Preliminary Evidence Supports that Long-Term Consumption of Higher-Protein Breakfast Promotes Higher Expression of Select miRNA Associated with Cardiometabolic Health in Adolescents

Abstract

Background

Increased dietary protein at breakfast promotes cardiometabolic health; however, whether these improvements occur at the molecular level is unknown.

Objectives

The objective was to examine whether long-term consumption of breakfast, varying in protein quantity, alters the expression of circulating microRNAs (miRNAs) associated with cardiometabolic health in “breakfast-skipping” adolescents.

Methods

Thirty adolescents (age: 19 ± 1 y; body mass index: 25.4 ± 3 kg/m²) completed a 6-mo tightly controlled breakfast trial in which participants consumed 350 kcal normal-protein (NP, 10 g protein) or higher-protein (HP, 30 g protein) breakfasts or continued to BS for 6 mo. Fasting blood samples were collected at baseline (PRE) and 6 mo (POST) for assessment of 12 a priori circulating plasma miRNA expression levels (real-time quantitative polymerase chain reaction), glucose, insulin, IL-6, and C-reactive protein.

Results

No main effects of group were observed for any miRNAs; however, a time-by-group interaction was detected for the expression of miR-126-3p (P = 0.05). HP breakfast tended to increase miR-126-3p expression throughout the study (POST-PRE, P = 0.09) leading to greater expression at POST compared with BS (P = 0.03), whereas NP breakfast did not. Additionally, several miRNAs predicted fasting concentrations of IL-6: miR-320a-3p, -146a-5p, -150-5p, -423-5p, -122-5p, glucose: miR-24-3p, -126-3p; insulin: miR-24-3p, -126-3p, -15b-5p; insulin sensitivity: miR-24-3p, -126-3p, -199a-5p, -15b-5p; and β-cell function: miR-15b-5p (R² between 0.2 and 0.39; P < 0.05) from PRE and POST samples across groups.

Conclusions

These data support the daily consumption of a HP breakfast to promote cardiometabolic health, potentially through changes in miRNA expression, in a sensitive life-stage where early intervention strategies are critical to reduce the risk of adult-onset chronic disease.

Trial registration number

NCT03146442.

Introduction

The prevalence of breakfast skipping(BS) in children and adolescents has continued to rise concurrently with the increase in overweight/obesity in this population, with overweight/obesity identified as a growing health concern that impacts >390 million lives worldwide [1,2]. In addition to an increased risk of unhealthy weight gain and obesity, BS has a detrimental effect on cardiometabolic outcomes including blood pressure, blood lipids, and insulin resistance in young people [3,4]. Conversely, the consumption of breakfast has been shown to have a protective effect; however, the composition of the breakfast plays a more significant role than previously thought [5]. Data from our laboratory demonstrate that the consumption of breakfast, particularly higher in dietary protein, decreases cardiometabolic risk through improvements in diet quality, weight management, and glycemic control in adolescent populations with overweight/obesity [[6], [7], [8]]. However, the underlying molecular mechanism contributing to these positive health outcomes is not fully understood.

MicroRNAs (miRNAs), a family of non-coding RNAs working as post-transcriptional modifiers and emerging as key influencers of chronic disease, may be involved in mediating high protein breakfast’s impact on cardiometabolic risk factors. Emerging evidence demonstrates that miRNAs can regulate gene expression in distal cells as it is disseminated into circulation [9,10]. Metabolic dysregulation as seen in obesity, type 2 diabetes (T2D), and cardiovascular disease (CVD) is known to influence the expression of miRNA [[11], [12], [13]]. In addition, recent work in adults with chronic disease has also shown that dietary factors, including an increase in dietary protein and/or timing of consumption, may also influence miRNA expression [14].

For example, acute high-protein breakfast consumption in individuals with metabolic dysfunction (that is, insulin resistance) has shown to differentially influence expression levels of circulating miRNAs when compared with healthy controls [15]. However, less is known about the habitual consumption of high protein breakfast on miRNA expression including miRNAs associated with cardiometabolic disease risk. Elucidating this connection is of importance given the evidence surrounding the prevalence of BS in adolescents and the negative impact skipping breakfast inflicts on cardiometabolic health.

