Substitution of dietary monounsaturated fatty acids from olive oil for saturated fatty acids from lard increases low-density lipoprotein apolipoprotein B-100 fractional catabolic rate in subjects with dyslipidemia associated with insulin resistance: a randomized controlled trial

Abstract

Background

The substitution of monounsaturated acids (MUFAs) for saturated fatty acids (SFAs) is recommended for cardiovascular disease prevention but its impact on lipoprotein metabolism in subjects with dyslipidemia associated with insulin resistance (IR) remains largely unknown.

Objectives

This study aimed to evaluate the impact of substituting MUFAs for SFAs on the in vivo kinetics of apolipoprotein (apo)B-containing lipoproteins and on the plasma lipidomic profile in adults with IR-induced dyslipidemia.

Methods

Males and females with dyslipidemia associated with IR (n = 18) were recruited for this crossover double-blind randomized controlled trial. Subjects consumed, in random order, a diet rich in SFAs (SFAs: 13.4%E; MUFAs: 14.4%E) and a diet rich in MUFAs (SFAs: 7.1%E; MUFAs: 20.7%E) in fully controlled feeding conditions for periods of 4 wk each, separated by a 4-wk washout. At the end of each diet, fasting plasma samples were taken together with measurements of the in vivo kinetics of apoB-containing lipoproteins.

Results

Substituting MUFAs for SFAs had no impact on triglyceride-rich lipoprotein apoB-48 fractional catabolic rate (FCR) (Δ = –8.9%, P = 0.4) and production rate (Δ = 0.0%, P = 0.9), although it decreased very low-density lipoprotein apoB-100 pool size (PS) (Δ = −22.5%; P = 0.01). This substitution also reduced low-density lipoprotein cholesterol (LDL-C) (Δ = −7.0%; P = 0.01), non–high-density lipoprotein cholesterol (Δ = −2.5%; P = 0.04), and LDL apoB-100 PS (Δ = −6.0%; P = 0.05). These differences were partially attributed to an increase in LDL apoB-100 FCR (Δ = +1.6%; P = 0.05). The MUFA diet showed reduced sphingolipid concentrations and elevated glycerophospholipid levels compared with the SFA diet.

Conclusions

This study demonstrated that substituting dietary MUFAs for SFAs decreases LDL-C levels and LDL PS by increasing LDL apoB-100 FCR and results in an overall improved plasma lipidomic profile in individuals with IR-induced lipidemia.

Trial registration

This trial was registered as clinicaltrials.gov as NCT03872349.

Introduction

Insulin resistance (IR) and atherogenic dyslipidemia are major components of metabolic syndrome, which is a cluster of metabolic abnormalities associated with a 5-fold risk of developing type 2 diabetes (T2D) and a 2-fold risk of developing cardiovascular diseases (CVDs) [1,2]. IR-induced dyslipidemia is characterized by a so-called “lipid triad” including an elevation in plasma triglyceride (TG) levels associated with a reduction in HDL cholesterol concentrations and the presence of small dense (sd) LDL particles (<20 nm) [[3], [4], [5]]. Studies have shown that the overaccumulation of atherogenic triglyceride-rich lipoproteins (TRLs) observed in subjects with IR is the result of an elevated production rate (PR) together with a low fractional catabolic rate (FCR) of apolipoprotein (apo)B-containing TRLs of hepatic and intestinal origins [6,7]. This is of great interest because even though the prevalence of IR has risen at an alarming rate in recent years and is now afflicting over 40% of adults in the United States, there are still no evidence-based dietary guidelines targeting the prevention and treatment of dyslipidemia associated with IR [8].

Several lines of evidence from observational and intervention studies have shown the replacement of dietary SFAs with either PUFAs or MUFAs to be beneficial in reducing CVD risk mostly by lowering LDL cholesterol, which is the primary target of current clinical guidelines [[9], [10], [11]]. However, these cardioprotective effects do not always translate into improvements in other CVD risk factors and are highly dependent on the food matrix [12,13]. Recent studies have reported a beneficial role of MUFA-rich olive oil in primary and secondary CVD prevention, both as part of the Mediterranean diet and on its own [[14], [15], [16]]. In addition, previous work from our group showed that the substitution of ω-6 PUFAs from safflower oil for SFAs from lard decreases LDL apoB-100 PR and pool size (PS) in males with dyslipidemia associated with IR [17]. Still, the scarcity of data related to the impact of MUFA-rich olive oil consumption on the metabolism of TRLs in individuals with IR is limiting our ability to give targeted nutritional guidance to this population at high CVD risk.

