The Effects of High-Protein and High-Monounsaturated Fat Meals on Postprandial Lipids,Lipoprotein Particle Numbers,Cytokines,and Leptin Responses in Overweight/Obese Subjects

Abstract

Background:

Obesity is linked to dyslipidemia, proinflammatory state, and hyperleptinemia. The influence of high-protein (HP) versus high-monounsaturated fat (HMF) meals on postprandial lipids, lipoprotein particle numbers, cytokines, and leptin responses in overweight/obese (OW/O) subjects is unknown.

Methods:

Twenty-four OW/O participants consumed an HP (31.9% energy from protein) and HMF (35.2% fat and 20.7% monounsaturated fat) meal, of similar energy/carbohydrate content, in a random order. The outcome variables were assessed from blood samples collected in fasted and postprandial (3 hr) states.

Results:

Repeated measures analysis found significant (P < 0.05) meal condition by time interactions for triglycerides (TGs), very low-density lipoprotein particles (VLDLP), total high-density lipoprotein particles (T-HDLP), and the ratio of large-buoyant high-density lipoprotein 2b (LB-HDL2b) to T-HDLP, and meal effect on small-dense HDLP (SD-HDLP). Comparison of HP versus HMF condition showed significantly lower TG at 120 min [geometric mean (95% confidence interval, CI): 148 (125–175) vs. 194 (164–230) mg/dL] and 180 min [167 (138–203) vs. 230 (189–278) mg/dL] and VLDLP at 180 min [70.0 (58.2–84.3) vs. 88.0 (73.1–106) nmol/L]. HP versus HMF condition showed significantly lower LB-HDL2b/T-HDLP at 180 min [mean difference (95% CI): 0.021 (0.004–0.038)], and higher T-HDLP [671 (263–1079) nmol/L] and SD-HDLP [606 (292–920) nmol/L] at 120 min. Area under the curve was significantly lower for TG and higher for T-HDLP, SD-HDLP, and small-dense LDL III (SD-LDL III) in the HP condition. Cytokines and leptin were not different between conditions.

Conclusion:

OW/O subjects had lower TG and VLDLP, but less favorable SD-LDL III, SD-HDLP, and LB-HDL2b/T-HDLP ratio responses to the HP versus HMF meals.

Introduction

Nearly 40% of American adults are obese.¹ A major co-morbidity associated with excess body fat is dyslipidemia.² This condition is characterized by elevated total cholesterol (TC), triglyceride (TG), or low-density lipoprotein cholesterol (LDLC), or decreased high-density lipoprotein cholesterol (HDLC) concentrations.² In addition, the obesity-associated dyslipidemia is characterized by lipoprotein subfractions (determined by particle density and size), including small-dense LDLC, small-dense HDLC (SD-HDLC), and very low-density lipoprotein (VLDL) remnants.²

Obesity is also associated with increased levels of proinflammatory molecules, including monocyte chemoattractant protein-1 (MCP-1) and tumor necrosis factor (TNF)-α.³ Both dyslipidemia and proinflammatory molecules are linked to an increased risk for cardiovascular disease (CVD).^2,3 Obesity is also associated with high leptin concentrations, possibly because of leptin resistance.³ Although leptin regulates blood glucose and food intake and increases energy expenditure, it may also promote a proinflammatory state in obese subjects.³

The composition of the diet may have an impact on lipids and lipoproteins. According to two meta-analyses, replacing high-carbohydrate (HC) diets with high-monounsaturated fat (HMF) diets reduces fasting plasma TG and may increase HDLC in patients with type 2 diabetes.^4,5 Replacing HC diets with high-protein (HP) diets may also improve certain lipids. A meta-analysis by Wycherley et al., found that energy-restricted, isocaloric, HP, low-fat diets compared with standard protein, low-fat, higher carbohydrate diets led to lower TG in subjects with excess weight, type-2 diabetes, hyperinsulinemia, polycystic ovarian disease, and/or metabolic syndrome.⁶ These results indicate that it may be prudent to explore the lipid and lipoprotein responses to HMF and HP diets as an alternative to HC diets.

There are limited data comparing the effects of HP and HMF intakes on blood lipids and lipoprotein particle numbers in overweight/obese (OW/O) subjects. Appel et al.⁷ found a reduction in fasting TC, TG, and HDLC, and no difference in LDLC or non-HDLC on HP versus HMF diets, of similar carbohydrate content, in subjects with prehypertension or stage 1 hypertension. Luscombe-Marsh et al.,⁸ however, did not find a difference in lipid concentrations in OW/O subjects randomized to carbohydrate-restricted HP or HMF diets. These studies did not assess lipoprotein particle numbers, however. This is important to determine, given that obesity-associated dyslipidemia is characterized by a number of lipoprotein subfractions which may determine the risk for CVD. Small-dense LDLC and HDL particles, remnant lipoproteins, and very low-density lipoprotein particles (VLDLP) may increase the risk for CVD, whereas large HDL particles may reduce the risk for CVD.^9
–11 More recently, another study evaluated the effects of HP and HMF meals on postprandial lipids and lipoprotein particle numbers, and found that an HP meal decreased TG, non-high-density lipoprotein particles (HDLP), and VLDLP, but reduced large-buoyant high-density lipoprotein 2b (LB-HDL2b) particles, and increased SD-HDLP compared to an HMF meal.¹² This study, however, was conducted in normal-weight participants.

