Systematic Literature Review Shows That Appetite Rating does Not Predict Energy Intake

Abstract

Self-report ratings of appetite, particularly visual analogue scales (VAS) are commonly used to measure subjective appetite and to assess modifications thereof following an intervention (1,2) . Subjective rated appetite is a widely employed proxy measure for Actual Energy Intake (AEI) (3) , measurement of which requires greater time and resources. The validity of self-report measures of appetite as surrogate measures of AEI have not been systematically reviewed elsewhere.To identify the corpus of papers assessing both self -reported appetite and AEI in the same trial and to establish whether self-report scales reliably predict AEI.A literature search was undertaken spanning 1999 to 2015 identifying studies recording both VAS ratings and AEI, generally in response to a nutrient or food intervention. Outcomes were pre-defined as there being agreement between self-reported appetite and AEI (link) or no agreement between self-reported appetite scores and AEI (no link). The presence of statistical (direct) comparison between the two methods was also recorded, the type of intervention, subject or patient were also noted. Each paper was scored independently by two authors.462 papers were included in this review. Appetite scores failed to correspond with AEI in 51·3 % of total studies. Only 6 % of studies evaluated directly compared the two measures statistically. A Chi Squared test revealed a significant departure (P < 0·001) from observed and expected frequencies for direct assessment of a link when studies were separated by 'intervention type'. This effect appears to be, in part, due to the use of pharmaceutical interventions and particularly where satiety regulating hormones were manipulated.The very substantive corpus identified in this review allows us to conclude that self-reported appetite ratings of appetite do not reliably predict AEI. Caution should be exercised when deriving conclusions based from self -report appetite data alone in relation to prospective energy intake.

The Differential Role of Smell and Taste For Eating Behavior

“…Although subjective hunger and appetite ratings often predict food intake (De Graaf, Blom, Smeets, Stafleu, & Hendriks, 2004) , they cannot be directly extrapolated, that is the type and amount of food people indicate that they want to eat is not necessarily equivalent to what they eventually consume (Drapeau et al., 2007; Holt et al., 2016; Mattes, 1990; Parker et al., 2004; Stubbs et al., 2000). Indeed, literature shows a distinction between prospective measures of eating (e.g., appetite ratings and preference) and actual food intake and report conflicting results regarding food intake upon odor exposure.…”

The contribution of behavioural science to nutrition: Appetite control

“…Surprisingly, at least for some people, these scales are highly informative and they form part of a satiety toolbox that is a central part of the methodology of appetite control (for review see Blundell et al 2010). This summary statement by experts working in the field remains valid despite the conclusion reached by a recent poorly conducted review (Holt et al 2017).…”

“…This summary statement by experts working in the field remains valid despite the conclusion reached by a recent poorly conducted review (Holt et al . ).…”

Is reduction in appetite beneficial for body weight management in the context of overweight and obesity? Yes, according to the SATIN (Satiety Innovation) study

“…We thereby hypothesise that those reporting acute suppression of appetite when evaluated by VAS do experience reduction in appetite throughout the weight loss maintenance period, helping them to consume less during the weight loss maintenance period; thus, the satiating effect is beneficial for their weight loss maintenance. In general, suppression of appetite assessed by VAS is not necessarily translated into reduced energy intake at the following meals and results are not directly comparable (40,41) . Both measures are prone to self-reporting and social-desirability biases as well as being affected by potential variation in the instruction of the participants (33,42) .…”