The primary objective of this study was to examine how the substitution of MUFAs from olive oil for SFAs from lard modifies the in vivo kinetics of apoB-containing lipoproteins in males and females with dyslipidemia and IR. In exploratory analyses, we also examined how substituting MUFAs for SFAs impacts the plasma lipidome to elucidate the key mechanisms behind the observed changes in lipoprotein metabolism. We hypothesized that substituting MUFAs for SFAs reduces the PR of apoB-containing lipoproteins as well as increases their FCR. We also hypothesized that changes in the plasma lipidome profile after the substitution of MUFAs for SFAs are partially correlated with concurrent changes in the metabolism of apoB-containing lipoproteins.

Methods

The research protocol was approved by the Laval University Medical Center ethical review committee and written consent was obtained from all participants and recruitment was done between November 2019 and November 2020. The study was conducted between January 2020 and June 2021 at the Institute of Nutrition and Functional Food (INAF) of Laval University in Quebec City (Canada) and was registered at clinicaltrials.gov (NCT03872349).

Study subjects

All participants had to be adults between 18 and 65 y old with fasting plasma TG concentrations ≥1.3 mmol/L (150 mg/dL) and <7 mmol/L (619 mg/dL), and fasting insulin concentrations ≥60 pmol/L. Subjects were required to have a stable body weight for ≥3 months prior to the screening, and a waist circumference ≥94 cm (males) and ≥80 cm (females). Fasting plasma TG and insulin concentrations were measured at 2 different occasions before enrollment. Exclusion criteria included smoking (1 cigarette/d), alcoholism (>14 drinks/wk), illicit drug consumption, extreme dyslipidemias (e.g., familial hypercholesterolemia), T2D, a history of CVD or cancer, acute hepatic or renal disease (aspartate aminotransferase/alanine transaminase >1.5× the upper limit of normal, creatinine concentrations >176 μmol/L, and creatine phosphokinase >2× the upper limit of normal), HIV, uncontrolled high blood pressure (>160/110 mm Hg), or any other condition that could interfere with participation in the study. Subjects had to withdraw ω-3 supplementation 3 wk prior to the first screening visit and abstain from for the duration of the study. The use of antidepressants, antihypertensive drugs, or levothyroxine was allowed only if the dose had been stable for ≥3 mo prior to and remained constant during the study. Any other drug, dietary supplement, or natural health product had to be stopped ≥3 wk prior to the beginning and for the duration of the study. Participants had to notify one of the study coordinators if they had to initiate or modify any medication during the study. From the 72 subjects screened for eligibility, 18 participants (12 males and 6 females) with dyslipidemia associated with IR were recruited.

Study design

This crossover double-blind, randomized, controlled trial comprised 2 diets: the high-SFA, low-MUFA control diet (hereafter referred to as the SFA diet) and the low-SFA, high-MUFA experimental diet (hereafter referred to as the MUFA diet). Both diets were consumed in fully controlled conditions for periods of 4 wk each and were separated by a 4-wk wash out. A 2-wk run-in period preceding the first controlled feeding period to help participants familiarize themselves to the protocol and to estimate daily energy requirements (see below). Randomization to diet sequences was performed with 3 blocks of 8 subjects stratified by sex with an allocation ratio of 1:1. The control and experimental diets were consumed under carefully controlled, weight-maintaining conditions to avoid participants experiencing weight fluctuations during the study. Energy requirements of the study participants were estimated using Harris–Benedict’s formula and an assessment of their energy intakes calculated using data from 3 web-based, 24-h dietary recalls [18,19] completed during the run-in period. Body weight was monitored throughout the feeding periods on weekdays before lunch at the Clinical Investigation Unit (CIU) at the INAF under the supervision of a research assistant. If a participant’s weight changed by ±2 kg from the diet-specific baseline value, the participant’s caloric intake was adjusted to stop the weight variation.

Three meals and 1 snack were provided daily to all participants. A 7-d cyclic menu was used. In general, subjects consumed their lunch on weekdays at the CIU and received all their food for the rest of the day and the next morning. On Fridays, participants received their food for the entire weekend. Participants were instructed to consume all the food, and only the food that was provided to them. Subjects had free access to caffeine- and calorie-free beverages. Alcohol consumption was forbidden during the experimental diets and limited to ≤2 drinks/d and ≤7 drink/wk during the run-in and the washout periods. Participants had to complete a daily checklist to identify foods that they had or had not consumed. Dietary compliance was measured using this checklist. Participants had to maintain a constant physical activity level throughout the study. Fasting blood samples were collected at the 2 screening visits, at the beginning of each diet and on the morning of the kinetic study at the end of each diet (weeks –4, –3, 0, 4, 8, and 12). Kinetic studies were conducted on the last day (day 29) of each diet. High-intensity physical activity was forbidden 24 h prior to the kinetic studies. Allocations to treatments were concealed from investigators, participants, study coordinators, and laboratory and medical technicians throughout the study. Codes were unconcealed after all primary statistical analyses had been completed.

