Substituting Refined Sugars With Maple Syrup Decreases Key Cardiometabolic Risk Factors in Individuals With Mild Metabolic Alterations: A Randomized, Double-Blind, Controlled Crossover Trial

Abstract

Background

Maple syrup, a minimally transformed sweetener rich in polyphenols, can exert a action and improve metabolic parameters in animal models. However, no randomized clinical trial has investigated this.

Objectives

This study aims to determine whether replacing refined sugars with an equivalent quantity of maple syrup could decrease key cardiometabolic risk factors in individuals with mild metabolic alterations.

Methods

In a randomized, double-blind, controlled crossover trial with 42 overweight adults with mild cardiometabolic alterations, participants were instructed to substitute 5% of their total caloric intake from added sugars with either maple syrup or an artificially flavored sucrose syrup for 8 wk. The primary outcome included changes in glucose homeostasis, whereas secondary outcomes were changes in other cardiometabolic risk factors such as blood pressure, anthropometric indices, and blood lipid profiles. Exploratory outcomes involved analyzing changes in gut microbiota composition.

Results

Replacing refined sugars with maple syrup over 8 wk decreased the glucose area under the curve when compared with substituting refined sugars with sucrose syrup, as determined during the oral glucose tolerance test, leading to a significant difference between the intervention arms (−50.59 ± 201.92 compared with 29.93 ± 154.90; P < 0.047). Substituting refined sugar with maple syrup also significantly decreased android fat mass (−7.83 ± 175.05 g compared with 67.61 ± 206.71 g; P = 0.02) and systolic blood pressure (−2.72 ± 8.73 mm Hg compared with 0.87 ± 8.99 mm Hg; P = 0.03). No changes in the blood lipid profile were observed. As an exploratory outcome, we further observed that substituting refined sugars with maple syrup promoted selective taxonomic changes in the gut microbiota such as a significant reduction in the abundance of Klebsiella species and decreased microbial functions associated with bacterial-induced cytokine response, when compared with substitution with sucrose syrup.

Conclusions

Substituting refined sugars with maple syrup in individuals with mild metabolic alterations result in a significantly greater reduction of key cardiometabolic risk factors compared with substitution with sucrose syrup, in association with specific changes in gut microbiota. The role of the gut microbiota in these effects remains to be further explored.

This trial was registered at clinicaltrials.gov as NCT04117802.

Introduction

Metabolic syndrome is a set of risk factors including abdominal obesity, glucose dysmetabolism, dyslipidemia, and hypertension, which increase risk of type 2 diabetes (T2D) and cardiovascular diseases [1,2]. The prevalence of overweight and obesity, which are defined by a BMI (kg/m²) greater than 25 and 30, respectively, has nearly tripled over the past 1 decades [3]. A robust body of literature has linked added sugars, especially in the form of sugar-sweetened beverages, to weight gain and risk of T2D, and cardiovascular diseases [4]. To limit its consequences on metabolic health, the WHO, the American Heart Association, and the Institute of Medicine have developed recommendations for an upper daily limit of added and free sugars. The WHO recommends reducing the intake of free sugars, which include monosaccharides and disaccharides (e.g., sucrose) added to foods and beverages, and sugars naturally present in honey, syrups, fruit juices, and fruit juice concentrate, to less than 10% of the total energy intake [5].

However, there is growing scientific evidence that not all sugars have the same metabolic effects. It has been previously reported that a single dose of maple syrup induced a lower glycemic response than an equivalent amount of refined sugar in rats [[6], [7], [8]]. We recently reported that diet-induced obese rats receiving maple syrup and other natural sweeteners daily for 8 weeks had decreased insulin resistance, improved glucose metabolism, and reduced liver inflammation compared with rats receiving isocaloric amounts of sucrose [9]. Similar metabolic benefits effects were observed in diet-induced obese mice in which sucrose was substituted with maple syrup, which was associated with reduced carbohydrate digestion and a shift in gut microbiota composition [10].

Maple syrup is a minimally transformed product obtained by thermal evaporation of maple sap until a sugar content of 66°–67° Brix is obtained, yielding a thick, sweet syrup [11]. Sucrose is the primary carbohydrate present in maple syrup, with minor amounts of glucose and polysaccharides [12]. Beyond its sugar composition, maple syrup also contains organic acids, micronutrients, and polyphenols, such as lignans, which are associated with improved metabolic health [[12], [13], [14], [15]]. The process of sap evaporation concentrates these compounds and even creates new ones, such as quebecol [16]. In addition to its phytochemical composition, maple syrup also contains inulin, a prebiotic fiber [13,17]. Polyphenols and inulin are compounds specifically degraded by the gut microbiota, which in turn, modifies its composition [18]. It is now broadly recognized that the gut microbiota plays a central role in host metabolic health [19,20].

To date, no randomized clinical trial has investigated the impact of replacing refined sugars with maple syrup on cardiometabolic risk factors and the gut microbiota. Thus, this randomized, controlled, double-blind, crossover trial aimed to assess the impact of replacing refined sugars with an equivalent quantity of maple syrup representing 5% of the daily caloric intake on glucose homeostasis (primary outcome); cardiometabolic risk factors, including blood pressure, anthropometric measures, and blood lipid profiles (secondary outcome); and changes in gut microbiota (exploratory outcome) in participants with mild alterations in their cardiometabolic risk profile.

Methods

Study design and participants

The study is a randomized, double-blind, controlled, and crossover clinical trial composed of two 8-wk intervention periods, separated by a 4-wk washout period and preceded by a 2-wk lifestyle stabilization period (Figure 1). Based on previous data [21], we believed that the 4-wk washout period was sufficient to prevent any carryover effect from the first treatment before starting the second treatment. Participants were instructed to substitute 5% of their total caloric intake provided by added sugars with either maple syrup or an artificially flavored sucrose syrup for 8 wk.

FIGURE 1.

Graphical representation of the study protocol. 24h-R, 24-h recall; MRI, magnetic resonance imaging; OGTT, oral glucose tolerance test.

The primary outcome of this clinical trial focused on assessing the impact of this sugar substitution on glucose homeostasis, and the secondary outcomes focused on other key cardiometabolic parameters such as blood pressure and anthropometric measurements. Additionally, the study aimed to explore changes in the composition of gut microbiota.

