The efficacy of a home-use metabolic device (Lumen) in response to a short-term low and high carbohydrate diet in healthy volunteers.

Abstract

Based on stoichiometric assumptions, and real-time assessment of expired carbon dioxide (%CO2) and flow rate, the Lumen device provides potential for consumers/athletes to monitor metabolic responses to dietary programs outside of laboratory conditions. However, there is a paucity of research exploring device efficacy. This study aimed to evaluate Lumen device response to: i) a high-carbohydrate meal under laboratory conditions, and ii) a short-term low- or high-carbohydrate diet in healthy volunteers. Following institutional ethical approval, 12 healthy volunteers (age: 36±4yrs; body mass: 72.1±3.6 kg; height: 1.71±0.02 m) performed Lumen breath and Douglas bag expired air measures under fasted laboratory conditions and at 30 and 60min after a high-carbohydrate (2g·kg-1) meal, along with capilliarized blood glucose assessment. Data were analyzed using a one-way ANOVA, with ordinary least squares regression used to assess the model between Lumen expired carbon dioxide percentage (L%CO2) and respiratory exchange ratio (RER). In a separate phase, 27 recreationally active adults (age: 42±2yrs; body mass: 71.9±1.9 kg; height: 1.72±0.02 m) completed a 7-day low- (~20% of energy intake [EI]; LOW) or high-carbohydrate diet (~60% of EI; HIGH) in a randomized, cross-over design under free-living conditions. L%CO2 and derived Lumen Index (LI) were recorded daily across morning (fasted and post-breakfast) and evening (pre/post meal, pre-bed) periods. Repeated measures ANOVA were employed for main analyses, with Bonferroni post-hoc assessment applied (P ≤0.05). Following the carbohydrate test-meal, L%CO2 increased from 4.49±0.05% to 4.80±0.06% by 30min, remaining elevated at 4.76±0.06% by 60min post-feeding (P <0.001, ηp2=0.74). Similarly, RER increased by 18.1% from 0.77±0.03 to 0.91±0.02 by 30min post-meal (P =0.002). When considering peak data, regression analysis demonstrated a significant model effect between RER and L%CO2 (F=5.62, P =0.03, R2=0.20). Following main dietary interventions, no significant interactions (diet × day) were found. However, main diet effects were evident across all time-points assessed, highlighting significant differences for both L%CO2 and LI between LOW and HIGH conditions (P <0.003). For L%CO2, this was particularly noted under fasted (4.35±0.07 vs. 4.46±0.06%, P =0.001), pre-evening meal (4.35±0.07 vs. 4.50±0.06%, P <0.001), and pre-bed time-points (4.51±0.08 vs. 4.61±0.06%, P =0.005). Our findings demonstrated that a portable, home-use metabolic device (Lumen) detected significantly increased expired %CO2 in response to a high-carbohydrate meal, and may be useful in tracking mean weekly changes to acute dietary carbohydrate modifications. Additional research is warranted to further determine the practical and clinical efficacy of the Lumen device in applied compared to laboratory settings.