The association of industry sponsorship with outcomes of nutrition studies: a systematic review and meta-analysis of reviews

ID: 

161

Session: 

Poster session 1

Date: 

Monday 24 October 2016 - 10:30 to 11:00

Location: 

All authors in correct order:

Gardon N1
1 University of Sydney, Australia
Presenting author and contact person

Presenting author:

Nicholas Gardon

Contact person:

Abstract text
Background: It is important to know whether sponsorship should be part of the evaluation of bias for nutrition studies that are included in systematic reviews and dietary guidelines.

Objectives: The objective of this review was to determine if the presence of food industry sponsorship or author conflicts of interest (COI) with the food industry are associated with effect sizes, statistical significance of results and/or conclusions that are favorable to the sponsor or differ in their risk of bias.

Methods: We conducted a systematic review and meta-analysis of reviews based on Cochrane methods using Review Manager that investigated samples of primary research studies or reviews that quantitatively compared food industry-sponsored studies with those that had no or other sources of sponsorship, or study authors that had a COI with the food industry with those that had none.

Results: Twelve reviews met the criteria for inclusion. Most reviews (n = 8) assessed the association of industry sponsorship with authors’ conclusions. Our meta-analysis of these eight reviews found that industry-sponsored studies were significantly more likely to have favorable conclusions than non-industry sponsored studies, risk ratio (RR) 1.31 (95% confidence interval (CI) 0.99 to 1.72). Three of these five reviews either had industry sponsorship or authors with COI.

Conclusions: Our findings suggest that industry sponsorship and author COI are associated with conclusions that favor industry sponsors, but there has been little analysis of the influence of sponsorship or author COI on the statistical significance or effect size of research results or risks of bias.