The 2 intervention phases (maple syrup and sucrose syrup) were closely supervised by a registered dietitian to ensure that each participant maintained an isocaloric and stable diet in terms of macronutrient distribution. The dietary intervention is detailed in the Supplemental Methods. Participants were randomized to a treatment sequence using an in-house electronic management platform. Gender-stratified block randomization ensuring 2:2 treatment allocation for each gender to either the intervention group consuming maple syrup or the control group was used. The treatment sequence was hidden from participants, study coordinators, laboratory technicians, and students throughout the study. The randomization process was under the responsibility of the lead study coordinator. The sequence codes were revealed after the completion of all statistical analyses. The acceptability of substituting maple syrup for refined sugars was rated by participants (Supplemental Table 1).

The inclusion criteria were males and females aged between 18 and 75 years with a BMI between 23 and 40 and ≥1 of the following criteria: fasting plasma insulin >42 pmol/L using our new analytic method, corresponding to a threshold value of 60 pmol/L with the former method that was predictive of a higher risk of cardiovascular disease in the Québec population [22]; glucose between 5.5 and 6.9 mmol/L, glycated hemoglobin level of 5.7%–6.4% and triglycerides of >1.35 mmol/L [23]. Participants were excluded from the study if they had an aversion to the taste of maple syrup, regularly drank alcohol (>2 glasses/d), had a significant change in body weight in the past 3 mo (±5% of their body weight) or were in the process of losing weight, if they had taken regular probiotic supplements in the past 3 mo, had eating disorders, or had undergone major surgery 3 mo before the study or if they had one planned if they had intestinal malabsorption, cirrhosis, or chronic kidney disease. Females who were pregnant or planning a pregnancy as well as those who were breastfeeding were also excluded.

Using group electronic messages sent to university and Institute of Nutrition and Functional Foods (INAF) members, 299 individuals contacted the researchers for information on this study. Overall, 112 participants met e-mail or phone prescreening criteria and were scheduled for a screening visit at the clinical investigation unit of INAF (Figure 2). Before the visit, participants were asked to follow a 12-h overnight fast, abstain from alcohol for 48 h, and abstain from vigorous exercise 24 h before the visit. Following a preselection meeting with the registered dietitian, the eligibility of participants was validated (Supplemental Methods). The recruitment and intervention visits were held at the INAF, Université Laval, under the supervision of the lead study coordinator, with the help of 2 graduate students. The study visits took place at INAF, and the magnetic resonance imaging (MRI) was done at the Institut Universitaire de Cardiologie et Pneumologie de Québec (Quebec Heart and Lung Institute). The sample size calculation was based on a previous study, which reported a mean change in area under the curve (AUC) of −0.3 in 25 participants who consumed a control meal supplemented with products containing polyphenols present in maple syrup compared with that in those who consumed the control meal alone [24]. With a power of 80% and a type 1 error of 5%, it was determined that a minimum number of 22 participants was required to observe significant changes in glucose homeostasis. To account for potential dropouts, we aimed to recruit ≥30 participants who received both sucrose and maple syrup for the final analysis, assuming an attrition rate of 15%.We thus aimed to recruit ≥30 participants who had received the sucrose and maple syrup treatments for the final analysis. All participants provided written informed consent before they participated in the study. This clinical trial was registered on clinicaltrials.gov as NCT04117802 and was conducted according to the guidelines described in the Declaration of Helsinki and was approved by the Ethics Committee of Université Laval (approbation number 21793).

FIGURE 2.

Participant flow diagram for recruitment, randomization, and data collection.

Sample collection during intervention visits

Before and after each intervention phase, intervention visits were held, and participants were instructed to follow a 12-h fast. Anthropometric measures, blood pressure, heart rate, and body composition, as measured by dual-energy X-ray absorptiometry (DEXA) scanner, were measured. Blood was drawn to assess the plasma lipid profile. Then, a 3-h 75-g oral glucose tolerance test (OGTT) was performed. Abdominal adiposity and liver fat accumulation were measured by MRI (Figure 1). The detailed methods are listed in the Supplemental Methods.

Bacterial DNA extraction and 16S rRNA gene sequencing

Stool samples were collected before the beginning of each phase and after the sucrose and maple syrup treatment period using the EasySampler Stool Collection Kit (ALPCO). Samples were stored at −20 °C until they were delivered to the research center, where they were stored at −80 °C until analysis. Bacterial genomic DNA was extracted from 1 g of fresh fecal material from each sample. Samples were homogenized with 10 mL of Inhibitex (Qiagen). Part of the lysate was transferred to tubes containing 0.1 mm zirconium beads, 50 mg lysozyme, and 500 U of mutanolysin. Tubes were incubated at 37 °C for 1 h. The tubes were then vortexed 2 times for 1 min using a bead beater. The suspension was heated for 5 min at 95 °C. Then, the rest of the protocol was processed using a DNA extraction kit (Qiagen QIamp DNA Stool Mini Kit) following the manufacturer’s instructions. DNA concentrations were evaluated with Quant-IT PicoGreen dsDNA reagent (Invitrogen). Extracted DNA was stored at −20 °C until they were sent to Centre d’expertise et de services Génome Québec.

The composition of the distal gut microbiota was determined by 16S rRNA amplicon sequencing. Briefly, for each DNA sample, PCR amplification of the V3–V4 region was performed using the primers 341F and 805R and followed by sequencing on an Illumina MiSeq sequencing platform, as previously detailed [25,26]. The resulting sequences were processed using the DADA2 package (v1.16) [27] in the R environment (http://www.R-project.org), and associations to bacterial taxa were obtained using the RDP classifier algorithm (v2.2) [28] trained against the Silva database 138.1 [29]. To quantify bacterial α-diversity, Shannon and Simpson reciprocal indexes were calculated. Principal component analysis (PCA) was performed on the Aitchison distance matrix to measure β-diversity. Identification of differentially abundant bacteria between 2 distinct biological conditions was measured with DESeq2 [30]. P values were false discovery rate corrected using the benjamini hochberg (BH) method, and a P value of <0.05 was considered statistically significant.