Composition of the experimental diets

The nutritional composition of the SFA diet was designed to represent the typical Canadian diet [20], whereas the nutritional composition of the MUFA diet was designed to respect the Canadian Dietetic Association guidelines [21], which promote <25% MUFAs of daily total energy (E). The SFA and MUFA diets both provided ∼35%E as fat, ∼50%E as carbohydrates, and ∼15%E as protein (Table 1). The SFA diet contained 13.4%E as SFAs, 14.3%E as MUFAs, and 4.7%E as PUFAs. Dietary fats and SFAs were mainly provided from lard. The MUFA diet provided 7.1%E as SFAs, 20.6%E as MUFAs, and 5.0%E as PUFAs. Olive oil instead of lard was used to provide a higher MUFA content, mainly as oleic acid (18:1 cis-9). Other foods were fat-free or low-fat in both diets.

TABLE 1.

Composition of the experimental diets

	SFA diet Mean (±SD)	MUFA diet Mean (±SD)
Energy (kcal/kJ)	2665/11,510 ± 389/1628	2689/11,251 ± 333/1393
Alcohol (%)	0.0	0.0
Protein (%)	16.0	16.0
Carbohydrate (%)	49.1	49.1
Lipids (%)	35.0	35.0
Lipids (g)	117.5 ± 0.4	119.1 ± 0.4
SFAs (%)	13.4	7.1
SFAs (g)	44.8 ± 0.8	24.3 ± 2.1
Butyric acid (4:0) (g)	0.3 ± 0.1	0.3 ± 0.1
Caproic acid (6:0) (g)	0.1 ± 0.1	0.1 ± 0.0
Caprylic acid (8:0) (g)	0.2 ± 0.1	0.1 ± 0.1
Capric acid (10:0) (g)	0.4 ± 0.1	0.2 ± 0.1
Lauric acid (12:0) (g)	0.7 ± 0.6	0.5 ± 0.6
Myristic acid (14:0) (g)	2.3 ± 0.3	1.2 ± 0.4
Palmitic acid (16:0) (g)	25.9 ± 1.0	15.8 ± 0.8
Margaric acid (17:0) (g)	0.1 ± 0.0	0.1 ± 0.0
Stearic acid (18:0) (g)	13.8 ± 0.7	4.8 ± 0.7
Arachidic acid (20:0) (g)	0.03 ± 0.01	0.4 ± 0.0
Behenic acid (22:0) (g)	0.01 ± 0.01	0.1 ± 0.0
MUFAs (%)	14.3	20.6
MUFAs (g)	47.8 ± 1.3	69.9 ± 3.6
Myristoleic acid (14:1) (g)	0.03 ± 0.02	0.03 ± 0.02
Palmitoleic acid (16:1) (g)	2.8 ± 0.2	1.6 ± 0.1
Oleic acid (18:1) (g)	43.4 ± 1.2	67.3 ± 3.6
Gadoleic acid (20:1) (g)	0.9 ± 0.1	0.4 ± 0.1
Erucic acid (22:1) (g)	0.1 ± 0.1	0.1 ± 0.1
PUFAs (%)	4.7	5.0
PUFAs (g)	15.8 ± 1.3	17.2 ± 1.4
ω-6 PUFAs (g)	13.8± 1.7	14.7 ± 1.1
ω-3 PUFAs (g)	1.9 ± 0.6	1.9 ± 0.5
PUFA: SFA ratio	0.4	0.7
Trans fatty acids (%)	1.5	1.1
Dietary cholesterol (mg)	299 ± 39	298 ± 13
Fiber (g)	33.3 ± 6.8	33.3 ± 6.8
Sodium (mg)	2850 ± 479	2851 ± 416

Small amounts of flax oil and sunflower oil were used to achieve the same ω-6- and ω-3 PUFA content between the 2 diets. In addition, egg yolk was used in the MUFA diet to match the cholesterol content of the SFA diet. Overall, the main difference between the 2 experimental diets was the substitution of MUFAs for SFAs (6.3%E). Other macro- and micronutrients were matched in both diets. The breakfast meal represented ∼30% of the daily energy intake and the lunch and dinner meals each provided 35% of the daily energy intake. The composition of the diets was assessed using Nutrition Data System for Research software (University of Minnesota). Dietetic technicians prepared all experimental diets throughout the study. Although they were not blinded to treatment allocation, they had no contact with participants. Each food or ingredient was weighed with a precision of ±0.1 g.

Fasting plasma lipoprotein, glucose, and insulin concentration measurements

Twelve-hour fasting blood samples were obtained from an antecubital vein. Serum cholesterol, TG, and glucose concentrations were determined with a Siemens Dimension Vista 1500 using proper reagents. Insulin concentrations were measured with a Siemens Centaur XPT. HOMA-IR index values were calculated using the following formula: [fasting blood glucose (mmol/L) × fasting insulin (mU/L)/22.5] [22].

Lipoprotein subclass analysis

Determination of lipoprotein subclasses was performed by ¹H nuclear magnetic resonance (NMR) spectroscopy on a Bruker IVDr B.I. LISA platform at 600 MHz using 12-h fasting plasma samples as previously described [23].