The purpose of this exploratory, preliminary study was to evaluate the effects of breakfast consumption, varying in protein quantity, on the expression of circulating miRNAs associated with cardiometabolic disease risk in adolescents who habitually BS. As a secondary aim, we sought to explore whether miRNA expression predicts changes in cardiometabolic disease risk factors.

Methods

Experimental design

This secondary analysis was part of a larger study testing whether the long-term consumption of breakfast improves weight management in habitual BS adolescents with overweight (clinical trials number: NCT03146442). Using a randomized controlled trial design, 30 “breakfast-skipping” adolescents were provided with 350 kcal normal-protein (10 g) (NP; n = 10) or higher-protein (30 g) (HP; n = 10) breakfast meals or continued to BS (n = 10) for 6 mo. Fasting blood samples were collected at baseline (PRE) and 6 mo (POST) for the analyses of select circulating miRNAs and markers of cardiometabolic health.

Study participants

A total of 159 adolescents were recruited from the West Lafayette, IN and Austin, TX area to participate in the larger study and met the following criteria: 1) age: 13–21 y; 2) 50th–98th percentile for BMI for age; BMI: 18–34 kg/m²; 3) skips breakfast ≥4 times/wk; 4) never smoked or used other tobacco products; 5) willing to consume study breakfasts; 6) no clinical diagnosis of an eating disorder; 7) no metabolic, hormonal, and/or neural conditions/disease that influence metabolism or appetite; 8) not currently or previously on a weight loss or other special diet (in the past 6 mo); 9) has not gained/lost ≥4.5 kg over the past 6 mo; 10) not taking medication that would directly influence appetite (weight-loss drugs or antidepressant, steroid, or thyroid medication, unless dosage has been stable for ≥6 mo); and 11) has normal cognitive restraint (assessed via Three Factor Eating Habits Questionnaire) [16]. Of these, 159 met initial screening criteria, signed the study assent/consent, and began the study. A total of 106 completed all study procedures. For this secondary analysis, 30 participants completed the PRE/POST collection of fasting blood samples. Their demographic data are shown in Table 1 and baseline variables did not differ between groups. In short, the participants were on older adolescents (mean ± SD; age: 19 ± 1 y; BMI: 25.4 ± 3 kg/m²) who habitually skipped breakfast 6 ± 1 times/wk. All study participants were informed of the purpose, procedures, and potential risks and signed the consent forms. The study was approved by the Institutional Review Boards at Purdue University and The University of Texas at Austin. Each of the participants received a stipend for completing study procedures.

TABLE 1.

Baseline characteristics of the participants who completed the study.

Demographic information	Breakfast skipping (BS) Mean ± SD	Normal-protein (NP) Mean ± SD	Higher-protein (HP) Mean ± SD	Group effect1 P value
Sample size (n)	10	10	10	N/A
Males:Females (n)	5:5	5:5	5:5	N/A
Age (y)	19.6 ± 1.17	19.1 ± 1.10	19.7 ± 1.16	0.465
Weight (kg)	76.0 ± 14.34	67.9 ± 17.41	74.5 ± 11.25	0.487
Height (cm)	168.4 ± 9.79	165 ± 11.39	172 ± 8.99	0.303
BMI (kg/m2)	26.7 ± 3.13	24.6 ± 3.69	25 ± 3.46	0.374
Habitual breakfast skipping (n/wk)	5.9 ± 1.54	6.2 ± 1.16	6.2 ± 1.23	0.904

¹Mean comparison via 1-way ANOVA.