Metagenomics shotgun profiling

Stool samples from baseline and maple syrup treatment groups were also analyzed by shotgun metagenomic sequencing. Metagenomic libraries were prepared with NEBNext Ultra II DNA Library Prep Kit for Illumina (New England BioLabs) as per the manufacturer’s recommendations. Libraries were quantified using the Kapa Illumina GA with Revised Primers-SYBR Fast Universal kit. The average size fragment was determined using a LabChip GX (PerkinElmer) instrument. Libraries were sequenced on the Illumina NovaSeq X platform using a 2× 150-bp paired-end run. Metagenomics reads were quality-filtered using Trimmomatic with a cutoff of ≥Q20 [31] and further filtered to remove the host-origin reads using Bowtie2 [32]. Host-decontaminated reads were taxonomically profiled using Kraken2 [33] against the RefSeq bacterial genomes. The abundance of species in annotated reads from a metagenomic sample was computed using Bracken [34]. Species that were not present in ≥10% of all samples were discarded. Functional analyses were performed using HUMAnN3 [35], and MetaCyc [36] pathways were generated from HUMAnN3 output. The R Phyloseq package [37] was used to perform all diversity analyses. Shannon and Simpson diversity indices were used to reflect the richness and evenness of microbial representation in a sample. Additionally, the similarity of microbial communities between samples was calculated using PCA based on the Aitchison distance. Detection of differentially abundant taxa or pathways between groups was performed with DESeq2 [30] with significance set at 0.05.

Fecal short-chain fatty acids

Fecal short-chain fatty acids (SCFAs) were quantified by gas chromatography. Feces were collected and kept frozen at −80 °C until extraction. After the addition of 1 mL H₂O per 100 mg material, fecal suspensions were homogenized for 2 min, then centrifuged at 18,000 × g for 10 min at 4 °C. The supernatant was spiked with 4-methylvaleric acid and acidified with phosphoric acid 10%. To extract SCFAs, samples were mixed for 2 min with an equal volume of diethyl ether, then centrifuged at 18,000 × g for 10 min at 4 °C. Organic phase analysis was performed on a GC-FID system (Shimadzu), constituting a GC 2010 Plus gas chromatograph equipped with an AOC-20s autosampler, an AOC-20i autoinjector, and a flame ionization detector. The system was controlled by GC solution software. SCFA were separated on a Nukol capillary GC column (30 m × 0.25 mm internal diameter; 0.25-μM film thickness; Supelco analytical). The column flow was constant at 1.3 mL/min of hydrogen. The injector was set at 230 °C and the detector at 250 °C. The oven temperature was initially programmed at 60 °C, then increased to 200 °C at 12 °C/min and held for 2 min. SCFAs were quantified using a 5-point calibration curve prepared with a mixture of acetic acid, propionic acid, butyric acid, isobutyric acid, valeric acid, isovaleric acid, and internal standard 4-methyl valeric acid. Phosphoric acid was purchased from VWR. Diethyl ether (99.5%) and all 99% grade standards (acetic acid, propionic acid, isobutyric acid, butyric acid, isovaleric acid, valeric acid, and internal standard 4-methyl valeric acid) were purchased from Sigma-Aldrich.

Statistical analyses

All data were entered twice independently and crosschecked for accuracy by the study coordinator. Descriptive characteristics are reported as means ± SDs or medians with IQRs for continuous variables according to their distribution. Categorical variables were expressed as percentages.

To analyze the absolute change between data from week 8 minus week 0 for the first period and week 20 minus week 12 for the second period, daily dietary intakes, glucose homeostasis indices, and indices of cardiometabolic health, a 2-period crossover statistical design using a linear mixed model was performed. Three fixed factors were defined: 1 related to the sequences, comparing the subjects (n = 23) starting with sucrose syrup and switching to maple syrup after a 4-wk washout period with subjects (n = 19) starting with maple syrup and receiving sucrose syrup in the second period. This factor was used to validate that carryover effect, and no significant result was observed. The second factor was related to the period effect, and the last fixed factor was the comparison between maple syrup diet and sugar syrup diet. These 2 factors were analyzed as repeated measurements using a compound symmetry covariance structure.

The Kenward–Roger approximation for denominator degrees of freedom was used to improve the finite sample performance. A period-heterogeneous variance and a treatment-heterogeneous variance were investigated. Using the smallest Akaike information criterion and Schwarz Bayesian criterion, we concluded that no heterogeneous variance structure was necessary.

The normality assumption was tested with the Shapiro–Wilk test, using the residuals from the statistical model and transforming them by the Cholesky metric. The graphical representation of the marginal linear predictor with studentized residuals was used to study the homogeneity of variances. Some variables were log-transformed to meet normality, and variance assumptions and descriptive values reported in the tables used medians and IQRs. The logarithm transformations were applied on data, and changes between periods were analyzed as the logarithm of ratios. All analyses were adjusted for age, gender, and BMI. Statistical analysis was not planned for multiplicity of tests for secondary outcomes. Results with P values of ≤0.05 were considered significant. Data were analyzed using the statistical packages SAS statistical software, version 9.4, and SAS Studio, version 3.8.

Results

Trial flow, baseline characteristics, and compliance

Recruitment of study participants took place in the greater Québec City metropolitan area between September 2019 and March 2021. Of the 112 participants screened, 47 were included in the randomized controlled trail. During stabilization, 5 participants withdrew from the study for personal reasons. Forty-seven participants were randomized in the first phase of the study. Four participants withdrew during phase 1 (lost to follow-up, n = 1; personal reason related to COVID-19, n = 2; dislikes syrup taste, n = 1). A total of 43 participants were involved in phase 2. Only 1 participant withdrew from phase 2 (dislikes syrup taste, n = 1). Analyses were conducted on a final sample of 42 participants who completed the 2 intervention phases (Figure 2).

Characteristics of volunteers at baseline are presented in Table 1. Baseline characteristics for age, gender, anthropometric, physiologic, or metabolic variables were well distributed at the baseline, and no significant differences were detected between the groups and remained nonsignificantly different after adjustment for age, gender, and BMI (Table 1).

TABLE 1.

Sequence comparison of baseline characteristics obtained at visit 1 (V1).