Kinetic studies

To determine the kinetics of apoB-48- and apoB-100-containing lipoproteins, subjects underwent a primed-constant infusion of l-[5,5,5-d₃] leucine while they were maintained in a constant fed state on the last day of each diet. Subjects received 30 small, identical snacks every half hour for 15 h, each containing one-thirtieth of their estimated daily food intake. The nutritional composition of the snacks provided during the kinetic studies was identical to that of the diet. The snacks of the kinetic study performed at the end of the MUFA diet provided 20.6% of energy as MUFAs and 7.1% as SFAs. The snacks of the kinetic study performed at the end of the SFA diet provided 14.3% of energy as MUFAs and 13.4% as SFAs. Blood samples were collected at predetermined times during the test. ApoB-48 and apoB-100 concentrations were measured by commercial ELISA kits (Fujifilm WAKO Chemicals, R&D Systems Inc.). Sample processing, laboratory measurements, and lipoprotein kinetic analyses were performed by liquid chromatography-tandem mass spectrometry (LC-MS/MS) with multiple reaction monitoring, and the SAAM II program (SAAM Institute) was used to fit the model to the observed tracer data as previously described [24].

Lipidomic profiling

Untargeted LC-MS/MS lipidomic data acquisition was performed from 12-h fasting plasma samples collected at the end of each diet, with an EquiSPLASH LIPIDOMIX (Avanti Polaris) internal standard solution added to each sample, on a Vanquish UPLC equipped with an Accucore C30 column and coupled to a Thermo Fischer Scientific Orbitrap Fusion Mass spectrometer. Lipidomic data were acquired using Xcalibur software version 4.6 from Thermo Fischer Scientific. Chromatographic alignment, peak picking, and deconvolution were done using Progenesis QI software version 3.0 from Nonlinear Dynamics. The internal standard used allowed the identification and normalization of 9 lipid classes. Lipids were identified according to their fragmentation pattern and retention time similarities with their respective internal standard. A total of 484 lipids were identified at class levels and 194 lipids were matched to the LIPID MAPS structure database [25].

Sample size calculations

A power calculation was conducted using the change in LDL cholesterol as the primary outcome based on a previous study from our group [17]. This calculation indicated that a sample size of 12 subjects would enable us to detect a clinically significant difference of 5.7% in LDL cholesterol between the 2 diets with a power of 80% at a 2-sided 5% significance level. The SD of the difference between the 2 values for the same patient for LDL cholesterol used for this calculation was 6.4%. An estimated attrition rate of 20% was considered to determine that 15 subjects had to be recruited.

Statistics

Posttreatment values for biochemical, NMR, and kinetic data are reported in percentage difference (%Δ), and statistical significance analyses were performed with a nonparametric Wilcoxon signed-ranked test using JMP Statistical Software (version 16, SAS Institute). Differences between the 2 diets were considered statistically significant at P ≤ 0.05. For lipidomics, a 1-factor statistical analysis in MetaboAnalyst 5.0 was used to find features of interest and a parametric paired t test was performed with a false discovery rate–corrected P values threshold of ≤0.05 to identify 89 features of interest out of 484 identified lipids. Boxplots reporting the standardized concentrations for 8 lipid classes of interest analyzed with a Wilcoxon signed-ranked test were generated using GraphPad version 9.5.1. The absence of carryover effect was confirmed by comparison of the pretreatment values between each diet using a Wilcoxon signed-ranked test and validated in mixed models for repeated measurements with the diet as the fixed effect and the sequence of the intervention as the random effect for all endpoints.

Results

A flowchart of the study is presented in Figure 1. During the intervention, 3 subjects dropped out of the study for personal and COVID-19-related reasons (busy work schedule n = 1; aversion to foods, n = 1; and moved out of town during lockdowns, n = 1). A fourth subject was excluded before the end of week 12 for health reasons. Baseline demographic and fasting biochemical characteristics of the 14 subjects who completed the study are presented in Table 2. Subjects exhibited features of IR and atherogenic dyslipidemia with elevated HOMA-IR indexes, waist circumferences, fasting TG concentrations, and reduced HDL cholesterol concentrations.

FIGURE 1.

Flowchart of participants.

TABLE 2.