Breakfast patterns

Participants were randomly assigned to either continue to BS (control) or consume 1 of 2 breakfast patterns for 6 mo. The BS group skipped breakfast every morning by consuming nothing but water until 10:00 am. Both breakfast groups were provided with breakfasts to consume every morning between 6:00 and 9:00 am throughout the intervention period. The energy and nutrition characteristics of each breakfast type are shown in Table 2. The energy content of the breakfast meals in the NP and HP groups were 350 kcals and standardized to provide ∼18% kcal/d for adolescents ages 13–21 y. The breakfast meals in both groups were matched for fat content (∼25% of energy from fat) but varied in protein and carbohydrates. The NP breakfasts contained ∼12% of energy from protein and ∼63% of energy from carbohydrates, whereas the HP breakfasts contained ∼35% of energy from protein and ∼40% of energy from carbohydrates. The NP and HP breakfast meals provided to the participants contained similar types of meals (that is, burritos, quesadillas, sandwiches, breakfast bowls, etc.) that included combinations of animal-based proteins (egg, dairy, and beef). Participants were provided with an 8-d breakfast meal plan to include variety and strengthen adherence. Adherence of >80% to the intervention was met by all participants included in this exploratory study.

TABLE 2.

Dietary characteristics of the breakfast meals.

Daily intake	Breakfastskipping(BS)	Normal-protein(NP) Mean ± SD	Higher-protein(HP) Mean ± SD
Energy (kcal/d)	0	347 ± 8.9	347 ± 10.8
Protein (g/d)	0	10.8 ± 1.0	30.4 ± 0.9
Carbohydrate (g/d)	0	54.6 ± 1.3	35.1 ± 1.2
Fat (g/d)	0	9.8 ± 0.7	10 ± 0.5
Total sugar (g/d)	0	21 ± 6.4	12.1 ± 4.2
Fiber (g/d)	0	3.8 ± 0.9	4.8 ± 1.5
Saturated fat (g/d)	0	4 ± 1.1	4.1 ± 0.9
Added sugar (g/d)	0	11 ± 7.5	4.8 ± 4.2

All of the study breakfasts were assembled, cooked, and package sealed in a metabolic testing kitchen with each of the ingredients weighed to the nearest tenth of a gram. Each participant, regardless of treatment group, had a one-on-one meeting with the study nutrition staff each week to discuss any questions, concerns, or issues with following their respective treatment. Additionally, participants completed daily breakfast check-off logs throughout the intervention period. For the BS group, the logs included a statement reminding them to eat or drink nothing (but water) before 10:00 am. For the breakfast consuming groups, the logs included a list of food items and quantities required to be consumed at breakfast each day. They were instructed to check off all foods that were consumed along with the time of day.

Fasting blood sampling

After a 12-h overnight fast, blood samples (4 mL/sample) were collected. The samples were collected in test tubes containing EDTA, centrifuged at 4°C for 10 min and the plasma subfraction was then separated and stored at −80°C for future analysis.

Plasma metabolic and inflammatory markers

The plasma fasting markers assessed were as follows: insulin (Ultrasensitive Insulin ELISA; ALPCO), C-reactive protein (CRP ELISA; ALPCO), IL-6 (Human IL-6 High Sensitivity ELISA; Invitrogen), and glucose (colorimetric assay; Cayman Chemical).

miRNA isolation and cDNA synthesis

Total RNA was isolated from 200 μL of plasma (miRNeasy plasma/serum Advanced Kit) (Cat. No. 217204; Qiagen). A spike-in template for miRNA UniSp2 was added to each sample before extraction to serve as a control for RNA isolation, given the limited amount of RNA in plasma (Cat. No. 339390; Qiagen). There were no differences detected in UniSp2 between treatments or time point (C_T values mean ± SD; BS PRE: 16.63 ± 0.75; BS POST: 16.18 ± 0.81; NP PRE: 16.53 ± 0.56; NP POST: 15.73 ± 0.60; HP PRE: 16.75 ± 0.86; HP POST: 15.92 ± 0.63). Using a fixed amount of RNA (2 μL), cDNA was synthesized by reverse transcription using the miRCURY LNA RT Kit (Cat. No. 339340; Qiagen) in a C1000 Touch Thermal Cycler (Bio-Rad).