Variable	V1 (n = 42)
	Sucrose syrup (n = 23)	Maple syrup (n = 19)
Age, y	53.78 ± 16.99	56.10 ± 14.64
Gender, % female (n)	47.83 (11)	68.42 (13)
Weight (kg)	81.90 ± 12.99	80.33 ± 15.78
Height (m)	1.70 ± 0.11	1.67 ± 0.10
BMI (kg/m2)	28.12 ± 3.08	28.46 ± 4.16
Waist circumference (cm)	98.15 ± 11.72	100.09 ± 12.80
SBP (mm Hg)	120.32 ± 13.14	119.42 ± 13.24
DBP (mm Hg)	73.68 ± 9.94	73.17 ± 9.19
Heart rate (bpm)	68.24 ± 7.55	68.47 ± 9.45
ApoB (g/L)	1.01 ± 0.16	1.03 ± 0.25
Total-C (mmol/L)	5.24 ± 0.83	5.37 ± 1.18
HDL-C (mmol/L)	1.42 ± 0.41	1.36 ± 0.39
LDL-C (mmol/L)	3.12 ± 0.67	3.22 ± 0.93
Total-C/HDL-C	3.94 ± 1.09	4.16 ± 1.20
TG (mmol/L)	1.53 ± 0.67	1.59 ± 0.79
Fasting free fatty acids (mmol/L)	0.46 ± 0.14	0.46 ± 0.14
Fasting glycemia (mmol/L)	5.14 ± 0.52	5.19 ± 0.46
Fasting insulinemia (pmol/L)	97.07 ± 72.13	84.26 ± 42.16
Fasting C-peptide (pmol/L)	516.83 ± 158.15	589.10 ± 175.58
AST (U/L)	19.74 ± 7.36	19.94 ± 5.81
ALT (U/L)	27.61 ± 11.50	31.33 ± 14.42

Results are presented as raw means ± SD.

Abbreviations: ALT, alanine aminotransferase; ApoB, apolipoprotein B; AST, aspartate aminotransferase; DBP, diastolic blood pressure; HbA1C, glycated hemoglobin; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; SBP, systolic blood pressure; total-C, total cholesterol; TG, triglyceride.

No adverse events related to the intervention were reported during the study. Compliance, as measured by the daily intake log, was 92%. Based on the volume of maple syrup and sucrose syrup returned following the interventions, compliance was 96%.

Dietary intake

There were no significant differences in energy intake and total and free sugar consumption between the groups during the protocol (Table 2).

TABLE 2.

Changes in daily dietary intakes of all participants.

Variable	Sucrose syrup		Maple syrup		Δ change (n =34–38)		P	Adjusted P1
	Pre	Post	Pre	Post	Sucrose syrup	Maple syrup
Calories (kcal)	2290.31 ± 619.54	2363.39 ± 686.12	2393.26 ± 611.04	2329.89 ± 658.30	73.08 ± 434.77	−63.37 ± 478.83	0.20	0.24
Carbohydrates (g)	277.05 ± 83.12	277.27 ± 94.00	285.59 ± 89.10	275.94 ± 76.18	0.23 ± 51.93	−9.65 ± 72.44	0.58	0.64
Total sugars (g)	125.28 ± 54.61	120.21 ± 54.65	127.47 ± 53.92	120.46 ± 46.27	−5.07 ± 28.11	−7.01 ± 35.97	0.93	0.95
Free sugars (g)	74.65 ± 47.02	73.77 ± 40.82	77.45 ± 39.90	69.80 ± 34.94	−0.88 ± 20.51	−7.66 ± 27.80	0.15	0.15
Fiber (g)	22.21 ± 7.42	21.61 ± 7.32	21.81 ± 9.00	23.08 ± 7.68	−0.60 ± 5.73	1.27 ± 8.49	0.13	0.13
Total fat (g)	91.37 ± 32.32	97.52 ± 34.29	97.12 ± 28.54	92.66 ± 31.52	6.15 ± 28.61	−4.45 ± 27.51	0.11	0.10
Proteins (g)	87.99 ± 23.63	90.64 ± 26.56	93.99 ± 27.46	94.42 ± 32.79	2.64 ± 19.31	0.73 ± 23.88	0.69	0.61
Alcohol (g)	6.74 ± 9.86	6.54 ± 8.76	5.45 ± 9.40	7.08 ± 9.98	−0.20 ± 7.69	1.63 ± 7.30	0.13	0.15

Results are presented as means ± SD. Analyses were conducted with SAS Studio, and comparisons of the Δ values were performed using the mixed procedure with age, gender, BMI, baseline value, period, and treatment by sequence interaction Three fixed factors were defined: 1 related to the sequences, 1 related to the period effect, and the last fixed factor was the comparison between maple syrup diet and sugar syrup diet. These 2 factors were analyzed as repeated measurements using a compound symmetry covariance structure. The Kenward–Roger approximation for denominator degrees of freedom was used. The mixed procedure was performed with squared root transformed values for alcohol. The P values associated with the “treatment × sequence” interaction were not significant and thus were not included in the table.

¹Adjusted for age, gender, and BMI.

Indices of glucose homeostasis

Following the substitution of refined sugars with maple syrup, we observed a statistically significant decrease in glucose AUC by 50.59 ± 201.92 arbitrary units measured during the OGTT (P = 0.047) and a nonstatistically significant decrease in incremental AUC glucose by 42.91 ± 147.36 arbitrary units (P = 0.11) compared with replacing refined sugars with an equivalent amount of energy from sucrose syrup, leading to an increased AUC of 29.93 ± 154.90 units and incremental AUC of 6.03 ± 130.86 units (Figure 3A–D). No changes in insulin and C-peptide excursion curves and incremental AUC were detected between groups (Figure 3E–J). No differences were found in insulin secretion, insulin sensitivity, insulin clearance, and adipose tissue insulin sensitivity indices between the groups (Table 3).

FIGURE 3.

Effects of treatment on glucose homeostasis. Glycemia before and after (A) sucrose or (B) maple syrup intervention. (C) D incremental area under the curve (iAUC) for glycemia. (D) D AUC for glycemia. Insulinemia before and after (E) sucrose and (F) maple syrup intervention. (G) D iAUC for insulinemia. C-peptide before and after (H) sucrose and (I) maple syrup intervention. (J) D AUC for C-peptide. Results are presented as means ± SD. Analyses were conducted with SAS Studio, and comparisons of the Δ values were performed using the mixed procedure with age, gender, BMI, baseline value, sequence, and treatment by sequence interaction. The model included fixed effects for treatment and a random effect for the participants [random subject (sequence)]. The Kenward–Roger approximation for denominator degrees of freedom was used. ∗P < 0.05.

TABLE 3.

Effects of treatment on glucose homeostasis indices.

