Weekly curry unlikely to delay dementia

While we’re all familiar with the old “apple a day” adage, the Daily Mail recently asked us to consider “Could a curry a week PREVENT dementia?”. While it’s not the first time that spicy food has been promoted as having a possible role in protecting our thinking skills as we age, the headlines and the actual research were not entirely in agreement. The Telegraph offered a slightly more measured headline “Eating curry may help fight off dementia, new study suggests” adding those caveats of “may” and “study suggests” to keep us guessing, while The Spectator left no room for doubt, declaring “Sorry, a weekly curry won’t prevent dementia”. Given that these headlines were all drawn from the same research study, it wouldn’t be surprising if people were left unsure what to think (and eat).

What did the researchers do?

The study was primarily based at Edith Cowan University in Western Australia, and was recently published in the British Journal of Nutrition. The researchers summarised existing evidence that curcumin, “a constituent of the widely used spice Turmeric”, might interfere with the accumulation of the main proteins that lead to the development of Alzheimer’s disease. A lot of that evidence has been drawn from animal models or short-term studies with humans; their study therefore set out to explore how curcumin might affect thinking skills in a group of adults over a period of one year.

Very briefly, 160 people were recruited into the study, and all were free from existing cognitive impairments based on objective assessments (standardised tests specifically designed to examine someone’s cognitive status). The group were randomly split with half receiving curcumin capsules (150mg, 3 times a day) and the other half receiving placebo capsules (identical in size, shape and colour, and taken at the same frequency as those getting the curcumin capsules, but containing no active ingredients). To monitor adherence, all participants visited the clinic every 3 months and swapped their empty capsule bottles for their next dose and submitted their “capsule diaries”. Anyone with less than 70% adherence was not included in the main analysis.

Because of non-compliance and a number of other reasons (health problems or adverse reactions to taking the capsules), 96 people were retained in the final study with an average age of 66 years old, 57 in the placebo group and 39 in the curcumin group. Everyone had completed a range of mental tests before being randomised to their group which were then repeated at a 6-month follow-up and again at 12 months. The assessments included a general cognitive ability screening test, a test of learning and memory, a test of generating words according to specific rules, and a test of perceptual motor speed.

And the results? Well, the results are somewhat underwhelming (especially given the media reports we’ll come onto in a moment). For one of the mental tests, the general cognitive ability test, it was suggested there was a difference in performance between the curcumin versus placebo groups. Those in the curcumin group seemed to be scoring better than the placebo group. However, when the researchers examined this, it was only apparent at the 6-month follow-up, and appeared to result from an improvement in performance in the curcumin group as well as a decrease in the placebo group. There were, however, no differences at the 12-month follow-up. And there were no differences on any of the other measures.

Or to put it simply, there was no cognitive benefit over a 12-month period associated with the consumption of curcumin.

The researchers carefully noted this in their report: “this association was driven by a decline in function within the placebo group at the 6-month follow-up that was not observed in the curcumin treatment group. No differences were observed between the treatment groups in all other clinical and cognitive measures.” As the researchers then went on to suggest, it might be there is a benefit, but perhaps the group they were studying were generally too healthy to see those over just a year, or maybe there are other thinking skills that would be spared by curcumin consumption but that weren’t assessed in the current study.

A good point to note, however, is that the study is the kind that we would consider at the peak for being able to make causal inferences (if we had found an association, this design allows us to suggest it’s likely to be the thing we altered that “caused” the outcome to occur). It was an intervention study, and more specifically a randomised, placebo-controlled, double-blind study, meaning people didn’t get to choose what group they were in, the effect of the proposed “treatment” was compared to a group who also thought they were getting the same treatment, and neither the participants or those doing the assessments knew who was in what group.

What did the media say?

Generally, the media led on the curry. The first line of the Huffington Post article stated “Eating a portion of curry once a week may reduce your risk of developing dementia, new research suggests”. No, the research didn’t say that. The Huff piece went on “The researchers found curcumin has the potential to improve memory function for adults by blocking proteins that destroy neurones in the brain.” Again, no. The researchers simply didn’t say anything like this and they did not measure any proteins in the brain in their study.