Very few studies have compared the effect of HP and HMF meals on postprandial TNF-α, leptin, and MCP-1 responses. Parvaresh Rizi et al.¹³ found no meal effect on TNF-α in obese and lean subjects in response to HC, HP, and HMF meals. Raben et al.¹⁴ found no difference in leptin responses to an HP versus an HMF meal of similar energy density and fiber content in normal-weight subjects. Holmer-Jensen et al.¹⁵ reported suppression in MCP-1 following consumption of high-fat meals supplemented with whey, cod protein, gluten, or casein in obese subjects.

The objective of this study was to compare the effects of HP and HMF meals on postprandial lipids, lipoprotein particle numbers, cytokines, and leptin responses in OW/O participants.

Materials and Methods

Participants

Twenty-four OW/O volunteers, aged 18–65 years, were recruited for the study. Data on the effect of an HMF versus an HP meal on hormones that control blood glucose in these participants have been published earlier.¹⁶ The OW/O subjects met two of the following criteria: >25% body fat in men and >35% body fat in women, >102 cm waist circumference in men (≥90 cm in Asian men) and >88 cm in women (≥80 cm in Asian women), and ≥25 kg/m² (≥ 23 kg/m² in Asians) body mass index (BMI).^17,18 Exclusion criteria included trying to lose weight, using anti-hyperglycemic agents, having liver, kidney, untreated thyroid, malabsorption, or eating disorders, smoking, drinking >1–2 drinks/d, or pregnancy/lactation.

Each subject signed an informed consent approved by the Institutional Review Board before enrolling in the study. The study was performed according to the Declaration of Helsinki principles for research regarding human subjects.

Experimental design

Each subject was fed an HP and an HMF meal, in a random order, determined by a blocked randomization schema, separated by ≥4 days. Blood samples were collected after a 12-hr overnight fast and for 3 hr during the postprandial state.

Test meals

The test meals were in a beverage form and contained the same energy, added sugar, and volume. Men received 840 kcal and women received 700 kcal. This was about 35% of their energy needs. The HP meal (31.9% energy from protein, 15.5% from total fat, 4.3% from monounsaturated fat, 9.9% from saturated fat, and 52.6% from carbohydrates) was prepared using nonfat Greek yogurt, plain whole milk yogurt, and sugar. The HMF meal (35.2% energy from total fat, 20.7% from monounsaturated fat, 12.6% from protein, 8.7% from saturated fat, and 52.3% from carbohydrates) was prepared using plain low-fat yogurt, avocado, and sugar.

Study protocol

Each subject came to our laboratory, at the same time on two different days, after an overnight fast. Females of childbearing age were scheduled during the follicular stage of their cycle. All subjects were instructed to consume the meals in 20 min and to take the same time in finishing both meals. A venous catheter was inserted in the antecubital vein. Blood samples were collected in the supine position, in the fasting state, and at 30, 60, 120, and 180 min from when the meal began. The samples were centrifuged and plasma was stored at −80°C. Foods other than the test meal were not allowed and water intake was controlled during the postprandial period. Energy intake and exercise levels on the day before the study days were controlled with no difference between those days. Body weight was similar on both study days. Data on energy intake, exercise duration, and body weight are given in section “Results” and in Table 1.

Table 1.

Baseline Data on Lipids, Lipoprotein Particle Numbers, Leptin, Cytokines, Body Weight, Body Mass Index, and Twenty-Four-Hour Recall on Energy Intake and Exercise Duration by Meal Condition

Variable	HP condition (n = 24)	HMF condition (n = 24)	P^a
TG (mg/dL)	116 (85–149)	122 (97–167)	0.70
TC (mg/dL)	177 ± 42	177 ± 38	0.98
LDLC (mg/dL)	106 ± 31	107 ± 30	0.90
HDLC (mg/dL)	42.5 ± 11.7	42.6 ± 12.0	0.98
VLDLP (nmol/L)	62.5 (41.5, 100.5)	61.0 (49.5–80.5)	0.72
RLP (nmol/L)	138 ± 64	133 ± 49	0.69
Non-HDLP (nmol/L)	934 ± 217	882 ± 163	0.17
T-LDLP (nmol/L)	854 ± 191	811 ± 155	0.20
SD-LDL III (nmol/L)	315 ± 122	287 ± 88	0.16
SD-LDL IV (nmol/L)	104 ± 24	100 ± 19	0.35
T-HDLP (nmol/L)	7796 ± 1155	7138 ± 1085	0.01
LB-HDL2b (nmol/L)	2058 ± 611	1857 ± 580	0.06
SD-HDLP (nmol/L)	5737 ± 810	5281 ± 661	0.01
LB-HDL 2b/T-HDLP ratio	0.261 ± 0.005	0.257 ± 0.046	0.56
Leptin (pg/mL)	11853 (5600–23,773)	12787 (4625–25,980)	0.65
TNF-α (pg/mL)	2.8 ± 1.3	2.9 ± 1.5	0.93
MCP-1 (pg/mL)	82.6 ± 35.1	77.0 ± 29.9	0.49
Body weight (kg)	91.8 ± 14.8	91.7 ± 15.5	0.71
Body mass index (kg/m²)	31.6 ± 3.9	31.6 ± 4.1	0.53
EI over past 24 hr (kcal)	1941 ± 693	1973 ± 668	0.62
ED over past 24 hr (minutes)	0 (0, 0)	0 (0, 0)	1.0

P values indicate differences by meal condition and were determined by Wilcoxon signed-rank test for ED and paired t-test for the remaining variables. The data on TG, VLDLP, and leptin were log-transformed before analysis.