Quantitative polymerase chain reaction

miRCURY LNA miRNA PCR Assays (Cat. No. 339306) were selected to assist in quantifying the following miRNAs of interest that have been linked to the progression and/or severity of obesity, T2D, and CVD [[11], [12], [13]]: miR-21-5p, miR-320a-3p, miR-146a-5p, miR-150-5p, miR-423-5p, miR-15b-5p, miR-122-5p, miR-223-3p, miR-199a-5p, miR-214-3p, miR-24-3p, miR-126-3p. Quantification via RT-qPCR was performed using miRCURY LNA SYBR Green PCR Kit (Cat. No. 339347) and the CFX96 Touch Real-Time PCR Detection System (Bio-Rad). RNase free water was used as a negative control, and each sample was measured in duplicate. miRNA met inclusion criteria if the average C_T ≤ 35. Hemolysis was assessed by the ratio generated by the difference in ΔC_T values of miR-16-5p and miR-23a-3p [17]. This analysis revealed that no sample had a miR ratio ≥ 7 and there was no difference in this marker between treatments or timepoint (C_T values mean ± SD; BS PRE: 3.99 ± 1.44; BS POST: 4.17 ± 1.84; NP PRE: 3.46 ± 1.61; NP POST: 3.64 ± 2.57; HP PRE: 4.49 ± 0.66; HP POST: 5.29 ± 1.20). Expression levels of miRNAs queried were normalized to the geometric mean of UniSp2 (external control) and SNORD38B and SNORD49A (internal controls) (C_T values mean ± SD, CV; BS PRE: 26.06 ± 0.85, 3.3%; BS POST: 25.61 ± 0.71, 2.8%; NP PRE: 25.65 ± 1.14, 4.4%; NP POST: 25.17 ± 0.81; 3.2%; HP PRE: 26.32 ± 0.78, 2.9%; HP POST: 25.64 ± 0.59; 2.3%). The ΔΔC_T method [18] was used for quantification of miRNA and relative expression is reported as 2^-ΔΔCT. Fold-change data were expressed relative to mean PRE values within a treatment.

Data and statistical analyses

Power analysis was performed using data from our laboratory assessing the differential expression of circulating miR-15b-5p after dietary interventions comparing the impact of protein quality (animal compared with plant) [19]. An effect size of 1.39 rendered that a sample size of n = 8/group would provide 80% power to detect meaningful differences in miRNA expression.

HOMA was utilized to evaluate metrics of IR, beta cell function (%B), and insulin sensitivity (%S) via the HOMA2 Calculator (https://www.dtu.ox.ac.uk/homacalculator/) (Diabetes Trials Unit, University of Oxford, Radcliffe Department of Medicine).

Summary statistics were computed for all data. No missing data occurred in this study. The Shapiro-Wilk test was used to assess for normal distribution of the data before statistical testing. All miRNA data were log transformed (log₂) for statistical analyses. One-way Analysis of Variance (ANOVA) was applied to identify group differences at baseline. Age, BMI, % body fat, gender at birth, ethnicity, and race were examined via Pearson correlational analyses to assess whether these potential confounders were associated with study outcomes. Those that were associated, were covaried in the model. Linear mixed models were performed to examine the main effects of time (PRE, POST), group (BS, NP, HP), and time-by-group interactions for all outcomes. When main effects and/or interactions were detected, post hoc pairwise comparisons were performed using Bonferroni adjustment. For these analyses, the criterion for statistical significance was designated as P ≤ 0.05. In addition, multivariate backward regression analyses were performed on the PRE and POST fold-change values across all groups to assess whether the expression of select miRNAs served as predictors of cardiometabolic disease risk factors measured. Specifically, the analysis includes the iterative removal of the least statistically significant predictor variable with a criterion set at a probability of F-to-remove ≥0.100. These procedures continue, sequentially, until only the statistically significant predictor variables remain. All statistical analyses were performed in IBM SPSS Statistics (Version: 29.0.0.0). The criterion for statistical significance was designated as P ≤ 0.05.

Results

All 12 a priori circulating miRNA were detected at each time point within all groups (Figure 1). Regardless of group, a main effect of time was observed for miR-423-5p and miR-24-3p (both, P ≤ 0.05) with both decreasing over time (POST-PRE, Figure 1). No other time effects and no group effects were observed (Figure 1). However, a time-by-group interaction was detected for expression of miR-126-3p (P = 0.05). Specifically, the HP breakfast group tended to increase miR-126-3p expression throughout the study (POST-PRE, P = 0.09) leading to greater expression at POST compared with BS (P = 0.03), whereas the NP breakfast group did not (Figure 2).