Baseline (week 0) characteristics of the subjects (n = 14)

	Mean (±SD)	Normal range Male/female
Age (y)	41.6 ± 11.0	—
Gender (M/F)	10/4	—
Weight (kg)	92.6 ± 12.3	—
BMI (kg/m2)	31.5 ± 3.3	18.5–24.9
Waist circumference (cm)	104 ± 9.8	—
TC (mmol/L)	5.45 ± 1.24	<5.2
TG (mmol/L)	2.50 ± 1.57	<1.3
HDL-C (mmol/L)	1.04 ± 0.29	>1.3/>1
LDL-C (mmol/L)	3.46 ± 1.06	<3.4
Non-HDL-C (mmol/L)	4.41 ± 1.24	<3.4
TC/HDL-C	5.61 ± 1.99	<5
ApoB (g/L)	1.13 ± 0.23	0.8–1.2
Glucose (mmol/L)	4.98 ± 0.29	4–7
Insulin (pmol/L)	110 ± 40.7	<60
HOMA-IR index	3.51 ± 1.42	0.5–1.4

Abbreviations: apo, apolipoprotein; TC, total cholesterol; TG, triglyceride.

The reported compliance to the experimental diets was very high for both diets (>99%). No differences were observed at the end of the MUFA diet compared with the SFA diet for weight (Δ = −0.1%, P = 0.5), BMI (Δ = 0.7%, P = 0.1), or waist circumference (Δ = −0.2%, P = 0.4).

Table 3 shows that substituting MUFAs for SFAs had no impact on the PR, FCR, and PS of both TRL apoB-48 and intermediate-density lipoprotein (IDL) apoB-100. However, consumption of a MUFA-rich diet lowered VLDL apoB-100 PS (Δ = −22.5%, P = 0.01) compared with the SFA diet but differences in PR (Δ = −15.9%, P = 0.9) and FCR (Δ = +20.5%, P = 0.4) did not reach statistical significance. Furthermore, this substitution resulted in a decrease in LDL apoB-100 PS (Δ = −6.0%, P = 0.05) that was partially attributable to an increase in LDL apoB-100 FCR (Δ = +1.6, P = 0.05) because there was no evidence of a statistical difference in PR (Δ = −8.1% P = 0.2) between the 2 diets. No significant differences between diets were observed in the VLDL-to-IDL, VLDL-to-LDL, and IDL-to-LDL conversion rates.

TABLE 3.

Kinetic parameters of TRL apoB-48 and VLDL, IDL, and LDL apoB-100 of the subjects at the end of each diet

	SFAs Mean ± SD	MUFAs Mean ± SD	%Δ SFAs → MUFAs	P1
TRL apoB-48
PS (mg)	22.7 ± 15.4	23.5 ± 14.8	+3.5	0.6
FCR (pools/d)	8.36 ± 2.61	7.62 ± 4.01	−8.9	0.4
PR (mg/kg/d)	2.14 ± 1.85	2.14 ± 1.93	0.0	0.9
VLDL apoB-100
PS (mg)	106.3 ± 42.6	82.4 ± 24.9	−22.5	0.01
FCR (pools/d)	7.91 ± 2.39	9.53 ± 5.34	+20.5	0.4
PR (mg/kg/d)	9.21 ± 4.62	7.75 ± 2.98	−15.9	0.9
Direct catabolic rate (%)	15.1 ± 16.3	11.4 ± 17.4	−25.0	0.4
IDL apoB-100
PS (mg)	47.2 ± 21.3	45.9 ± 20.2	−2.8	0.9
FCR (pools/d)	6.73 ± 2.68	7.30 ± 2.86	+8.5	0.4
PR (mg/kg/d)	3.43 ± 2.01	3.31 ± 1.25	−3.5	0.5
Direct catabolic rate (%)	99.9 ± 0.0	99.9 ± 0.0	0.0	1.0
LDL apoB-100
PS (mg)	1597 ± 398	1501 ± 497	−6.0	0.05
FCR (pools/d)	0.249 ± 0.169	0.253 ± 0.097	+1.6	0.05
PR (mg/kg/d)	4.18 ± 2.87	3.84 ± 1.31	−8.1	0.2
Conversion rate
VLDL to IDL (%)	39.1 ± 15.6	41.2 ± 14.0	+5.4	0.2
VLDL to LDL (%)	45.8 ± 12.8	47.5 ± 12.8	+3.7	0.6
IDL to LDL (%)	0.08 ± 0.03	0.08 ± 0.04	0.0	0.6

Abbreviations: apo, apolipoprotein; FCR, fractional catabolic rate; IDL, intermediate-density lipoprotein; PR, production rate; PS, pool size; TRL, triglyceride-rich lipoprotein.

¹P values were calculated with a Wilcoxon signed-rank test for repeated measurements.

Standard biochemical analyses were conducted to determine if the changes in LDL apoB-100 kinetics were associated with a variation in circulating molecules known to impact lipoprotein metabolism (Table 4). The significant reduction in fasting concentrations of both LDL cholesterol (Δ = −7.0%, P = 0.01) and non-HDL cholesterol (Δ = −2.5 %, P= 0.04) after the MUFA diet compared with the SFA diet was in line with the detected reductions in LDL apoB-100 PS and VLDL apoB-100 PS (Table 4). No changes were observed in the concentrations of other lipid biomarkers such as TC, HDL cholesterol, TG, apoB, and TC/HDL cholesterol ratio. Lastly, substituting MUFAs for SFAs had no impact on glucose and insulin homeostasis.