Crucially, the Huffington Post focussed entirely on the differences observed at 6 months and didn’t refer to these differences no longer being apparent at the end of the 12 month follow-up. Luckily, the piece did at least lift a quote from The Telegraph coverage in which Laura Phipps, from Alzheimer’s Research UK, said “there is currently no conclusive evidence [curcumin] could prevent or treat dementia in people”. Sadly, however, readers needed to wait to the latter part of the media report for that. The need for that expert opinion isn’t simply to ensure balance in a contested area, but rather it is a vital piece of the coverage as there is actually no story to report in the first place.

It was common for the coverage to start almost identically, the Mail opening with “Eating curry at least once a week may help ward off dementia, according to scientists”, and continuing “the study shows that a spice commonly used in the dish boosts brain power and protects against memory loss in old age. Curcumin, the key chemical in turmeric used in everything from mild Kormas to the hottest Vindaloos, is believed to delay or prevent dementia symptoms. A study of middle aged and elderly people found those who popped a capsule of the stuff three times a day had better memories than those given a dummy pill.” Based on the message above, hopefully you can see these opening remarks are, at best, overstatements, or at worst simply untrue. Most worrying of course are those in the latter category, including “The year long trial published in the British Journal of Nutrition found evidence curcumin blocks rogue proteins called beta amyloid which clumps together and destroys neurons”. Again (again), no. The researchers talked about how these proteins might be implicated in the mechanism (if there was any benefit, which remember there wasn’t at 12 months), but as discussed above the study didn’t include any measurements of these.

Interestingly, the quotes attributed to the lead author of the study, Stephanie Rainey-Smith, focus on what the study sought to investigate and that more work is required, but there was no direct quote from her or other members of the team making the claims reported. It’s not clear how the media have got this one so wrong as there doesn’t appear to be an overzealous press release from the researchers, the university or the journal (though I’m happy to be corrected if that exists; you can read more about the potential pitfalls of press release reporting here). Ever odder, the Telegraph and Mail pieces share many similar or identical phrases though were written by different journalists. When based on press releases, such consistency isn’t unexpected (see our How to “Research the Headlines” series for more).

How to evaluate the coverage then? There are errors that might be easier to spot. Curry consumption is being conflated with the curcumin capsules used in the study. The research paper only has the word curry three times, though two of these are in the reference list where it is included in the title of earlier papers; the one occurrence in the main text simply describes one of those studies. At no point do the authors make the link between their very specific and controlled study and how that might relate to real-world curry consumption. A second key error is that Alzheimer’s or dementia are often being used as shorthand in the media reports for the cognitive assessments used in the study, though these are clearly very different outcomes. For any media reporting of a dietary factor being associated with thinking skills, look out for those points. When the results themselves are being misrepresented is, unfortunately, much less easy to spot without access to the published study and a knowledge of how this was designed an analysed.

On a final small point, as the study wasn’t about curry consumption per se the researchers didn’t give an indication of how their 1500mg curcumin per day (spread over three doses) might equate to your average curry. The Spectator suggested the dosage used was “higher than in a mild curry”, and even though the piece also misreported a potential cognitive benefit of curcumin consumption (wrongly suggesting this was also apparent at the 12 month follow-up), overall their coverage was far more sceptical of the likely real-world impact of the findings, going as far as to conclude the “lay press hasn’t let this lack of causality get in the way of a spicy headline”.

The bottom line.

You’ll pardon the pun, but any dietary factor being reported as delaying or stopping dementia or other similar conditions needs to be taken with a pinch of salt. In this case, the dietary factor was conflated with a larger group of foods, and the outcome being reported (Alzheimer’s or dementia) wasn’t specifically studied. On the whole, those working in this field would suggest there are unlikely to be any dietary “silver bullets”, and that a balanced whole-diet approach is what people need to focus on. For now, heart-smart diets are those best encouraged for all round health.

Rainey-Smith S. R. et al. (2016). Curcumin and cognition: a randomised, placebo-controlled, double-blind study of community-dwelling older adults. British Journal of Nutrition. DOI: 10.1017/S0007114516001203