Data on TG, VLDLP, leptin, and ED are shown as medians and 25th and 75th percentiles and data on the remaining variables are shown as mean and standard deviation.

HP, high protein; HMF, high-monounsaturated fat; TG, triglyceride; TC, total cholesterol; LDLC, low-density lipoprotein cholesterol; HDLC, high-density lipoprotein cholesterol; VLDLP, very low-density lipoprotein particles; RLP, remnant lipoprotein particles; non-HDLP, non-high-density lipoprotein particles; T-LDLP, total low-density lipoprotein particles; SD-LDL III, small-dense low-density lipoprotein III; SD-LDL IV, small-dense low-density lipoprotein IV; T-HDLP, total high-density lipoprotein particles; LB-HDL2b, large-buoyant high-density lipoprotein 2b; SD-HDLP, small-dense high-density lipoprotein particles; TNF-α, tumor necrosis factor alpha; MCP-1, monocyte chemoattractant protein-1; EI, energy intake; ED, exercise duration.

Measures

Demographics and anthropometry

Demographic data were collected by questionnaire. BMI (kg/m²) was computed from measured weight and height. Percent body fat was measured by dual-energy x-ray absorptiometry. Waist circumference, at the level of the umbilicus, was measured.

Lipids and lipoprotein particle numbers determination

TC, HDLC, and TG were measured by LabCorp, a Clinical Laboratory Improvement Amendments (CLIA) certified laboratory. LDLC was computed using the Friedwald equation.¹⁹ Lipoprotein particle numbers were determined by SpectraCell Laboratories (Houston, TX), a CLIA-certified laboratory. The particle numbers were determined by a continuous gradient produced by ultracentrifugation as described elsewhere,²⁰ with a coefficient of variation (CV) of 5%. The particles were defined by density and included VLDLP (density: 1.0–1.006 g/mL), remnant lipoprotein particles (RLP; 1.004–1.019 g/mL), total low-density lipoprotein particles (T-LDLP; 1.006–1.063 g/mL), non-HDLP (1.000–1.063 g/mL), small-dense low-density lipoprotein III (SD-LDL III; 1.034–1.044 g/mL) particles, small-dense LDL IV (SD-LDL IV; 1.044–1.063 g/mL) particles, total high-density lipoprotein particles (T-HDLP; 1.063–1.2 g/mL), LB-HDL2b (1.063–1.1 g/mL) particles, SD-HDLP (1.1–1.2 g/mL), and LB-HDL2b/T-HDLP ratio.

Leptin, TNF-α, and MCP-1 were assessed using the MILLIPLEX^® MAP Human Metabolic Hormone Magnetic Bead Panel kits (EMD Millipore Corporation, Billerica, MA). The intra-assay CV were 5.7%, 6.5%, and 8.0%, respectively.

Power calculation and statistical analysis

A sample size of 24 participants provided 88% power at α = 0.05 to detect the mean difference of 24 mg/dL in serum TGs with a standard deviation of the difference of 36 mg/dL. The mean difference and standard deviation correspond to the postprandial data collected at 60 and 120 min in a previous study, with a similar design comparing the effect of an HMF compared to an HP meal on lipids and lipoproteins.¹²

A paired t-test was used to compare baseline data on lipids, lipoprotein particle numbers, leptin, cytokines, body weight, BMI, and energy intake from 24-hr recall by meal condition. Data on TG, VLDLP, and leptin were log-transformed before analysis. Exercise duration over the past 24 hr was compared between conditions using the Wilcoxon signed-rank test.

A mixed-effects repeated measures analysis model was used to evaluate the impact of meal condition, time, and the interaction between meal condition and time on lipids, lipoprotein particle numbers, TNF-α, MCP-1, and leptin concentrations. Least-square mean contrasts were computed to determine differences in outcome variables by meal condition and time. Variables that were not normally distributed (TG, VLDLP, and leptin) were log-transformed before analyses and presented in the figures as geometric means and 95% confidence intervals (CIs). The remaining variables were normally distributed and presented in the figures as arithmetic means and 95% CI. A significant model meal condition effect or significant interaction between meal condition and time indicated statistically different lipid responses between the meal conditions.

The area under the curve (AUC) was also computed for each outcome variable and compared by meal condition using repeated measures analysis. Variables that were not normally distributed (TGs, VLDLP, and leptin) were log-transformed before analyses. The AUC data are shown in the figures as medians, 25th and 75th percentiles, and 10th and 90th percentiles.