TABLE 4.

Fasting biochemical characteristics of the subjects at the end of each diet

	SFA diet Mean ± SD	MUFA diet Mean ± SD	%Δ SFAs → MUFAs	P1
Glucose (mmol/L)	5.05 ± 0.43	5.01 ± 0.44	−0.8	0.6
Insulin (ρmol/L)	115 ± 42.5	116 ± 29.3	+0.2	0.2
HOMA-IR index	3.70 ± 1.23	3.72 ± 3.73	+0.5	0.3
TC (mmol/L)	5.03 ± 1.01	4.93 ± 1.12	−2.0	0.3
TG (mmol/L)	1.99 ± 1.03	2.09 ± 1.06	+5.0	0.6
HDL-C (mmol/L)	1.04 ± 0.24	1.04 ± 0.24	0.0	0.7
LDL-C (mmol/L)	3.15 ± 0.97	2.93 ± 1.02	−7.0	0.01
Non-HDL-C (mmol/L)	3.99 ± 1.07	3.89 ± 1.08	−2.5	0.04
TC/HDL-C	5.11 ± 1.62	4.97 ± 1.55	−2.7	0.2
ApoB, (g/L)	1.05 ± 0.25	1.03 ± 0.27	−1.9	0.4

Abbreviations: apo, apolipoprotein; TC, total cholesterol; TG, triglyceride.

¹P values were calculated with a Wilcoxon signed-rank test for repeated measurements.

Figure 2 shows the results of the NMR lipoprotein subclass analysis. The reduction in LDL particle number (Δ = −7.4% P = 0.001) after the MUFA diet compared with the SFA diet was consistent with the diminutions of LDL cholesterol and LDL apoB-100 PS (Box A). Additionally, this substitution increased the proportion of sdLDL (Δ = +6% P= 0.001) (Box B).

FIGURE 2.

Bar graph representing the concentrations of total LDL particles after each diet (Box A) and the percentage of sdLDL after each diet (Box B). Differences in values after the MUFA diet compared with the SFA diet are expressed in percentages. P values were calculated with a Wilcoxon signed-rank test for repeated measurements. sdLDL, small dense low-density lipoprotein.

Figure 3 presents box plot diagrams with the standardized concentrations of the 8 most prevalent lipid classes identified from fasting plasma samples collected at the end of each diet. The substitution of MUFAs for SFAs increased plasma concentrations of lysophosphatidylcholine (LPC), phosphatidylcholine (PC), diglyceride (DG) and reduced sphingomyelin (SM) and TG levels. Ceramide (Cer), lysophosphatidylethanolamine (LPE), and phosphatidylethanolamine (PE) showed no significant differences between the 2 diets.

FIGURE 3.

Box plot diagram representing the standardized concentrations of the 8 most prevalent lipid classes identified from fasting plasma samples taken at the end of each diet. Concentrations of LPC, PC, and DG were higher after the MUFA diet compared with the SFA diet, whereas the latter presented elevated SM and TG concentrations. Cer, LPE, and PE showed no significant difference between the 2 diets. Boxes contain the second and third quartiles of data with the horizontal line inside them representing the median value; whiskers represent the range. Linked points represent the values of individual participants at the end of each diet. Significant P values ≤ 0.05 are shown for each lipid class. P values were calculated with a Wilcoxon signed-rank test for repeated measurements. Cer: ceramide, DG, diglyceride; LPC: lysophosphatidylcholine. LPE, lysophosphatidylethanolamine; NS, nonsignificant; PC, phosphatidylcholine; PE, phosphatidylethanolamine; SM, sphingomyelin; TG, triglyceride.

Discussion

This study assessed the effects of the substitution of MUFAs from olive oil for SFAs from lard on lipoprotein metabolism in males and females with IR-induced dyslipidemia. Overall, the substitution of MUFAs for SFAs had no impact on TRL apoB-48 kinetics but resulted in a 22.5% reduction in VLDL apoB-100 PS. Furthermore, this substitution decreased non-HDL cholesterol by 2.5% and LDL cholesterol by 7.0% as well as LDL apoB-100 PS by 6.0% with a 1.6% increase in FCR. These changes were accompanied by a 7.4% reduction in total LDL particle number but also a 6% higher proportion of sdLDL. Additionally, each diet presented a distinct lipidomic profile with the MUFA diet showing significantly lower concentrations in plasma TG and SM together with elevated LPC, PC, and DG levels compared with the SFA diet.

