Kaur Asha, Scarborough Peter, Rayner Mike
Centre on Population Approaches for Non-Communicable Disease Prevention, Nuffield Department of Population Health, University of Oxford, Oxford, England.
Int J Behav Nutr Phys Act. 2017 Jul 11;14(1):93. doi: 10.1186/s12966-017-0548-1.
Health-related claims are statements regarding the nutritional content of a food (nutrition claims) and/or indicate that a relationship exists between a food and a health outcome (health claims). Their impact on food purchasing or consumption decisions is unclear. This systematic review measured the effect of health-related claims, on pre-packaged foods in retail settings, on adult purchasing decisions (real and perceived).
In September 2016, we searched MEDLINE, EMBASE, PsychINFO, CAB abstracts, Business Source Complete, and Web of Science/Science Citation Index & Social Science Citation Index for articles in English published in peer-review journals. Studies were included if they were controlled experiments where the experimental group(s) included a health-related claim and the control group involved an identical product without a health-related claim. Included studies measured (at an individual or population level); actual or intended choice, purchases, and/or consumption. The primary outcome was product choices and purchases, the secondary outcome was food consumption and preference. Results were standardised through calculating odds ratios and 95% confidence intervals (CI) for the likelihood of choosing a product when a health-related claim was present. Results were combined in a random-effects meta-analysis.
Thirty-one papers were identified, 17 of which were included for meta-analyses. Most studies were conducted in Europe (n = 17) and the USA (n = 7). Identified studies were choice experiments that measured the likelihood of a product being chosen when a claim was present compared to when a claim was not present, (n = 16), 15 studies were experiments that measured either; intent-rating scale outcomes (n = 8), consumption (n = 6), a combination of the two (n = 1), or purchase data (n = 1). Overall, 20 studies found that claims increase purchasing and/or consumption, eight studies had mixed results, and two studies found consumption/purchasing reductions. The meta-analyses of 17 studies found that health-related claims increase consumption and/or purchasing (OR 1.75, CI 1.60-1.91).
Health-related claims have a substantial effect on dietary choices. However, this finding is based on research mostly conducted in artificial settings. Findings from natural experiments have yielded smaller effects. Further research is needed to assess effects of claims in real-world settings.
PROSPERO systematic review registration number: CRD42016044042 .
与健康相关的声明是关于食品营养成分的陈述（营养声明）和/或表明食品与健康结果之间存在某种关系（健康声明）。它们对食品购买或消费决策的影响尚不清楚。本系统评价衡量了零售环境中预包装食品上与健康相关的声明对成年人购买决策（实际和感知）的影响。
2016年9月，我们检索了MEDLINE、EMBASE、PsychINFO、CAB文摘、商业资源全文数据库以及科学网/科学引文索引和社会科学引文索引，以查找同行评审期刊上发表的英文文章。如果研究是对照实验，实验组包含与健康相关的声明，对照组涉及不含与健康相关声明的相同产品，则纳入研究。纳入的研究在个体或群体层面上测量；实际或预期的选择、购买和/或消费。主要结局是产品选择和购买，次要结局是食品消费和偏好。通过计算出现与健康相关声明时选择产品的可能性的比值比和95%置信区间（CI）对结果进行标准化。结果合并进行随机效应荟萃分析。
共识别出31篇论文，其中17篇纳入荟萃分析。大多数研究在欧洲（n = 17）和美国（n = 7）进行。已识别的研究是选择实验，测量存在声明时与不存在声明时产品被选择的可能性，（n = 16），15项研究是测量以下方面的实验：意向评分量表结果（n = 8）、消费（n = 6）、两者结合（n = 1）或购买数据（n = 1）。总体而言，20项研究发现声明会增加购买和/或消费，8项研究结果不一，2项研究发现消费/购买减少。对17项研究的荟萃分析发现，与健康相关的声明会增加消费和/或购买（比值比1.75，CI 1.60 - 1.91）。
与健康相关的声明对饮食选择有重大影响。然而，这一发现基于大多在人工环境中进行的研究。自然实验的结果影响较小。需要进一步研究以评估声明在现实环境中的影响。
PROSPERO系统评价注册号：CRD42016044042 。