Variable	Sucrose syrup		Maple syrup		Δ change (n = 33–42)		P	Adjusted P
	Pre	Post	Pre	Post	Sucrose syrup	Maple syrup
HOMA-2%S (insulin)	76.73 ± 42.06	86.39 ± 60.14	85.61 ± 43.74	84.74 ± 52.59	9.66 ± 53.44	−0.87 ± 43.05	0.41	0.41
HOMA-2%B (C-peptide)	101.19 ± 30.16	105.56 ± 31.61	104.11 ± 29.31	103.11 ± 33.45	4.37 ± 18.14	−1.00 ± 15.84	0.21	0.23
Insulinogenic index (pmol/mmol)1	1.24 (0.71, 1.97)	1.35 (0.81, 1.86)	1.10 (0.56, 2.18)	1.01 (0.65, 2.10)	.06 (−0.48, 0.41)	0.09 (−0.10, 0.36)	0.68	0.68
ADIPO-IR1	33.61 (25.65, 48.81)	38.89 (19.92, 57.38)	38.27 (20.75, 46.63)	34.33 (22.01, 58.77)	−2.63 (−11.32, 5.61)	−1.81 (−11.14, 14.37)	0.37	0.44
Insulin clearance: 120 min1	3.98 (3.40, 4.69)	3.67 (2.96, 4.52)	3.89 (3.04, 5.43)	3.91 (3.00, 5.06)	−0.18 (−0.92, 0.44)	−0.13 (−0.76, 0.43)	0.92	0.91
Insulin clearance: 30 min	2.69 ± 0.87	2.70 ± 1.45	2.92 ± 1.56	2.71 ± 0.76	0.01 ± 1.69	−0.22 ± 1.31	0.67	0.63
SI (dL/kg/min per μU/mL)	1.15 × 10−3 ± 8.73 × 10−4	1.07 × 10−3 ± 8.24 × 10−4	1.24 × 10−3 ± 1.60 × 10−4	1.09 × 10−3 ± 9.20 × 10−4	−9.38 × 10−5 ± 7.60 × 10−4	−1.65 × 10−4 ± 9.39 × 10−4	0.90	0.85
SID (dL/kg/min per μU/mL)	4.85 × 10−4 ± 4.16 × 10−4	4.21 × 10−4 ± 3.52 × 10−4	4.48 × 10−4 ± 5.36 × 10−4	4.40 × 10−4 ± 3.67 × 10−4	−6.70 × 10−5 ± 4.00 × 10−4	−1.30 × 10−5 ± 3.27 × 10−4	0.14	0.16
DISi (pmol/mmol × dL/kg/min per μU/mL)	1.56 × 10−3 ± 1.71× 10−3	2.05 × 10−3 ± 2.81 × 10−3	1.65 × 10−3 ± 2.07 × 10−3	2.36 × 10−3 ± 5.98 × 10−3	3.84 × 10−4 ± 2.86 × 10−3	7.59 × 10−4 ± 6.35 × 10−3	0.27	0.39
DISiD (pmol/mmol × dL/kg/min per μU/mL)1	10.56 × 10−4 (4.73 × 10−4, 18.90 × 10−4)	7.42 × 10−4 (4.60 × 10−4, 15.75 × 10−4)	7.08 × 10−4 (3.79 × 10−4, 17.23 × 10−4)	7.42 × 10−4 (4.60 × 10−4, 15.75 × 10−4)	−0.06 × 10−4 (−5.27 × 10−4, 3.46 × 10−4)	0.53 × 10−4 (−5.61 × 10−4, 3.83 × 10−4)	0.92	0.88

Results are presented as means ± SD or median with IQR. Analyses were conducted with SAS Studio and comparisons of the Δ values were performed using the mixed procedure with age, gender, BMI, baseline value, period, and treatment by sequence interaction. Three fixed factors were defined: 1 related to the sequences, 1 related to the period effect, and the last fixed factor was the comparison between maple syrup diet and sugar syrup diets. These 2 factors were analyzed as repeated measurements using a compound symmetry covariance structure. The Kenward–Roger approximation for denominator degrees of freedom was used. Mixed procedure was performed with log10-transformed values for clearance insulin 120, clearance insulin 30, SI, and DISi^D. The P values associated with the “treatment × sequence” interaction were not significant and, thus, were not included in the table.

Abbreviations: HOMA, homeostasis model assessment; ADIPO-IR, adipose tissue insulin resistance index; SI, insulin sensitivity; SI^D, dynamic insulin sensitivity; DISi, disposition index; DISi^D, dynamic disposition index.

¹Median with IQR.

Indices of cardiometabolic health

Replacing 5% of the daily energy intake with maple syrup was not associated with changes in body weight and waist circumference compared with substitution with sucrose syrup (Table 4). However, body composition analyzed by DEXA differed between the groups: android fat decreased by 7.83 g ± 175.05 g in the maple syrup group, whereas it increased in the sucrose syrup group by 67.61 g ± 206.71 g (P = 0.02) (Table 4). No differences regarding changes in total body fat, gynoid fat, and total lean mass were found between the groups (Table 4). When measuring the hepatic, visceral, and subcutaneous fat by MRI, no changes or group differences were detected (Supplemental Table 2).

TABLE 4.

Indices of cardiometabolic health by intervention group.