This study is, to our knowledge, the first to evaluate the impact of the substitution of MUFAs for SFAs on the metabolism of both apoB-48- and apoB-100-containing TRLs using tracer analyses in IR subjects. Our findings that this dietary substitution had no significant effects on TRL apoB-48 kinetics bring an important insight because previous investigations using other methodological approaches have had variable outcomes [[26], [27], [28]]. In addition, previous kinetic studies in both IR and insulin-sensitive subjects from our group and others also reported mixed results concerning the effects of MUFA consumption on VLDL apoB-100 metabolism [[29], [30], [31]]. In this study, our demonstration that the substitution of dietary MUFAs for SFAs reduces VLDL apoB-100 PS and non-HDL cholesterol levels is of particular interest because, even though we did not observe significant changes in VLDL apoB-100 FCR, it could nonetheless suggest an increase in atherogenic remnant-like particle (RLP) clearance because we also observed lower plasma SM levels that are markers for the clearance of RLP [32]. We also found the LDL cholesterol–reducing effects of MUFA consumption compared with SFAs reported in meta-analyses and other kinetic studies to be the result of a reduction in LDL apoB-100 particle number partially caused by a slightly accelerated LDL apoB-100 turnover [10,30,31,33]. However, the reductions in these apoB-100-containing lipoproteins only resulted in a nonsignificant decrease in total apoB (Δ: −1.9% P = 0.4) that has been previously associated with the substitution of MUFAs for SFAs [33]. This could be because of the limited statistical power of our sample size together with the absence of impact on TRL apoB-48 and a slightly higher proportion of atherogenic sdLDL particles, which is also a known effect of this substitution [12,34,35]. Nonetheless, changes in sdLDL particle number were very small, and current evidence shows that larger LDL particles are not benign, making the overall impact of this dietary fatty acid substitution on serum cholesterol potentially beneficial for CVD prevention in IR subjects [36]. In addition, although nonsignificant in this study, a decrease in LDL apoB-100 PR resulting from the reduction in VLDL apoB-100 PS may also have contributed to these effects as we have previously shown that the substitution of ω-6 PUFAs for SFAs in IR subjects reduces LDL cholesterol levels by this mechanism, albeit with no effect on LDL apoB-100 FCR [17]. However, other compensatory mechanisms could be involved because PUFAs but not MUFAs modulate key transcription factors involved in lipid and lipoprotein metabolism such as sterol regulatory element-binding proteins, peroxisome proliferator-activated receptors, and liver X receptors [37,38].

An increase in serum TG is a hallmark of IR but lipidome-wide association studies have found that individuals with IR also present elevations in Cer and SM together with a decrease in LPC and PC [[39], [40], [41], [42], [43]]. Furthermore, this blood lipid profile is associated with an increased CVD risk because plasma SM concentrations are positively and independently associated with these diseases, making them potential biomarkers [[44], [45], [46], [47], [48]]. In this regard, our plasma lipidomic profiling showed that the substitution of dietary MUFAs for SFAs resulted in an improved circulating lipid profile with significantly lower TG and SM together with elevated LPC and PC as reported in previous studies [49,50]. However, this substitution did not result in significant improvements in clinical biomarkers of IR that also mirrored the findings of recent systematic reviews and meta-analyses [51,52].

In addition, several studies, both in vitro and in vivo, also reported that SM inhibits the LDL receptor (LDLR) by interfering with apoB and apoE binding as well as reducing the activity of the lipoprotein lipase (LPL) and lecithin–cholesterol acyltransferase (LCAT) enzymes [[53], [54], [55], [56], [57], [58], [59]]. In cell cultures, the addition of sphingomyelinase resulted in a 2- to 5-fold increase in LDL uptake and degradation together with an increase in VLDL-TG hydrolysis and HDL cholesterol esterification [[54], [55], [56], [57]]. In apoE-KO mice, serine palmitoyltransferase (SPT) inhibitor myriocin decreased both hepatic and plasma SM, resulting in higher HDL cholesterol along with reduced VLDL-TG, LDL cholesterol, and atherosclerotic lesions [59]. So far, the mechanisms behind the circulatory SM-reducing effects of substituting MUFAs for SFAs remain unclear, although animal studies have shown that a high-SFA diet induces SPT activity, resulting in elevated Cer levels compared with both a high-PUFA diet and a low-fat diet [60]. Consequently, the decrease in plasma SM observed after the MUFA diet compared with the SFA diet could be the result of the differential effects of this substitution on the various enzymes involved in their synthesis together with the reduced availability of palmitic acid that is the SPT’s preferred substrate in the rate-limiting step of the sphingolipid de novo synthesis pathway [61].

