Protein Valuation in Food Choice Is Positively Associated with Lean Mass in Older Adults

Abstract

Background: Calorie for calorie, protein is more satiating than carbohydrate or fat. However, it remains unclear whether humans perceive calories derived from these macronutrients equally and whether lean mass is associated with a tendency to "value" protein when dietary decisions are made.

Objectives: This study aimed to determine the test-retest reliability of a novel method for quantifying macronutrient valuations in human volunteers and to determine whether "protein valuation" is associated with a higher fat-free mass index (FFMI) in older adults.

Methods: A 2-alternative, forced-choice task in which 25 foods were compared in 300 trials was undertaken in 2 studies. In study 1, participants (age range 19-71 y, n = 92) attended 2 test sessions, spaced 1 wk apart. In study 2, older adults (age range 40-85 y; n = 91) completed the food-choice task and assessed the test foods for liking, expected satiety, and perceived healthiness. Body composition and habitual protein intake were assessed in both studies. Data were analyzed through the use of individual binomial logistic regressions and multilevel binomial logistic regressions.

Results: In study 1, measures of macronutrient valuation showed excellent test-retest reliability; responses in the forced-choice task were highly correlated (week 1 compared with week 2; protein, r = 0.83, P < 0.001; carbohydrate, r = 0.90, P < 0.001; fat, r = 0.90, P < 0.001). Calorie for calorie, protein and carbohydrate were stronger predictors of choice than fat (P < 0.001). In study 2, protein was a stronger predictor than both carbohydrate (P = 0.039) and fat (P = 0.003), and a positive interaction was observed between protein valuation and FFMI (OR = 1.64; 95% CI: 1.38, 1.95; P < 0.001). This was the case after controlling for age, gender, liking for foods, and habitual protein consumption.

Conclusions: Together, these findings demonstrate that adult humans value calories derived from protein, carbohydrate, and fat differently, and that the tendency to value protein is associated with greater lean mass in older adults.

Keywords: aging; body composition; fat-free mass index; food choice; lean mass; protein valuation; sarcopenia.

FIGURE 1

Relations between ORs obtained for men and women aged 19–71 y in session 1 and 2 in study 1. Panels show associations for (A) fat, (B) carbohydrate, (C) protein, and (D) energy. Note: short dashed lines represent OR = 1 (no significant effect on food choice). Long dashed lines show the correlation (shaded ±1 95% CI) between participant ORs between sessions. Each data point shows 2 ORs, obtained from a single participant (total n = 84) tested on 2 occasions, separated by a 1-wk interval.

FIGURE 2

Box and whisker plots describing ORs for predictors of choice for men and women aged 19–81 y in study 1 and study 2. Separate panels show ORs for macronutrients in study 1 (A) and study 2 (B) and psychological predictors (expected satiety and healthiness) for study 1 (C) and study 2 (D). Note: dashed line indicates no effect on choice. In cases where a 95% CI fails to cross this line, then the associated variable has a nonrandom effect on choice. ORs for study 1 were averaged across test sessions. For all figures, the black triangle indicates mean OR, the black line represents the median, the upper edge of the box represents the 75% quartile, and the lower edge represents the 25% quartile. Black dots represent outliers.