Variable	Sucrose syrup		Maple syrup		Δ change (n = 38–42)		P	Adjusted P
	Pre	Post	Pre	Post	Sucrose syrup	Maple syrup
Weight (kg)	81.54 ± 14.16	81.94 ± 14.42	81.31 ± 14.22	81.47 ± 14.32	0.40 ± 1.32	0.17 ± 1.29	0.42	0.42
BMI (kg/m2)	28.41 ± 3.65	28.55 ± 3.74	28.32 ± 3.62	28.38 ± 3.70	0.14 ± 0.46	0.06 ± 0.46	0.44	0.44
Waist circumference (cm)	99.17 ± 12.20	99.39 ± 11.92	98.93 ± 12.50	99.61 ± 12.29	0.39 ± 1.21	0.34 ± 1.53	0.88	0.88
SBP (mm Hg)	118.42 ± 13.13	119.29 ± 12.41	118.96 ± 12.75	116.24 ± 11.56	0.87 ± 8.99	−2.72 ± 8.73	0.03	0.03
DBP (mm HG)	73.65 ± 9.90	74.10 ± 9.98	72.98 ± 9.80	71.70 ± 7.44	0.44 ± 5.32	−1.29 ± 5.70	0.08	0.08
Heart rate (bpm)	67.73 ± 7.94	67.98 ± 8.00	67.33 ± 8.91	67.30 ± 10.19	0.63 ± 5.13	0.20 ± 6.48	0.62	0.64
ApoB (g/L)1	1.06 (0.90, 1.17)	1.02 (0.92, 1.13)	1.00 (0.93, 1.16)	0.98 (0.87, 1.16)	−0.01 (−0.11, 0.08)	−0.03 (−0.09, 0.00)	0.20	0.21
Total-C (mmol/L)	5.327 ± 1.132	5.261 ± 0.971	5.265 ± 0.948	5.120 ± 0.946	−0.066 ± 0.563	−0.173 ± 0.447	0.34	0.37
HDL-C (mmol/L)	1.400 ± 0.399	1.356 ± 0.351	1.399 ± 0.376	1.381 ± 0.365	−0.044 ± 0.169	−0.015 ± 0.134	0.37	0.40
LDL-C (mmol/L)	3.210 ± 0.938	3.129 ± 0.798	3.155 ± 0.756	3.052 ± 0.786	−0.074 ± 0.488	−0.131 ± 0.436	0.43	0.46
Total-C/HDL-C	4.005 ± 1.107	4.087 ± 1.192	3.965 ± 1.075	3.901 ± 1.059	0.081 ± 0.533	−0.089 ± 0.404	0.64	0.67
TG (mmol/L)1	1.38 (1.17, 1.78)	1.48 (1.06, 1.84)	1.39 (1.01, 1.70)	1.40 (1.13	0.11 (−0.19, 0.36)	0.00 (−0.26, 0.22)	0.48	0.46
AST (U/L)1	19.5 (15, 24)	21.00 (16, 26)	20 (16, 26)	21 (17, 25)	1 (−1, 3)	0 (−3, 5)	0.89	0.93
ALT (U/L)1	25 (22, 39)	26 (20, 35)	24 (21, 37)	25 (20, 39)	0 (−2, 3)	−1 (−4, 2)	0.43	0.38
FFA	0.47 ± 0.14	0.48 ± 0.12	0.50 ± 0.16	0.48 ± 0.14	−0.0006 ± 0.17	−0.0359 ± 0.18	0.95	0.73
Total body fat (g)	30379.2 ± 8822.5	30891.4 ± 8942.2	30222.7 ± 8857.6	30436.9 ± 8920.0	512.2 ± 1203.8	214.1 ± 1105.9	0.28	0.29
Trunk fat (g)	15782.8 ± 5140.6	16151.9 ± 5197.5	15701.8 ± 5212.5	15771.3 ± 5163.0	369.0 ± 850.3	69.6 ± 735.7	0.12	0.13
Android fat (g)	2892.5 ± 1048.0	2960.1 ± 1055.3	2888.6 ± 1080.6	2875.2 ± 1032.4	67.6 ± 206.7	−13.4 ± 173.4	0.02	0.02
Gynoid fat (g)	4792.9 ± 1479.5	4846.0 ± 1526.6	4807.6 ± 1479.3	4817.1 ± 1560.6	53.1 ± 265.3	9.5 ± 288.6	0.51	0.55
Total lean mass (g)	48559.9 ± 10019.5	48264.2 ± 9852.1	47983.2 ± 9783.9	48000.4 ± 9813.1	−295.7 ± 988.2	17.1 ± 1220.6	0.29	0.27

Results are presented as means ± SD or median with IQR. Analyses were conducted with SAS Studio and comparisons of the Δ values were performed using the mixed procedure with age, gender, BMI, baseline value, period, and treatment × sequences interaction. Three fixed factors were defined: 1 related to the sequences, 1 related to the period effect, and the last fixed factor was the comparison between maple syrup diet and sugar syrup diet. These 2 factors were analyzed as repeated measurements using a compound symmetry covariance structure. The Kenward–Roger approximation for denominator degrees of freedom was used. Significant P values are in bold (P < 0.05). Analyses were performed on log10-transformed values for ApoB and total-C/HDL-C.

Abbreviations: ALT, alanine aminotransferase; AST, aspartate aminotransferase; ApoB, apolipoprotein B; DBP, diastolic blood pressure; HbA1C, glycated hemoglobin; SBP, systolic blood pressure; TG, triglyceride; total-C, total cholesterol; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol.

¹Median with IQR.

Maple syrup substitution for refined sugars significantly decreased systolic blood pressure (SBP) by 2.72 ± 8.73 mm Hg compared with an increase of 0.87 ± 8.99 mm Hg when substituting refined sugars with sucrose syrup (P = 0.03) (Table 4). Diastolic blood pressure was also decreased by 1.29 ± 5.7 mm Hg in the maple syrup intervention arm compared with an increase of 0.44 ± 5.32 mm Hg in the sucrose syrup arm, although this finding did not reach statistical significance (P = 0.08) (Table 4). The intervention was not associated with significant changes in heart rate, aspartate aminotransferase, alanine aminotransferase, free fatty acids, and other plasma lipids (Table 4).

Gut microbiota and SCFAs

To investigate changes in gut microbiota composition following the dietary intervention, fecal samples were collected before and after each trial arm, and 16S rRNA sequencing analysis was performed. The diversity of the intestinal microbiota was estimated at the amplicon sequence variant level. No differences in microbial richness were observed between all groups included in this study, as shown by the α-diversity Simpson and Shannon indices (Figure 4A, B). Specifically, both indices reflect community diversity; Simpson index is more sensitive to species evenness, although Shannon index is more sensitive to species richness [38]. The determination of β-diversity PCA on the Aitchison distance matrix revealed that the gut microbiota composition was not stratified according to the type of sugars consumed (Figure 4C). However, DESeq2 analysis of the 16S rRNA sequencing, which allows the identification of significantly different bacterial abundance, revealed that the gut microbiota of participants in the maple syrup group harbored significantly less Klebsiella species and Bacteroides pectinophilus than those in the sucrose syrup group (after maple syrup compared with after sucrose syrup intervention) (Figure 4D).

FIGURE 4.

Effects of dietary intervention on gut microbiota composition: (A) Simpson reciprocal index; (B) Shannon index; (C) principal component; (PC) analysis of gut microbial composition. (D) DESeq2 analysis was performed to explore the taxa that most strongly discriminated between dietary interventions. DESeq2 analysis was performed to explore the (E) taxa and (F) pathways that most strongly discriminated before and after maple syrup (MS) intervention.

