiastic about and read their reviews. Movies that everyone likes (e.g., 94 rating) are rare.

Another possibility is that I’m going to the movies with several friends and I just want to make sure no one is going to hate the chosen movie. Then I’d probably want to see the minimum ratings, not the average ratings.

So: different questions, wildly different “averages”. I have never heard a statistician or textbook make this point except trivially (if you want the “middle” number choose the median, a textbook might say).  The possibility of “averages” wildly different from the mean or median is important because averaging is at the heart of how medical and other health treatments are evaluated. The standard evaluation method in this domain is to compare the mean of two groups — one treated, one untreated (or perhaps the two groups get two different treatments).

If there is time to administer only one treatment, then we probably do want the treatment most likely to help. But if there are many treatments available and there is time to administer more than one treatment — if the first one fails, try another, and so on — then it is not nearly so obvious that we want the treatment with the best mean score. Given big differences from person to person, we might want to know what treatments worked really well with someone. Conversely, if we are studying side effects, we might want to know which of two treatments was more likely to have extremely bad outcomes. We would certainly prefer a summary like the minimum (worst) to a summary like the median or mean.

Outside of emergency rooms, there is usually both a wide range of treatment choice and plenty of time to try more than one. For example, you want to lower your blood pressure. This is why medical experts who deride “anecdotal evidence” are like people trying to speak a language they don’t know — and don’t realize they don’t know. (Their cluelessness is enshrined in a saying: the plural of anecdote is not data.) In such situations, extreme outcomes, even if rare, become far more important than averages. You want to avoid the extremely bad (even if rare) outcomes, such as antidepressants that cause suicide. And if a small fraction of people respond extremely well to a treatment that leaves most people unchanged, you want to know that, too. Non-experts grasp this, I think. This is why they are legitimately interested in anecdotal evidence, which does a better job than means or medians of highlighting extremes. It is the medical experts, who have read the textbooks but fail to understand their limitations, whose understanding has considerable room for improvement.

Flaxseed Oil Alert: Don’t Take When Pregnant

From a press release:

A study has found that the risks of a premature birth quadruple if flaxseed oil is consumed in the last two trimesters of pregnancy. The research was conducted by Professor Anick Bérard of the Université de Montréal’s Faculty of Pharmacy and the Sainte-Justine Hospital Research Center and Master’s student Krystel Moussally.

In Canada, 50 percent of pregnant women take prescription medication. Yet many of them prefer to use natural health products during the pregnancy. “We believe these products to be safe because they are natural. But in reality, they are chemical products and we don’t know many of the risks and benefits of these products contrarily to medication,” says Bérard.

Bérard and Moussally set out to conduct one of the largest studies ever undertaken on by analyzing data from 3354 Quebec women. The first part of the research established that close to 10 percent of women between 1998 and 2003 used natural health products during their pregnancy. Before and after pregnancy they were respectively 15 and 14 percent to use these products. The increase means that about a third of women consuming natural health products stopped during the pregnancy.

The most consumed natural health products by pregnant women are chamomile (19 percent), green tea (17 percent), peppered mint (12 percent), and flaxseed oil (12 percent). Bérard and Moussally correlated these products to premature births and only one product had a very strong correlation: flaxseed oil.

“In the general population, the average rate of premature births is 2 to 3 percent. But for women consuming flaxseed oil in their last two trimesters that number jumps up to 12 percent,” says Bérard. “It’s an enormous risk.”

The correlation existed only with flaxseed oil, yet women consuming the actual seed were unaffected. Even if more studies must be undertaken to verify these results, Bérard recommends caution when it comes to consuming flaxseed oil.

Thanks to Joyce Cohen.