Holford Watch: Patrick Holford, nutritionism and bad science

Patrick Holford wants everyone who has an interest in childcare in whatever capacity to watch Trevor MacDonald 13 July, 20:00 where he will present his results to the world by the rigorous mechanism of an easy-listening TV programme rather than peer-review. The programme will report the outcome of a Food for the Brain (FFTB) pilot project that was run in Chineham Park Primary, a poorly performing school in Basingstoke. Holford is trailing that the programme will reveal significant improvements in the children’s SAT scores, behaviour, attention and hyperactivity.

The impressive changes in the children’s behaviour and performance were achieved by following an optimal diet, supplement and exercise programme for eight months, under the guidance of a team of experts, including nutritionists, cookery and catering consultants, educational psychologists and fitness teachers. Parents, teachers, contract caterers and the children themselves were all involved in the re-education programme and the results speak for themselves. The school’s SAT scores have improved dramatically, along with significant improvement in the children’s behaviour, attention and learning abilities – especially among children with challenging behaviour, attention and learning difficulties.

“I am really delighted with these results.” Says Gwen Clifford, the school head. “The project has made a big difference.”

It is interesting to see what impact such an input of expertise can have in Chineham Park Primary. Of course, it may be difficult to disentangle any one element as the major contributor to any results because they do seem to work in a synergistic way. Mrs Clifford is to be commended for the introduction of a number of innovations such as the acclaimed Family Reading Scheme which appears to have been running concurrently with the FFTB project. We’ve consulted the schools rankings in the DfES Standard Attainment Tests (SATS) league tables and it looks like Mrs Clifford has already presided over a near-doubling of points within the year preceding the project.

Sir Trevor is a genteel host so we are unlikely to see any Paxman style questioning of the parties involved in the collaboration. It’s going to be impossible to appreciate the full achievement of the pupils, school, expert input and the innovative additional schemes in a 30 minute programme so Holford Watch looks forward to the release of the report. Not least because it is rare to see a good explanation and understanding of the statistical niceties of potentially confounding variables but we’re sure that FFTB will have had access to the requisite expertise and resources, courtesy of the members of its Scientific Advisory Board.

The public domain information about the Chineham Park Primary School is very unsatisfactory. The only freely-accessible information that we could find didn’t even make arithmetical sense (a recurrent theme with Holford’s work and enterprises) which is why we hope for a better analysis this time. You will notice that there were 9 children in the testing group but it was reported thus:

Based on pre-test scores seven of those tested had a score that
would have suggested ADHD and three had scores that were within normal limits.

These numbers add up to 10; possibly there were 10 in the baseline group but one did not complete the tests, this should be clearer because it muddies the already odd analysis of such a small group. Commenting on the results the writer enthuses:

Post test scores indicated that out of the nine tested only four remained in the ADHD classification and five has [sic] scores within normal limits. Of the four that had scores within the ADHD classification, two individuals had significantly reduced that score, one individual did not complete the test and one had a higher score.

Now, to Holford Watch, this looked like 4/7 children remained within the ADHD classification because 3 had presented at baseline with scores within the normal range (not that the test is diagnostic) so…It was impossible to comment on the score reduction because we had no idea whether or not it was within the standard deviation for the range. [Edited: 20:30, FFTB's experts have posted expert comments on the results some of which are comparatively sensible and others impenetrable.]

In order to aid our fuller understanding and to promote the public discussion of some important issues, it would be helpful to know the study protocol including:

• inclusion/exclusion criteria
• ethical review submission for testing physiological parameters, biochemical biomarkers and psychometric evaluations
• parental consent procedure
• additional teacher/parent training in behavioral assessment
• physiological and psychometric tests
• recording and validation of diet
• what supplements were given and how compliance was validated
• how the baseline and other data were collected and evaluated
• the weighting analysis used to filter out contributions such as the Family Reading Scheme and other innovations
• it would help if the children who were tested in the sub-group were given identifiers so it is easier to know who wasn’t tested and how this might have affected the results. [Edited in: 20:45 having seen the abysmal summary of the results and particularly the Woodhouse account.

We accept that it is difficult to make experimental protocols seem like much-watch television but that is the problem when you disseminate your findings by television rather than more soberly. Nonetheless, we at Holford Watch would like to know the above because we are convinced that the British Dietetic Association is agog at the prospect of learning more about its shortcomings and poverty of imagination/experience on the true impact of food/nutrition on health from Patrick Holford. There are rumours that some people would like to pop along a response to the splendid Chocolate Pudding Letters Review which has an excellent mission. The Review is soliciting contributions:

Please send an email to either of the Editors (refer to UIUC psychology website for contact info) stating your affiliation and which type of article you are submitting: a) short report, b) confession, c) diatribe, or d) ad hominem comment on others' work(s). Type (d) submissions are especially welcome for the first issue, but we prefer to receive these before the target article is composed and/or submitted. Videos and other multimedia works are encouraged. Please send us your most sincere manuscripts only. For authors interested in hilarious mockery of academic literautre, we recommend "The Savory Biscuit Pudding Report", "Journal of Experimental Pudding: Learning, Memory, and Pudding", "Psychological Pudding", and "Sensory and Pudding Skills". [I haven't sic'd the errors.]

Flickr details: 1. IMG_2441, 2. IMG_2446

Created with fd’s Flickr Toys Mosaic.