To further document taxonomic and functional changes in the gut microbiota of participants who replaced their daily intake of sugar with maple syrup and potentially identify microbial determinants of the maple syrup response, complementary shotgun metagenomic analyses were performed on samples before and after the maple syrup consumption (premaple syrup compared with postmaple syrup). As with 16S rRNA sequencing, no changes were observed on α- and β-diversities (Supplemental Figure 1). However, DESeq2 analysis of the metagenomics data showed that the consumption of maple syrup significantly increased the growth of Lactocaseibacillus casei (originally classified as Lactobacillus casei) and Clostridium beijerinckii (Figure 4E). Additionally, and in line with the 16S rRNA data, consumption of maple syrup was associated with a decreased abundance of Klebsiella quasipneumoniae.

Metagenomics sequencing analysis further revealed that maple syrup intake was associated with a downregulation of specific gut bacterial pathways, such as the L-arginine and L-ornithine degradation (ORNARGDEG-PWY) pathway and L-arginine, putrescine, and 4-aminobutanoate degradation pathway (ARGDEG-PWY) (Figure 4F).

With regard to SCFAs, replacing sugar with maple syrup and sucrose syrup did not significantly alter the concentrations of acetic, propionic, isobutyric, butyric, isovaleric, valeric, and hexanoic acids in the feces of participants (Supplemental Table 3).

Discussion

This randomized crossover clinical trial documented the impact of replacing a refined sugar source (sucrose) with an equivalent quantity of maple syrup for 8 wk in overweight participants with mild metabolic alterations. Substituting refined sugars with maple syrup resulted in a significant reduction in the glucose AUC compared with replacing refined sugar with sucrose syrup, which led to an increase in the AUC. Likewise, the substitution of refined sugars with maple syrup was found to decrease android fat mass and systolic blood pressure, whereas replacing refined sugars with sucrose syrup increased those parameters. These results were obtained by substituting only 5% of the total caloric intake provided by refined sugars with maple syrup and observed after only 8 wk of intervention. Despite remaining a significant source of added sugar, which should be consumed in moderation, these results indicate that maple syrup does not exacerbate cardiometabolic risk factors compared with sucrose syrup when consumed in moderate amounts (e.g., between 20 and 40 mL/d in this study) for 8 wk in participants with mild metabolic alterations.

In agreement with this study, our group has previously shown that daily administration of maple syrup, representing 4%–10% of the total daily caloric intake of diet-induced obese rats and mice, was less harmful to glucose tolerance, insulin resistance, and hepatic steatosis compared with administration of equivalent amounts of sucrose [9,10]. Other animal studies have also reported that a single administration of maple syrup decreased glycemia compared with an equivalent dose of refined sugar [[6], [7], [8]]. Despite being a source of free sugars, maple syrup also contains micronutrients (vitamins, minerals) and is a source of unique polyphenols like quebecol and lignans [16,17]. Although maple and sucrose syrup both contain identical sugar composition and caloric content, polyphenols found in maple syrup may have limited glucose intestinal absorption by inhibiting α-glucosidase in the postprandial state [39]. In line with this, it was reported that the administration of maple syrup reduced intestinal glucose absorption in rats compared with an equivalent dose of sucrose [8]. We further reported that maple syrup reduces intestinal α-glucosidase activity in obese mice [10]. The potential effect of maple syrup to reduce glucose absorption could be of interest in preventing postprandial hyperglycemia in human subjects with metabolic alterations.

The observation of decreased android fat mass upon replacing refined sugars with maple syrup, compared with sucrose syrup, is of great interest. We speculate that a reduced glucose absorption, and therefore lower postprandial glycemic response (as suggested by lower AUC during the OGTT), may have contributed to this small, yet significant, decrease in android fat when comparing the maple syrup and sucrose syrup interventions. Although a 70-g fat mass difference between the 2 groups seems small in absolute terms, it is important to note that this effect was observed in only 8 wk and thus may become more clinically relevant after many months of dietary interventions. The android region, located above the pelvis and extending upward to include 20% of the abdomen until the chin, contains both subcutaneous and visceral adipose tissue. It is worth noting that although maple syrup and sucrose syrup groups gained a nonsignificant amount of weight during the intervention as evidenced by DEXA analyses, participants who consumed maple syrup tended to gain a lower amount of weight, which was mostly located in the gynoid region rather than the abdominal region. This finding is of physiologic importance and clinical relevance, as visceral adipose tissue is a major risk factor for cardiometabolic diseases [40]. Although we did not observe significant changes in visceral adipose tissue measured by MRI, this could partly be explained by the low number of participants who completed all 4 MRI visits (n = 28) than those who completed DEXA visits (n = 42).

Blood pressure is an important marker of cardiometabolic health, as hypertension is the leading modifiable risk factor for cardiovascular disease and premature death worldwide [41,42]. Replacement of refined sugars with maple syrup corresponding to 5% of the total energy intake for 8 wk decreased SBP, whereas this parameter increased in participants upon sucrose syrup substitution for refined sugars. Despite being modest, the 2.72-mm Hg decrease following substitution with maple syrup was observed after only 8 wk and could therefore lead to greater blood pressure reduction in the long term. Although this blood pressure–lowering effect is less than what can be expected from antihypertensive drugs, it is important to mention that dietary interventions aiming to decrease blood pressure, namely the Dietary Approaches to Stop Hypertension, or low-sodium:high-potassium; low-sodium, low-calorie; as well as low-calorie diets, were reported to reduce SBP by 3.07 mm Hg on average, similar to the same range as observed in this study [43]. Furthermore, lignans are a type of polyphenols found in maple syrup [7], and a meta-analysis of randomized controlled trials reported that supplementation with lignan-rich flaxseed compounds was associated with a significant reduction of SBP by 2.85 mm Hg [44]. In line with this, a cross-sectional study involving 6633 males showed that dietary lignans and other polyphenols were inversely correlated with blood pressure, fasting plasma glucose, and triglycerides [45] and associated with a significantly lower risk of coronary artery disease [46,47].

The bioactivity of lignans, like many other polyphenols, relies on their metabolization by the gut microbiota [14,18]. Beyond its polyphenolic content, maple syrup also contains inulin, a prebiotic fiber [13,17], which is defined as a substrate selectively used by host microorganisms to confer health benefits [48]. The microbial fermentation of both polyphenols and fibers produces an array of metabolites that can promote a thicker intestinal mucus layer and stronger barrier function, thus decreasing intestinal permeability and preventing translocation of microbial compounds such as the proinflammatory bacterial endotoxin LPS known to promote metabolic endotoxemia [18,20].