FIGURE 1.

Dot plots of circulating miRNA for comparison between time (PRE, POST) and treatment (Breakfast Skipping (BS, □), Normal Protein (NP, ◯), Higher Protein (HP, △) via linear mixed model. Mean ± SD. ∗Different from PRE, P ≤ 0.05. The following covariates were included in the models: %Fat, gender at birth, BMI, and ethnicity.

FIGURE 2.

Dot plots of circulating miRNA comparing time (PRE, POST) and treatment (Breakfast Skipping (BS, □), Normal Protein (NP, ◯), Higher Protein (HP, △) via linear mixed model with main effects of time-by-treatment interactions. Post hoc comparisons; different letters illustrate significance, P ≤ 0.05. Mean ± SD. The following covariates were included in the models: %Fat and gender at birth.

Several miRNAs were found to moderately predict fasting concentrations of IL-6 (miR-320a-3p, -146a-5p, -150-5p, -423-5p, 122-5p), glucose (miR-24-3p, -126-3p), and insulin (miR-24-3p, -126-3p, -15b-5p) as well as markers of insulin sensitivity (miR-24-3p, -126-3p, -199a-5p, -15b-5p) and β-cell function (miR-15b-5p) (Table 3). When applying linear background regression analyses with best model fit, 6 miRNAs explained 38.5% of the variance of fasting plasma IL-6 (P = 0.003). Regarding glucose regulation, 8 miRNAs explained 21.9% of the variance of fasting glucose (P = 0.008), whereas 7 miRNAs explained 22.4% of the variance of fasting insulin (P = 0.015). Lastly, 5 miRNAs explained 21.3% of the variance of HOMA-IR (P = 0.01), and 21.3% of the variance of HOMA-%S (P = 0.010).

TABLE 3.

Multivariate background regression analyses assessing miRNA as predictors of cardiometabolic markers measured in plasma.

	miRNA	β	t	P
IL-6	320a-3p	1.767	3.530	0.002∗
	146a-5p	−1.272	−3.487	0.002∗
	150-5p	−0.388	−2.342	0.027∗
	423-5p	−1.103	−2.990	0.006∗
	126-3p	0.609	1.695	0.102
	122-5p	−0.512	−3.400	0.002∗
Adjusted R2 = 0.385 F (6,27) = 4.438

	miRNA	β	t	P
CRP	126-3p	1.568	3.122	0.003∗
	214-3p	−0.277	−1.793	0.079
	199a-5p	−0.674	−2.277	0.027∗
	15b-5p	−0.968	−2.898	0.006∗
Adjusted R2 = 0.118 F (4,49) = 2.776

	miRNA	β	t	P
Glucose	21-5p	0.963	1.929	0.059
	320a-3p	0.686	1.713	0.093
	146a-5p	−0.819	−1.594	0.117
	423-5p	−0.533	−1.540	0.130
	24-3p	−0.404	−2.460	0.017∗
	126-3p	1.223	2.057	0.045∗
	199a-5p	−0.679	−1.916	0.061
	15b-5p	−0.670	−1.716	0.093
Adjusted R2 = 0.219 F (8,49) = 3.004

	miRNA	β	t	P
Insulin	21-5p	0.786	1.407	0.167
	320a-3p	0.377	1.168	0.250
	146a-5p	−0.774	−1.453	0.154
	24-3p	−0.701	−3.309	0.002∗
	126-3p	1.859	2.898	0.006∗
	199a-5p	−0.614	−1.822	0.076
	15b-5p	−1.450	−3.369	0.002∗
Adjusted R2 = 0.224 F (7,39) = 2.901

	miRNA	β	t	P
HOMA-IR	320a-3p	0.426	1.859	0.070
	24-3p	−0.599	−2.966	0.005∗
	126-3p	1.728	3.032	0.004∗
	199a-5p	−0.751	−2.633	0.012∗
	15b-5p	−1.178	−3.360	0.002∗
Adjusted R2 = 0.213 F (5,41) = 3.484

	miRNA	β	t	P
HOMA-B	126-3p	0.337	1.193	0.239
	15b-5p	−0.607	−2.149	0.037∗
Adjusted R2 = 0.088 F (2,44) = 3.211

	miRNA	β	t	P
HOMA-S	320a-3p	−0.426	−1.859	0.070
	24-3p	0.599	2.966	0.005∗
	126-3p	−1.728	−3.032	0.004∗
	199a-5p	0.751	2.633	0.012∗
	15b-5p	1.178	3.360	0.002∗
Adjusted R2 = 0.213 F (5,41) = 3.484

^∗P ≤ 0.05.

Discussion

In this exploratory, preliminary study, we examined the impact of consuming breakfast meals, varying in protein quantity, for 6 mo on the expression of select circulating miRNAs associated with cardiometabolic risk factors in “breakfast-skipping” adolescents. The long-term consumption of HP breakfasts altered the expression of miR-126-3p compared with BS, whereas the NP breakfasts did not. A number of miRNAs were found to predict cardiometabolic outcomes that supports their suggested roles in biological processes that contribute to obesity, T2D, and CVD. Although preliminary, these data suggest that the daily consumption of breakfast, particularly higher in protein, promotes cardiometabolic health, potentially through changes in miRNA expression known to regulate glucose metabolism and cardiovascular health. This study offers novel insight into the effects of dietary factors on miRNAs that were shown to be predictive of cardiometabolic risk factors in a population receptive of early intervention strategies to reduce the risk of adult-onset chronic disease.

Large cohort studies highlight the role of breakfast consumption for healthy weight management, glucose control, and cardiometabolic risk reduction [20]. However, breakfast quality is also a key factor of consideration to align with recommendations to incorporate protein for weight management and cardiometabolic health benefits [5]. Although the benefits of HP breakfasts are well documented, the molecular mechanisms by which protein at breakfast improves cardiometabolic health remain unknown.

To our knowledge, only one acute intervention study exists evaluating the impact of a HP breakfast on circulating miRNA expression. As shown in Ramzan et. al. [15], adult women with/without overweight and insulin resistance consumed a 600-kcal HP breakfast containing 27% of energy as protein (∼40 g protein) and circulating miRNA expression was assessed pre-breakfast and 2- and 4-h post-breakfast. In the healthy adult women, miR-15a-5p and miR-17-5p expression was reduced by 50% at 4-h post-breakfast compared with pre-breakfast, whereas this reduction did not occur in those overweight and insulin resistance. These specific miRNAs are involved in glucose and fatty acid metabolism such that reduction in their expression suggest metabolic flexibility after meal consumption.

In this study we found that miR-126-3p expression level was greater in the HP breakfast group compared with skipping breakfast. Vital to vascular health, miR-126-3p is highly enriched in endothelial cells and plays a pivotal role in angiogenesis [21]. In humans, miR-126-3p was found to be lower in adults with impaired fasting glucose and those diagnosed with T2D, and increases as a result of pharmaceutical, dietary, and/or exercise interventions [22]. Supporting evidence strengthens the influence of dietary intervention on miR-126-3p as expression levels increase as a result of weight loss [23]. In agreement with previous findings, this study found a higher level of expression of miR-126-3p as a result of dietary intervention, specifically a HP from animal sources breakfast. This result may suggest a potential improvement in indices of cardiovascular health, but further research is warranted.

There are a number of proposed pathways by which miR-126-3p modulate anti-inflammatory responses through inflammatory regulators, including nuclear factor kappa B and macrophage activation [24]. Furthermore, pathway analyses performed for this miRNA have shown that miR-126-3p is involved in the fatty acid β-oxidation pathway by targeting suppressors of AMP-activated protein kinase [25]. Although speculative, it could be interpreted that this miRNA modulates similar pathways to facilitate the production of specific metabolites that ameliorate metabolic disease risk factors contributing to conditions such as T2D and CVD [25]. However, future studies, such as metabolomics analyses, are warranted to explore this relationship and the influence of varying dietary strategies, including dietary protein consumption, on the changes in concentration of metabolites implicated in disease risk.

We identified several miRNAs that consistently appeared in models used to predict changes in cardiometabolic markers with respect to miRNA expression. Interestingly, we found that miR-126-3p predicted changes in CRP, glucose, insulin, HOMA-IR, and HOMA-%S such that for every unit increase of miR-126-3p, each of these markers would increase. Given the findings presented earlier concerning the higher expression of miR-126-3p were in agreement with previous studies, further exploration is warranted to understanding the impact of miR-126-3p on markers of cardiometabolic health. Furthermore, miR-24-3p predicted glucose, insulin, HOMA-IR, and HOMA%S responses; miR-199a-5p predicted CRP, HOMA-IR, and HOMA-%S responses; and miR-15b-5p predicted CRP, insulin, HOMA-IR, and HOMA-%S. Specifically, for each unit increase in the expression of miRNA, CRP, insulin, HOMA-IR, and HOMA-%B would decrease, but for every unit increase in miRNA expression, HOMA-%S would increase. A recent study from our laboratory also observed inverse associations between miR-15b-5p expression and indices of insulin resistance, beta cell function, and positive associations with insulin sensitivity [19]. These data illustrate that this subset of miRNA (miR-126-3p, miR-199a-5p, miR-24-3p, and miR-15b-5p) may serve as moderate predictors for cardiometabolic disease risk factors.

Limitations and considerations

Although the study intervention, duration, and participant adherence should be considered strengths of this study, several limitations also exist. First, it is suggested to consider that hemolysis may influence expression levels of specific miRNAs. In the protocol outlined by Shah et al. [17], samples are at low risk of hemolysis if the miR ratio is <5. In this study we found that none of the samples included in the analysis were ≥7, which suggests that these samples were at moderate to low risk. However, there were no differences in miR ratio between timepoint or treatment, and therefore, it is unlikely that our data were influenced by hemolysis. Within this study, the sample size was fairly small (n = 10/group), and although this was an exploratory study that met sample size estimates, we may have been underpowered to detect time and group differences in miRNA expression. In the current investigations, we assessed total circulating miRNA, which consists of miRNA bound in extracellular vesiculas, to include exosomes, lipoproteins, with RNA binding protein (Argonaute2), and apoptotic bodies that have been shed by tissues. For future studies, additional insight may be obtained by isolating a specific source of miRNA, particularly extracellular vesiculas that are actively released from cells and participate in cell-to-cell communication [26]. Lastly, the interpretations of these data are confined to a healthy, albeit overweight, adolescent population and similar outcomes may not be observed in a similar and/or older population with chronic disease. Additional research is warranted to understand the broader impact of an HP breakfast in a population with greater risk for developing or afflicted with cardiometabolic disease. Future efforts should also survey a greater spectrum of miRNAs and affiliated gene targets associated with integral pathways contributing to metabolic dysregulation.

The results from this study provide insight into the potential beneficial influence of animal-sourced HP breakfast meals on miRNAs associated with and suggested predictors of cardiometabolic disease risk factors. Collectively, these data provide new evidence in evaluating the influence of habitual BS compared with consumption at the molecular level specifically in post-transcriptional modifiers linked with cardiometabolic disease risk factors.

Author contributions

The authors’ responsibilities were as follows – KAP, JAG, HJL: developed the research question; KAP, LMM: developed the associated methodology; KAP: performed the statistical analyses and interpretation of the data; KAP: wrote the first draft of the manuscript; KAP, HJL: had primary responsibility for the final content; and KAP, LMM, JAG, HJL: read and approved the final manuscript.

Funding

National Institutes of Health (NIDDK), R01 DK107390 01. The National Institutes of Health supplied the funds to complete the study but was not involved in the design, implementation, analyses or interpretation of the data.

Data availability

Data described within the manuscript are available upon request from the corresponding author (HJL).

Conflict of interest

KAP, LMM, and JAG have no conflicts of interest to report. The opinions or assertions contained herein are the private views of the authors and are not to be construed as official or as reflecting the views of the Army or the Department of Defense. Any citations of commercial organizations and trade names in this report do not constitute an official Department of the Army endorsement or approval of the products or services of these organizations. HJL is an Editor for The Journal of Nutrition and played no role in the Journal's evaluation of the manuscript.