The meal sequence was assessed as a factor in the mixed-effects models and no meal sequence effects on the outcome variables were observed. The results were not given by gender because when controlling for sex in our model and when assessing interactions between sex and meal condition, we did not observe any difference in lipid or lipoprotein responses between men and women. Three participants used medications for dyslipidemia and two participants used medications for hypertension. They were retained in the analysis because the medication type/dose remained constant during both meal conditions, and a sensitivity analysis without these subjects did not affect the results. None of the participants had type 2 diabetes. The data were analyzed using SAS software, version 9.4 (SAS Institute, Cary, NC). A P-value <0.05 was considered statistically significant.

Results

Participant characteristics

The study included 12 females and 12 males. Racial distribution was 79.2% white, 12.5% Asians, 4.2% black, and 4.2% mixed race. The mean ± standard deviation was 38.7 ± 15.3 years for age, 31.6 ± 4.0 kg/m² for BMI, 47.7 ± 4.5% (females) and 33.2 ± 5.9% (males) for percent body fat, and 106.8 ±9.8 cm (females) and 106.9 ± 11.3 cm (males) for waist circumference.

Baseline data on lipids, lipoprotein particle numbers, leptin, cytokines, body weight, BMI, and 24-hr recall on energy intake and exercise duration by meal condition are presented in Table 1. There was no difference in any of these variables, except T-HDLP and SD-HDLP, which were significantly different at baseline.

Lipids

The lipid responses to meal condition are shown in Fig. 1.

FIG. 1.

Postprandial TG (A), TC (B), LDLC (C), and HDLC (D) responses to an HP and an HMF meal in 24 overweight/obese participants. TG response is shown as geometric means and 95% confidence intervals, whereas the remaining variable responses are shown as arithmetic means and 95% confidence intervals in the line graphs. The box plots show AUC. The line within the box is median, the lower and upper limits of the box are 25th and 75th percentiles, and the error bars are 10th and 90th percentiles. TG was log-transformed before being analyzed. Mixed-model repeated measures analysis revealed significant meal condition by time interactions for TG (P < 0.0001) and LDLC (P < 0.0001), and time effect (P < 0.0001) for HDLC. Repeated measures analysis showed that AUC for TG was significantly higher in the HMF versus HP meal condition (P = 0.003). HP versus HMF comparisons: **P < 0.01. Compared to time 0 within each condition: ^# P < 0.05. TG, triglyceride; TC, total cholesterol; LDLC, low-density lipoprotein cholesterol; HDLC, high-density lipoprotein cholesterol; HP, high-protein; HMF, high-monounsaturated fat; AUC, area under the curve.

Triglycerides

The study found a significant meal condition by time interaction (P < 0.0001) effect on TG. The responses were higher at 120 (P = 0.006) and 180 (P = 0.003) min during the HMF compared to the HP meal condition. The TG responses were higher at 60, 120, and 180 min during both conditions compared to the respective baseline values (P < 0.001). TG AUC was also higher during the HMF versus HP condition (P = 0.003).

Total cholesterol

No changes were observed in TC responses to the two meals.

Low-density lipoprotein cholesterol

A significant meal condition by time interaction (P < 0.0001) effect was found on LDLC. There was a tendency for LDLC to be higher at 180 min on the HP versus HMF condition (P = 0.07). Compared to the respective baseline values, LDLC responses were significantly lower at 60, 120, and 180 min in the HMF condition and at 60 and 180 min in the HP condition (P < 0.05). LDLC AUC was not different by condition.

High-density lipoprotein cholesterol

No meal condition by time interaction or meal effects were seen on HDLC. Time effect was significant (P < 0.0001). HDLC responses were significantly lower at 120 and 180 min in both conditions compared to the respective baseline values (P < 0.05). AUC was not different by condition.

Lipoprotein particle numbers

The VLDLP, non-HDLP, T-LDLP, RLP, SD-LDL III, and SD-LDL IV responses to meal condition are shown in Fig. 2.

FIG. 2.

Postprandial VLDLP (A), non-HDLP (B), T-LDLP (C), RLP (D), SD-LDL III particle (E), and SD-LDL IV particle (F) responses to an HP and an HMF meal in 24 overweight/obese participants. VLDLP response is shown as geometric means and 95% confidence intervals, whereas the remaining variable responses are shown as arithmetic means and 95% confidence intervals in the line graphs. The box plots show AUC. The line within the box is median, the lower and upper limits of the box are 25th and 75th percentiles, and the error bars are 10th and 90th percentiles. VLDLP was log-transformed before being analyzed. Mixed-model repeated measures analysis revealed significant meal condition by time interactions for VLDLP (P = 0.004), non-HDLP (P = 0.0001), and TLDLP (P = 0.0009). AUC was significantly higher during the HP versus HMF condition for SD-LDL III (P = 0.04). HP versus HMF comparisons: *P < 0.05. Compared to time 0 within each condition: ^# P < 0.05. VLDLP, very low-density lipoprotein particles; non-HDLP, non-high-density lipoprotein particles; T-LDLP, total low-density lipoprotein particles; RLP, remnant lipoprotein particles; SD-LDL III, small-dense, low-density lipoprotein III; AUC, area under the curve.

Very low-density lipoprotein particles

There was a significant meal condition by time interaction (P = 0.004) effect on VLDLP. VLDLP was higher on the HMF versus HP condition at 180 min (P = 0.02). Compared to the respective baseline values, VLDLP responses were significantly (P < 0.05) higher at 120 and 180 min in the HMF condition and at 120 min in the HP condition. VLDLP AUC was not different by meal condition.