16S rRNA gene–based analyses revealed that the gut microbiota of participants in which refined sugars were substituted with the sucrose syrup group harbored significantly more B. pectinophilus and Klebsiella species than the gut microbiota of those after maple syrup substitution. Likewise, shotgun metagenomic analyses revealed that maple syrup substitution for refined sugars decreased the abundance of K quasipneumoniae, an opportunistic pathogen causing antibiotic-resistant infections in the gastrointestinal tract [49]. Klebsiella sp. has previously been associated with deteriorated cardiometabolic outcomes [[50], [51], [52]]. For instance, a recent study demonstrated that Klebsiella sp. was increased in the feces of prehypertensive and hypertensive populations [51] and the gut microbiota of individuals with atherosclerotic cardiovascular diseases [53]. Compared with sucrose replacement, maple syrup substitution for refined sugars also decreased the expression of bacterial pathways related to L-arginine and L-ornithine degradation and to L-arginine, putrescine, and 4-aminobutanoate degradation. These bacterial pathways were previously associated with LPS-induced IL-6 proinflammatory response [54]. Klebsiella sp. was also previously reported to play a key role in the initiation and symptom severity in patients with Crohn disease through the production of proinflammatory cytokines [55,56]. This suggests that by altering the gut microbiota composition, maple syrup substitution for refined sugars may have decreased the gut inflammatory cytokine production capacity, thus preventing further cardiometabolic alterations compared with sucrose syrup substitution.

The metagenomic analyses also revealed that substituting refined sugars with maple syrup increased the abundance of both C. beijerinckii and L. casei. A robust body of literature supports that intake of the probiotic L. casei results in modest but consistent improvements in several metabolic risk factors in subjects with metabolic diseases [57]. In addition, supplementation of L. casei in murine models of T2D was shown to improve metabolic parameters and was associated with decreased circulating inflammatory markers [[58], [59], [60]]. Although this warrants further investigation, this positive shift in gut microbiota composition following the substitution of refined sugars with maple syrup may have contributed to preventing further deterioration of cardiometabolic risk markers compared with that after sucrose syrup substitution.

This randomized, double-blind, controlled crossover trial has some limitations. This study investigated the metabolic impact of substituting refined sugars with maple syrup in both males and females. Nevertheless, these results must be replicated in larger and longer clinical interventions and confirmed in more diverse populations. Owing to restricted access to the IUCPQ hospital during the COVID-19 pandemic, many participants did not complete their planned MRIs, which decreased the statistical power of this analysis. Multiple testing of secondary outcomes increases risk of family-wise error rate. To mitigate this, each outcome was analyzed separately, and multiple testing correction was applied using the Tukey–Kramer method. The decision to use a per-protocol approach in this study facilitated the detection of the true effect of the interventions. Nevertheless, it is important to recognize that a per-protocol approach may introduce a potential selective reporting bias, as it tends to focus on outcomes that confirm differences between the 2 interventions studied. However, the very low dropout rate in this study mitigates this problem considerably.

This study also has major strengths, including a robust, crossover study design. Since we did not observe any carryover effects as indicated by the absence of significant interaction between treatment and sequence, we concluded that the 4-wk washout period effectively minimized potential influences of treatment order and carryover effects. Furthermore, to our knowledge, most studies investigating the impact of refined sugar replacement by maple syrup on metabolic health have been limited to rodent models, and its impact on gut microbiota remains poorly studied, which further emphasizes the novelty of our work. As 90% of added sugar intake in the diet comes from industrial food and beverage products [61], replacing 5% of the energy intake from added sugar with maple syrup is not only clinically relevant but also achievable at the population level. In line with this, the effect of maple syrup over sucrose syrup on cardiometabolic risk factors was observed after only 8 wk in participants with mild metabolic alterations, which suggests that these outcomes could even be more important in longer-term studies.

In conclusion, replacing refined sugars representing only 5% of the total caloric intake with maple syrup for 8 wk led to reduced glycemic excursions during OGTT compared with sucrose syrup replacement. Likewise, the substitution of refined sugar with maple syrup led to a significant decrease in both android fat mass and systolic blood pressure compared with replacing refined sugar with sucrose syrup, which was found to increase such parameters. The changes in cardiometabolic risk factors were associated with specific changes in the gut microbiota composition and function, notably with higher growth of beneficial species and lower relative abundance of Klebsiella species. Although maple syrup represents a source of added sugar that should be consumed in moderation, results from this trial suggest that substituting refined sugars with maple syrup does not exacerbate cardiometabolic risk factors in individuals with mild metabolic alterations compared with sucrose syrup substitution.

Acknowledgments

We thank Jean-Philippe Songpadith and Pauline Navarro for their help with preparing the sucrose syrup solution. We also thank Steeve Larouche from the INAF nursing staff for his assistance with data collection.

Author contributions

The author’s responsibilities were as follows—A. Marette, M-CV, A-MC, J-PD, BJA, GP, CG: designed the research; A. Morissette, A-LA, TG, LR, JM, TVV, EL: conducted the research; A. Morissette, A-LA, TG, TVV: analyzed and interpreted the data; A. Morissette, A-LA: drafted the manuscript; GP, A. Marette, M-CV: revised the manuscript; KD, CG, BAJ, J-PD, A-MC: critically reviewed the manuscript; A. Marette, M-CV: have primary responsibility for final content; and all authors: read and approved the final manuscript.

Conflict of interest

A. Marette reports financial support was provided by the Producteurs et productrices acéricoles du Québec, Ministère de l’Agriculture, des Pêcheries et de l’Alimentation du Québec, and Producteurs et productrices acéricoles du Québec.

Funding

This study was funded by the Producteurs et productrices acéricoles du Québec (PPAQ) and the Ministère de l’Agriculture, des Pêcheries et de l’Alimentation du Québec (MAPAQ). The sponsors had no role in study design, in the collection, analysis, and interpretation of data, in the writing of the report, or in the decision to submit the article for publication. A. Morissette was awarded studentships from the Fonds de recherche du Québec: Santé (FRQS) and the PPAQ, A-LA was supported by the Foundation of the Centre Hospitalier Universitaire de Québec. A. Marette holds a Pfizer research Chair in the pathogenesis of insulin resistance and cardiovascular diseases. M-CV is a Canada Research Chair in Genomics Applied to Nutrition and Metabolic Health. A-MC, CG, and BJA are funded by career awards from the FRQS. TG is funded by a master’s degree award from the FRQS.

Appendix A. Supplementary data

The following is the Supplementary data to this article: