How long should washout be to reliably measure hepatic de novo lipogenesis using deuterated water? Exploratory analysis from a randomized crossover trial

Deuterated water (²H₂O) can be used to measure de novo lipogenesis (DNL) with minimal participant burden, but implications of tracer washout on repeated measures in humans are unclear. This study is an exploratory analysis of data from a crossover trial to determine the impact of duration between repeated ²H₂O dosing for sequential assessments of DNL, alongside day-to-day variability in DNL. A total of 22 nonobese adults (11 men and 11 women) completed three laboratory visits in a randomized, crossover design (35 ± 13-day washout). Participants consumed 3 g·kg−1 body water of ²H₂O the evening before laboratory visits. Blood was sampled before ²H₂O ingestion, and the following morning in a fasted state. Deuterium (2H) enrichment of plasma water and very-low-density lipoprotein-triacylglycerol (VLDL-TG)-palmitate were used to determine DNL. Although predosing plasma water 2H enrichments (mole percent) increased across visits from 0.017 ± 0.003% to 0.022 ± 0.008% and 0.027 ± 0.014%, respectively (P < 0.05), postdose enrichments did not display a systematic bias, and nor did 2H enrichments of VLDL-TG-palmitate or measures of DNL [largest mean difference = −1.2%; 95% confidence interval (95% CI) = −12.7% to 9.1%, P = 0.45]. The day-to-day standard deviation and coefficient of variation of fractional hepatic DNL was 2.39% (95% CI = 1.35%–3.42%) and 27% (95% CI = 19%–35%), respectively. Repeated 2H2O dosing does not systematically bias measures of fasting hepatic DNL when using a washout duration of ∼4 wk. We also found no evidence that DNL is biased by washout durations of 3 wk. Therefore, ²H₂O can be used to reliably assess human hepatic DNL in repeated measures designs with at least 3 wk between sequential measures.