Non-HDLP and T-LDLP

There was a significant meal condition by time interaction effect on non-HDLP (P = 0.0001) and T-LDLP (P = 0.0009). Compared to the respective baseline values, TLDLP and non-HDLP were significantly (P < 0.05) lower at 180 min in the HP condition and did not change in the HMF condition. AUC for non-HDLP (P = 0.09) and T-LDLP (P = 0.07) tended to be lower in the HMF versus HP condition.

RLP, SD-LDL III, and SD-LDL IV

There were no meal condition by time interactions, meal condition effects, or time effects on RLP, SD-LDL III, or SD-LDL IV. AUC was significantly higher during the HP versus HMF condition for SD-LDL III (P = 0.04), but was not different for RLP and SD-LDL IV.

T-HDLP, LB-HDL2b, SD-HDLP, and LB-HDL2b/T-HDLP ratio responses to meal condition are shown in Fig. 3.

FIG. 3.

Postprandial T-HDLP (A), LB-HDL 2b particles (B), SD-HDLP (C), and LB-HDL2b/T-HDLP ratio (D) responses to an HP and an HMF meal in 24 overweight/obese participants. The data are shown as arithmetic means and 95% confidence intervals in the line graphs. The box plots show AUC. The line within the box is median, the lower and upper limits of the box are 25th and 75th percentiles, and the error bars are 10th and 90th percentiles. Mixed-model repeated measures analysis revealed significant meal condition by time interactions for T-HDLP (P = 0.02), LB-HDL2b (P < 0.0001), and LB-HDL2b/T-HDLP (P < 0.0001), meal condition effect for SD-HDLP (P = 0.001), and time effect for SD-HDLP (P < 0.0001). Repeated measures analysis showed that AUC for T-HDLP (P = 0.002) and SD-HDLP (P = 0.0006) were significantly higher on the HP versus HMF meal condition. HP versus HMF comparisons: *P < 0.05; **P < 0.01; ***P < 0.001. Compared to time 0 within each condition: ^# P < 0.05. T-HDLP, total high-density lipoprotein particles; LB-HDL2b, large-buoyant, high-density lipoprotein 2b; SD-HDLP, small-dense high-density lipoprotein particles; AUC, area under the curve.

Total high-density lipoprotein particles

There was a significant meal condition by time interaction (P = 0.02) effect on T-HDLP. T-HDLP responses were significantly higher at baseline (P = 0.01) and 120 (P = 0.002) min in the HP compared to the HMF condition. T-HDLP responses were significantly higher during both conditions at 120 and 180 min compared to the respective baseline values (P < 0.05). AUC for T-HDLP was significantly higher on the HP versus HMF meal condition (P = 0.002).

Large-buoyant high-density lipoprotein 2b

A significant meal condition by time (P < 0.0001) effect was found on LB-HDL2b. It was higher (P = 0.0005) at 180 min compared to the baseline in the HMF condition. AUC for LB-HDL2b was not different by meal condition.

Small-dense high-density lipoprotein particles

Significant meal condition (P = 0.001) and time (P < 0.0001) effects, but no meal condition by time interaction effect, were found on SD-HDLP. The responses were significantly higher at baseline (P = 0.014) and at 120 (P = 0.0003) min and tended to be higher at 180 min (P = 0.098) in the HP compared to the HMF condition. SD-HDLP responses were significantly higher in both conditions at 120 and 180 min compared to the respective baseline values (P < 0.01). AUC was significantly higher on the HP versus HMF meal condition for SD-HDLP (P = 0.0006).

LB-HDL2b/T-HDLP ratio

A significant meal condition by time (P < 0.0001) effect was found on LB-HDL2b/T-HDLP ratio. LB-HDL2b/T-HDLP ratio was significantly higher at 180 min on the HMF compared to the HP condition (P = 0.017). Compared to the respective baseline values, LB-HDL2b/T-HDLP ratios were significantly lower at 120 min in the HMF condition and at 120 and 180 min in the HP condition (P < 0.001). LB-HDL2b/T-HDLP AUC tended to be higher in the HMF versus HP condition (P = 0.08).

Leptin, TNF-α, and MCP-1

Leptin, TNF-α, and MCP-1 responses to meal condition are presented in Fig. 4. A significant meal condition by time interaction (P = 0.02) effect was seen for leptin. Compared to the corresponding baseline values, leptin concentrations were significantly (P < 0.01) lower at 60, 120, and 180 min in the HMF condition and at 30, 60, 120, and 180 min in the HP condition. There were no meal condition by time interaction effects or meal condition effects, but significant time effects on TNF-α (P = 0.01) and MCP-1 (P = 0.003). MCP-1 was significantly (P = 0.02) higher at 120 min compared to baseline value during the HMF condition. The AUC values for leptin, TNF-α, and MCP-1 were not different by meal condition.

FIG. 4.

Postprandial leptin (A), TNF-α (B), and MCP-1 (C) responses to an HP and an HMF meal in 24 overweight/obese participants. The data are shown as geometric means and 95% confidence intervals for leptin and arithmetic means and 95% confidence intervals for TNF-α and MCP-1 in the line graphs. The box plots show AUC. The line within the box is median, the lower and upper limits of the box are 25th and 75th percentiles, and the error bars are 10th and 90th percentiles. Leptin was log-transformed before being analyzed. Mixed-model repeated measures analysis revealed significant meal condition by time interaction for leptin (P = 0.02), and time effects for TNF-α (P = 0.01) and MCP-1 (P = 0.003). Repeated measures analysis showed that AUC was not different by meal condition for any of the variables. Compared to time 0 within each condition: ^# P < 0.05. TNF-α, tumor necrosis factor-alpha; MCP-1, monocyte chemoattractant protein-1; AUC, area under the curve.

Discussion

This is the first study to examine the effects of HP versus HMF meals on postprandial lipoprotein particle numbers in obese subjects. Previous studies had limited the comparisons of lipoprotein particle responses to HP and HMF meals to normal-weight subjects or compared the effects of HP and HMF diets on lipids, but not lipoprotein particle numbers.

TG response was lower during the HP versus HMF meal condition. No significant differences in TC, LDLC, and HDLC were observed between the two meal conditions, however. A lower TG response, but no differences in TC, LDLC, and HDLC, has also been observed by Shah et al.¹² in a postprandial study comparing carbohydrate-controlled HP versus HMF meals in lean subjects. Appel et al.⁷ found lower fasting HDLC, TG, and TC concentrations and no difference in LDLC response following HP versus HMF diets with similar carbohydrate content in subjects with prehypertension or stage 1 hypertension. Luscombe-Marsh et al.,⁸ however, reported no differences in improvement in lipid concentrations between carbohydrate-restricted HP and carbohydrate-restricted HMF diets given to OW/O subjects. The HP meal in this study may have lowered TG through reduced chylomicron production (because of lower fat consumption), faster chylomicron clearance by increasing lipoprotein lipase stimulation, decreased lipid synthesis, and enhanced hepatic lipid oxidation to generate energy for amino acid catabolism.²¹ Furthermore, glucagon-like peptide-1, a gut hormone known to decrease hepatic lipid synthesis, was reported to be higher following an HP versus an HMF meal.^16,21

VLDLP response was lower in the HP versus the HMF condition at 180 min. As expected, the VLDLP response was similar to the TG response because VLDLP contain mostly TG. Spearman's correlation test showed that the AUC for VLDLP and TG was strongly correlated (HP condition: r = 0.93; P < 0.0001; HMF condition: r = 0.76; P < 0.0001). A lower postprandial VLDLP response to HP versus HMF meals has also been observed by Shah et al.¹² in lean subjects. Maki et al. reported that replacing 16% of the energy from refined carbohydrates and added sugars with a combination of egg protein powder and vegetable oil reduced TG and VLDLP compared to the baseline values in OW/O subjects with hypertriglyceridemia.²²

SD-LDL III AUC was higher in the HP versus HMF condition, whereas no difference was seen for SD-LDL IV. A previous study found no difference in postprandial SD-LDL III or SD-LDL IV responses to HP versus HMF meals in lean subjects.¹² Wang et al. reported a lower fasting SD-LDLP with diets high in monounsaturated fat versus lower in fat in OW/O participants,²³ and Damasceno et al. reported a lower SD-LDLP on a Mediterranean diet supplemented with nuts compared to a low-fat diet in participants at high risk for CVD.²⁴ These studies, however, did not directly compare HMF with HP diets. LDL is made from VLDL when the TG in VLDL are hydrolyzed by lipoprotein lipase. SD-LDL is made when cholesterol ester transfer protein (CETP) moves TG from the VLDL to the LDL. The TG in the LDL are then hydrolyzed forming SD-LDL. As expected, both the SD-LDL III AUC (HP condition: r = 0.51; P = 0.01; HMF condition: r = 0.45; P = 0.03) and the SD-LDL IV AUC (HP condition: r = 0.54; P = 0.007; HMF condition: r = 0.48; P = 0.02) were correlated with the TG AUC in this study. Because VLDLP and TG in this study were higher in the HMF condition, the lower SD-LDL III for the HMF versus HP meal condition is probably more due to increased LDL clearance rather than decreased LDL production. Studies have reported that monounsaturated fatty acids upregulate the LDL receptor activity, which accelerates clearance of LDL.^25,26 Another reason for a lower LDL with the HMF meal condition may be due to the avocado in the HMF meal. Avocados are rich in phytosterols. Sialvera et al.²⁷ reported that supplementation with phytosterols decreased SD-LDL in patients with metabolic syndrome.

Both T-HDLP and SD-HDLP were higher during the HP versus HMF condition. The higher T-HDLP in the HP condition may be due to the higher SD-HDLP during that condition. LB-HDL2b was not different, but LB-HDL2b/T-HDLP was lower during the HP condition versus HMF condition. Shah et al. have also observed a higher T-HDLP and SD-HDLP and lower LB-HDL2b/T-HDLP ratio in lean subjects given HP compared to HMF meals.¹² They also observed a lower LB-HDL2b on the HP versus HMF meal.¹² Damasceno et al. reported an increase in large HDL with a Mediterranean diet supplemented with nuts or olive oil in patients with high risk for CVD.²⁴ Wang et al., however, reported no difference in T-HDLP or large HDLP between diets rich in monounsaturated fat compared to a lower fat diet among OW/O subjects, possibly because the difference in monounsaturated fat content between the diets was much smaller than that between the HP and HMF meals in this study.²³ The lower LB-HDL2b/T-HDLP during the HP versus HMF condition, in this study, indicates that there was a smaller proportion of the large-buoyant, more cardio-protective HDL, in relation to the total HDL, in the HP condition. A large compared to a small HDL facilitates more reverse cholesterol transport, a process by which excess cholesterol is transferred from peripheral tissues, including atherosclerotic lesions, and taken to the liver and removed through biliary secretion.^28,29 Apolipoprotein (Apo) A1 in HDL plays a role in transporting cholesterol from peripheral tissues to HDL.²⁹ Apo A1, however, dissociates from SD-HDL, and may be lost by renal clearance,²⁸ which may affect cholesterol removal from peripheral tissues. The lower SD-HDLP with HMF compared to the HP meal may be due to decreased CETP production. Replacing diets high in saturated fat with cis-monounsaturated fat decreases CETP production.³⁰ CETP transfers TG from VLDL to HDL. The TG in the HDL are then hydrolyzed by lipoprotein lipase leading to SD-HDL. A reduction in CETP related to an HMF diet would thus result in lower SD-HDLP. In this study, TG AUC was not significantly correlated with the SD-HDLP AUC (HP: r = 0.28; P = 0.18; HMF: r = 0.10; P = 0.63), however.

Leptin, TNF-α, and MCP-1 responses did not vary by meal composition. Raben et al. also found no difference in postprandial leptin responses between meals rich in protein and fat in normal-weight healthy subjects.¹⁴ The two meals were different in their carbohydrate content, and the type of fat in the high-fat meal was not shown, however.¹⁴ This makes it difficult to compare the results from this study to the results in the present study. Parvaresh Rizi et al. found no postprandial meal effect on TNF-α in insulin-resistant obese and insulin-sensitive lean subjects in response to HC, HP, and HMF meals.¹³ Holmer-Jensen et al.¹⁵ reported suppression of MCP-1 after consumption of a high-fat meal supplemented with whey, cod protein, gluten, or casein.¹⁵ However, an HP meal was not compared to an HMF meal in this study, and it was limited by a small sample size.

This study has some limitations. The postprandial response was limited to one meal and 3 hr. However, peak TG response has been seen at 3 hr in response to a high-fat meal.³¹ Nevertheless, a longer postprandial duration may be prudent. Our sample may have a selection bias because the subjects were not randomly selected. The study had several strengths. The two meals were similar in energy, carbohydrate, and added sugar content. Energy intake and physical activity were controlled before the study days. Body weight was the same on the study days. The study was a randomized crossover design.

In conclusion, obese subjects had lower TG and VLDLP, but less favorable SD-LDL III, SD-HDLP, and LB-HDL2b/T-HDLP ratio responses to the HP than the HMF meals. Longer term studies need to confirm these results. Obese individuals with high SD-LDL III, SD-HDLP, and low LB-HDL2b/T-HDLP may benefit from HMF meals, whereas those with high TG may benefit from HP meals.

Footnotes

Acknowledgments

The authors would like to acknowledge Brooke Bouza, Lauren Nelson, Bethany Schneider, Manall Jaffery, Mike Levitt, and Dr. Sheena Shah-Simpson for assisting with the study. This study was partly funded by Invests in Scholarship grant from Texas Christian University.

Author Disclosure Statement

No conflicting financial interests exist.

References

Hales

, Carroll

, Fryar

, et al. Prevalence of obesity among adults and youth: United States, 2015–2016. NCHS Data Brief, 2017:1–7.

Bays

, Toth

, Kris-Etherton

, et al. Obesity, adiposity, and dyslipidemia: A consensus statement from the National Lipid Association. J Clin Lipidol, 2013; 7:304–383.

Jung

, Choi

. Obesity and its metabolic complications: The role of adipokines and the relationship between obesity, inflammation, insulin resistance, dyslipidemia and nonalcoholic fatty liver disease. Int J Mol Sci, 2014; 15:6184–6223.

Garg

. High-monounsaturated-fat diets for patients with diabetes mellitus: A meta-analysis. Am J Clin Nutr, 1998; 67(3 Suppl):577S–582S.

Qian

, Korat

, Malik

, et al. Metabolic effects of monounsaturated fatty acid-enriched diets compared with carbohydrate or polyunsaturated fatty acid-enriched diets in patients with type 2 diabetes: A systematic review and meta-analysis of randomized controlled trials. Diabetes Care, 2016; 39:1448–1457.

Wycherley

, Moran

, Clifton

, et al. Effects of energy-restricted high-protein, low-fat compared with standard-protein, low-fat diets: A meta-analysis of randomized controlled trials. Am J Clin Nutr, 2012; 96:1281–1298.

Appel

, Sacks

, Carey

, et al. Effects of protein, monounsaturated fat, and carbohydrate intake on blood pressure and serum lipids: Results of the OmniHeart randomized trial. JAMA, 2005; 294:2455–2464.

Luscombe-Marsh

, Noakes

, Wittert

, et al. Carbohydrate-restricted diets high in either monounsaturated fat or protein are equally effective at promoting fat loss and improving blood lipids. Am J Clin Nutr, 2005; 81:762–772.

