The spice must flow: will chili powder extend your life?

For the video version of this post, click here. As a huge Frank Herbert fan, the opportunity to review an article linking consumption of a spice with longevity was just too good to pass up.  But this is not the spice melange, hailing from Arrakis, Dune, the Desert Planet. Unfortunately, today, we’re talking about chili powder.

An article appearing in The BMJ purports to show a link between higher consumption of spicy foods and lower overall mortality.  Let’s get something out of the way right up front. I am rather dubious about any study that claims that a single dietary element has any long-term effects.  Dietary “patterns”, sure.  But bacon, probably not.

This was a study using the very large China Kadoorie Biobank. Starting in 2004, roughly 500,000 individuals from all over China underwent a baseline survey, and were followed for a median of around 7 years.  The survey included a qualitative dietary questionnaire, which asked, among other things, about how frequently spicy food was consumed. It was limited in that it didn’t ask much about other dietary factors, beyond a rather simple frequency of red meat, vegetable, and fruit consumption.

Roughly 50% of those surveyed reported spicy food intake of less than once per week.  About a third said they ate spicy food almost every day.  The rest fell somewhere in between.  Eating spicy food was associated with smoking cigarettes, consuming alcohol, and was very strongly associated with rural living.

The study excluded individuals with pre-existing cancer, stroke, or heart disease, so death rates were rather low: around 1 death per every 160 person-years of follow-up in the no-spice group and 1 death per every 170 person-years of follow-up in the spice-every-day group.  If you do the math, you’d find that you’d need to take around 2800 mild food eaters and turn them into spice-aholics to save one life.

Speaking of spice-ahol, I should mention that this effect was only seen in those who didn’t drink alcohol – the one thing that makes spicy food worthwhile. <Sigh>.

We should give the authors credit for some decent sensitivity analyses – excluding those who died within a couple of years of enrollment, for instance (as sicker people might be avoiding spicy foods). But there was limited detail on how physical activity was captured – and the strong rural skew of spice consumption makes me wonder if some of this is being confounded by daily physical activity. I should also note that there was no dose-response here – which is usually a red flag when you’re trying to infer causality.

I’d like to soapbox for a minute if I may.  This trend to look for “functional foods” – specific dietary elements that have some specific benefit: your Kale, your quinoa, your chili powder – is tempting because of the feeling of control it provides.  You can take your life into your own hands – you don’t need a prescription, you don’t need a doctorate, you just need this piece of knowledge – perhaps handed down through the ages – to protect you from any manner of ills. It’s a nice thought.  But perhaps we should be gratified that it is not so.  Or else he who controls the spice, could control the universe.

The P-Value is Ruining Medical Science

The P-value, that most vaunted of metrics. For decades, it has been the standard by which we judge whether studies are positive or negative.  But it has a major, major problem.

I blogged about it this week on medpagetoday.com.

Take a look.

Lying, Cheating and Stealing with Statistics: A How-To Guide

 

If we learned anything from Spiderman, it's that with great power comes great responsibility. Statistics is a great power - but people continue to misuse it, harming all of us by publishing studies that are not fully supported by the data.

Using the great fake dataset (a bunch of completely random data) I walk you through some of the techniques researchers can use to massage positive results out of a negative study.  Read the full article here.

Exercise is good. But is intense exercise good?

This week in 150 seconds, I took a look at a randomized trial appearing in the Annals of Internal Medicine that looked at how exercise intensity might affect waist circumference.  The bottom line?  Anything is better than nothing.  The study was marred by a high drop-out rate, but this is par for the course in exercise studies.  Overall, the weight of the evidence continues to favor getting a move on.

Video below:

Workouts for Weight Loss: How Much is Enough? | Medpage Today.