Accordingly, the increase in LDL apoB-100 FCR along with the decrease in VLDL and LDL apoB-100 PS observed in the MUFA diet compared with the SFA diet observed in this study could be the outcome of an increase in LDLR activity resulting from reduced plasma SM levels (Figure 4). Furthermore, this elevated LDLR activity could also explain the increased proportion of sdLDL after the MUFA diet because these subfractions have a decreased affinity for the LDLR, thus favoring the elimination of large buoyant LDL particles [62]. Moreover, an increase in VLDL-TG hydrolysis induced by enhanced LPL activity could explain the circulating TG reduction observed in the MUFA diet. In addition, because PC is a phosphocholine group donor in SM synthesis, a putative decrease in SM synthase activity caused by substituting MUFAs for SFAs could potentially negate the PC depletion observed in IR and, together with an increase in LCAT activity, could result in more PC being degraded into LPC [63]. Although these interpretations are hypothetical at this point, these findings are of great interest because they not only strengthen our knowledge of the mechanisms behind the cholesterol-lowering effects of substituting dietary MUFAs for SFAs in subjects with IR-induced dyslipidemia but also provide new research avenues because modulating SM levels through dietary interventions could prove to be an effective tool in managing these disorders. Therefore, these effects merit further investigation, especially by examination of the lipidomic profile resulting from other dietary fatty acid substitutions like PUFAs for SFAs, or different food matrices such as novel high-oleic canola oil because it could lead to targeted dietary recommendations for the prevention of CVD in high-risk populations such as those with dyslipidemia associated with IR.

FIGURE 4.

Graphical summary of the proposed mechanism explaining the cholesterol-lowering effects of substituting MUFAs for SFAs. Reduced SFA intakes downregulate sphingolipid hepatic de novo synthesis by limiting palmitate availability, resulting in lower intracellular SM and DG together with elevated PC. The ensuing reduced SM:PC ratio in circulating apoB-100-containing lipoproteins enhances LDLR activity that increases LDL uptake resulting in lowered LDL-C and non-HDL-C. LPL activity is also increased, resulting in lower circulating TG. Reduced SM content in HDL increases LCAT activity, therefore elevating circulating LPC, which is the other product of HDL-C esterification. apo, apolipoprotein; DG, diglyceride; FCR, fractional catabolic rate; LPC, lysophosphatidylcholine; PC, phosphatidylcholine; PS, pool size; SM, sphingomyelin; SMase, sphingomyelinase; SMS, sphingomyelin synthase; SPT, serine palmitoyl transferase; TG, triglyceride.

The use of fully controlled diets in a crossover design and a subject sample of both males and females with IR-associated dyslipidemia are major strengths of this dietary intervention that enabled us to evaluate the effects of the substitution MUFAs for SFAs as recommended in dietary guidelines. Additionally, the combination of lipoprotein kinetic studies, NMR lipoprotein subclass analysis, and LC/MS/MS lipidomic profiling provides a unique appreciation of the mechanistic explanations behind the observed effects on blood lipid homeostasis. However, these results only reflect the short-term effects of a MUFA–SFA substitution, and the relatively small sample size may not have allowed sufficient statistical power to detect changes in some lipoprotein kinetic parameters and cardiometabolic risk factors. Finally, the fact that only a small number of participants were females and that all were IR subjects from Quebec may limit the generalizability of our results for other populations.

In conclusion, our study showed that the substitution of dietary MUFAs from olive oil for SFAs from lard in subjects with dyslipidemia associated with IR had no impact on TRL apoB-48 kinetics but, decreased VLDL apoB-100 PS, and increased LDL apoB-100 FCR that resulted in reduced non-HDL cholesterol, LDL cholesterol, and total LDL particle number. In addition, reduced plasma SM along with elevated LPC and PC levels after the MUFA diet suggest that an increase in LDLR, LPL, and LCAT activities may partly underly these changes that represent an improvement to the proatherogenic lipid profile found in IR subjects. Overall, this study represents a step forward in our ability to understand the underlying mechanisms behind the cardioprotective effects of substituting MUFAs for SFAs, further supporting current dietary guidelines for managing CVD risk in individuals with IR-induced dyslipidemia.

Author contributions

The authors’ responsibilities were as follows – PC, BL, J-PD-C: designed the study; AJT, MR-B, AC, VL: conducted the research; L-CD, FB, JC, AJT, EJS, BL, PC: analyzed the data; L-CD, AJT, BL, PC: wrote the manuscript; PC: had the primary responsibility for the final content of the manuscript; and all authors: read and approved the final version.

Conflict of interest

The authors report no conflicts of interest.

Funding

This study was funded by the Canadian Institutes of Health Research (CIHR, NUT-408430). The funder had no role to play in the design of the study and in the interpretation of the results. L-CD is funded by a graduate studentship from the CIHR. J-PD-C is a research scholar of the Fonds de recherche du Québec – Santé. BL has received funding from the Canadian Institutes for Health Research (on going), the Fonds de recherche du Québec – Santé (FRQS) (on going), Fonds de recherche du Québec – Nature et technologies (NT) (on going), the Ministère de la santé et des services sociaux (MSSS) du Québec (on going), and Health Canada (completed in 2022). PC has received funding from the Canadian Institutes for Health Research (on going).

Data availability

Data described in the manuscript, code book, and analytic code will be made available upon request pending approval by our ethical review board.