Kugiyama

, Doi

, Takazoe

, et al. Remnant lipoprotein levels in fasting serum predict coronary events in patients with coronary artery disease. Circulation, 1999; 99:2858–2860.

10.

Berneis

, Krauss

. Metabolic origins and clinical significance of LDL heterogeneity. J Lipid Res, 2002; 43:1363–1379.

11.

Kontush

. HDL particle number and size as predictors of cardiovascular disease. Front Pharmacol, 2015; 6:218.

12.

Shah

, Jaffery

, Adams-Huet

, et al. Effect of meal composition on postprandial lipid concentrations and lipoprotein particle numbers: A randomized cross-over study. PLoS One, 2017; 12:1–18.

13.

Parvaresh Rizi

, Baig

, Shabeer

, et al. Meal rich in carbohydrate, but not protein or fat, reveals adverse immunometabolic responses associated with obesity. Nutr J, 2016; 15:100.

14.

Raben

, Agerholm-Larsen

, Flint

, et al. Meals with similar energy densities but rich in protein, fat, carbohydrate, or alcohol have different effects on energy expenditure and substrate metabolism but not on appetite and energy intake. Am J Clin Nutr, 2003; 77:91–100.

15.

Holmer-Jensen

, Karhu

, Mortensen

, et al. Differential effects of dietary protein sources on postprandial low-grade inflammation after a single high fat meal in obese non-diabetic subjects. Nutr J, 2011; 10:115.

16.

Shah

, Franklin

, Adams-Huet

, et al. Effect of meal composition on postprandial glucagon-like peptide-1, insulin, glucagon, C-peptide, and glucose responses in overweight/obese subjects. Eur J Nutr, 2017; 56:1053–1062.

17.

WHO Expert Consultation. Appropriate body-mass index for Asian populations and its implications for policy and intervention strategies. Lancet, 2004; 363:157–163.

18.

Inoue

, Zimmet

, Caterson

, et al. The Asia-Pacific Perspective: Redefining Obesity and Its Treatment. Sydney, Australia: Health Communications Australia Pty Limited. 2000:1–56.

19.

Friedewald

, Levy

, Fredrickson

. Estimation of the concentration of low-density lipoprotein cholesterol in plasma, without use of the preparative ultracentrifuge. Clin Chem, 1972; 18:499–502.

20.

Oliver

, Joubert

, Caldwell

, et al. A longitudinal study examining the effects of a season of American football on lipids and lipoproteins. Lipids Health Dis, 2015; 14:35.

21.

El Khoury

, Anderson

. Recent advances in dietary proteins and lipid metabolism. Curr Opin Lipidol, 2013; 24:207–213.

22.

Maki

, Palacios

, Lindner

, et al. Replacement of refined starches and added sugars with egg protein and unsaturated fats increases insulin sensitivity and lowers triglycerides in overweight or obese adults with elevated triglycerides. J Nutr, 2017; 147:1267–1274.

23.

Wang

, Bordi

, Fleming

, et al. Effect of a moderate fat diet with and without avocados on lipoprotein particle number, size and subclasses in overweight and obese adults: A randomized, controlled trial. J Am Heart Assoc, 2015; 4:e001355.

24.

Damasceno

, Sala-Vila

, Cofan

, et al. Mediterranean diet supplemented with nuts reduces waist circumference and shifts lipoprotein subfractions to a less atherogenic pattern in subjects at high cardiovascular risk. Atherosclerosis, 2013; 230:347–353.

25.

Kurushima

, Hayashi

, Shingu

, et al. Opposite effects on cholesterol metabolism and their mechanisms induced by dietary oleic acid and palmitic acid in hamsters. Biochim Biophys Acta, 1995; 1258:251–256.

26.

Rumsey

, Galeano

, Lipschitz

, et al. Oleate and other long chain fatty acids stimulate low density lipoprotein receptor activity by enhancing acyl coenzyme A: Cholesterol acyltransferase activity and altering intracellular regulatory cholesterol pools in cultured cells. J Biol Chem, 1995; 270:10008–10016.

27.

Sialvera

, Pounis

, Koutelidakis

, et al. Phytosterols supplementation decreases plasma small and dense LDL levels in metabolic syndrome patients on a westernized type diet. Nutr Metab Cardiovasc Dis, 2012; 22:843–848.

28.

Barter

, Kastelein

. Targeting cholesteryl ester transfer protein for the prevention and management of cardiovascular disease. J Am Coll Cardiol, 2006; 47:492–499.

29.

Curtiss

, Valenta

, Hime

, et al. What is so special about apolipoprotein AI in reverse cholesterol transport?. Arterioscler Thromb Vasc Biol, 2006; 26:12–19.

30.

Jansen

, Lopez-Miranda

, Castro

, et al. Low-fat and high-monounsaturated fatty acid diets decrease plasma cholesterol ester transfer protein concentrations in young, healthy, normolipemic men. Am J Clin Nutr, 2000; 72:36–41.

31.

Adamska

, Ostrowska

, Goscik

, et al. Intake of meals containing high levels of carbohydrates or high levels of unsaturated fatty acids induces postprandial dysmetabolism in young oerweight/obese men. Biomed Res Int, 2015:147196.