Exploratory analyses on the effect of time since last meal on concentrations of amino acids, lipids, one-carbon metabolites, and vitamins in the Hordaland Health Study

Abstract

Purpose: Dietary intake may have pronounced effects on circulating biomarker concentrations. Therefore, the aim was to provide a descriptive overview of serum metabolite concentrations in relation to time since last meal, focusing on amino acids, lipids, one-carbon metabolites, and biomarkers of vitamin status.

Methods: We used baseline data from the observational community-based Hordaland Health Study, including 2960 participants aged 46-49 years and 2874 participants aged 70-74 years. A single blood draw was taken from each participant, and time since last meal varied. Estimated marginal geometric mean metabolite concentrations were plotted as a function of time since last meal, up to 7 h, adjusted for age, sex, and BMI.

Results: We observed a common pattern for nearly all amino acids and one-carbon metabolites with highest concentrations during the first 3 h after dietary intake. Homocysteine and cysteine were lowest the 1st hour after a meal, while no patterns were observed for glutamate and glutamic acid. The concentrations of phylloquinone and triglycerides were highest 1 h after dietary intake. Thiamine and thiamine monophosphate concentrations were highest, while flavin mononucleotide concentrations were lowest within the first 2 h after a meal. No clear patterns emerged for the other fat-soluble vitamins, blood lipids, or B-vitamin biomarkers.

Conclusion: Our findings suggest that distinguishing between "fasting" and "non-fasting" blood samples may be inadequate, and a more granular approach is warranted. This may have implications for how to account for dietary intake when blood sampling in both clinical and research settings.

Keywords: Biomarkers; Categorization; Epidemiology; Fasting; Metabolites; Postprandial.

Fig. 1

Flowchart illustrating the inclusion and exclusion process of participants in the two age cohorts in the Hordaland Health Study 1997–1999

Fig. 2

The concentration of glucose as a function of time since last meal using cross-sectional data from 5834 participants in the Hordaland Health Study 1997–1999. The solid line indicates estimated marginal geometric means (from a linear regression model adjusted for age cohort, sex, and BMI), while the shaded area represents 95% geometric confidence intervals. The p-value indicated in the figure is for time since last dietary intake. Note, the origin of the y-axis ≠ 0. An overview of the number of observations at each timepoint, and the number of missing observations for each metabolite at each timepoint is provided in Supplementary Table 4

Fig. 3

The concentration of amino acids as a function of time since last meal using cross-sectional data from 5834 participants in the Hordaland Health Study 1997–1999. The solid line indicates estimated marginal geometric means (from a linear regression model adjusted for age cohort, sex, and BMI), while the shaded area represents 95% geometric confidence intervals. The p-value indicated in the figure is for time since last dietary intake. Note, the origin of the y-axis ≠ 0, and the y-axes are scaled to be compatible with the metabolite concentration ranges. An overview of the number of observations at each timepoint, and the number of missing observations for each metabolite at each timepoint is provided in Supplementary Table 4

Fig. 4

The concentration of blood lipids as a function of time since last meal using cross-sectional data from 5834 participants in the Hordaland Health Study 1997–1999. The solid line indicates estimated marginal geometric means (from a linear regression model adjusted for age cohort, sex, and BMI), while the shaded area represents 95% geometric confidence intervals. The p-value indicated in the figure is for time since last dietary intake. Note, the origin of the y-axis ≠ 0, and the y-axes are scaled to be compatible with the metabolite concentration ranges. An overview of the number of observations at each timepoint, and the number of missing observations for each metabolite at each timepoint is provided in Supplementary Table 4. HDL high-density lipoprotein, LDL low-density lipoprotein

Fig. 5

The concentration of one-carbon metabolites as a function of time since last meal using cross-sectional data from 5834 participants in the Hordaland Health Study 1997–1999. The solid line indicates estimated marginal geometric means (from a linear regression model adjusted for age cohort, sex, and BMI), while the shaded area represents 95% geometric confidence intervals. The p-value indicated in the figure is for time since last dietary intake. Note, the origin of the y-axis ≠ 0, and the y-axes are scaled to be compatible with the metabolite concentration ranges. An overview of the number of observations at each timepoint, and the number of missing observations for each metabolite at each timepoint is provided in Supplementary Table 4

Fig. 6

The concentration of lipid-soluble vitamins as a function of time since last meal using cross-sectional data from 5834 participants in the Hordaland Health Study 1997–1999. The solid line indicates estimated marginal geometric means (from a linear regression model adjusted for age cohort, sex, and BMI), while the shaded area represents 95% geometric confidence intervals. The p-value indicated in the figure is for time since last dietary intake. Note, the origin of the y-axis ≠ 0, and the y-axes are scaled to be compatible with the metabolite concentration ranges. An overview of the number of observations at each timepoint, and the number of missing observations for each metabolite at each timepoint is provided in Supplementary Table 4

Fig. 7

The concentration of B-vitamin biomarkers as a function of time since last meal using cross-sectional data from 5834 participants in the Hordaland Health Study 1997–1999. The solid line indicates estimated marginal geometric means (from a linear regression model adjusted for age cohort, sex, and BMI), while the shaded area represents 95% geometric confidence intervals. The p-value indicated in the figure is for time since last dietary intake. Note, the origin of the y-axis ≠ 0, and the y-axes are scaled to be compatible with the metabolite concentration ranges. An overview of the number of observations at each timepoint, and the number of missing observations for each metabolite at each timepoint is provided in Supplementary Table 4