Effects of diacylglycerol on postprandial energy expenditure and respiratory quotient in healthy subjects

Introduction 

Diacylglycerol (DAG) oil is a natural oil that is present at a concentration of several percentages in vegetable oils [1], has a long history of use as a human food, and is widely used as cooking oil. DAG has been used as a food additive in small amounts, but the 1,3-specific lipase-catalyzed reverse reaction currently allows for the large-scale production of DAG, which is commercially available as cooking oil or processed oil and fat products in the United States and Japan. Recent studies have indicated that DAG is effective for postprandial hyperlipidemia and for obesity prevention [2], [3], [4], [5]. As a result, DAG is approved as a food for specific health use in Japan [6].

In trials of long-term ingestion in Japanese and Western populations, DAG oil decreased body weight and body fat to a greater extent than did triacylglycerol (TAG) oil [4], [5]. DAG and TAG oils of similar fatty acid composition have similar energy values and undergo similar amounts of absorption in rats [7], suggesting that these properties are not associated with the effectiveness of DAG. In contrast, Watanabe et al. [8] reported that a single administration of a diolein emulsion resulted in increased oxygen consumption compared with that of a triolein emulsion in rats, suggesting that an increase in postprandial energy expenditure (EE) is one of the mechanisms underlying the antiobesity effect of DAG. Diet-induced thermogenesis (DIT) results in a postprandial increase in EE, which accounts for only about 10% of total daily EE [9]. However, there are some reports that DIT is decreased in obesity [10], and some researchers have claimed that a long-term decrease in DIT has serious consequences for weight gain [11], suggesting that foodstuffs with a high DIT contribute to the prevention of obesity.

In a human trial, Kamphuis et al. [12] analyzed energy metabolism for 24 h after ingestion of a DAG-containing meal using a respiration chamber and reported a significant increase in the amount of lipid oxidation. However, there was no significant difference between DAG and TAG meals in total 24-h EE or total DIT after three meals plus between-meal snacks. Thus, their results differed from those of the animal experiment of Watanabe et al.

To investigate metabolic dynamics after loading with a DAG meal in more detail, we analyzed breath and serum for time-dependent changes in factors related to lipid and sugar metabolism.

Results 

EE and respiratory quotient 

The ΔEE 3 h after loading with the DAG meal tended to be higher than that after loading with the TAG meal (Fig. 1A; P < 0.1). The cumulative ΔEE after loading with the DAG meal was also higher than that after loading with the TAG meal, with a 74-kJ difference at the 5-h time point, but not significantly (Fig. 1B). The changes in respiratory quotient (ΔRQ) 2 and 5 h after loading with the DAG meal were significantly lower than those after loading with the TAG meal (Fig. 2).

View Large Image Download to PowerPoint

Fig. 1. Postprandial energy expenditure. (A) Changes in Δ values. (B) Changes in cumulative Δ values. Value are expressed as mean ± standard deviation. Dashed line connecting open circles indicates triacylglycerol meal, and solid line connecting solid circles indicates diacylglycerol meal. #Significant difference between triacylglycerol and diacylglycerol meals by paired t test, P < 0.1. Δ, change in amount.

View Large Image Download to PowerPoint

Fig. 2. Postprandial respiratory quotient. Values are expressed as mean ± standard deviation. Dashed line connecting open circles indicates triacylglycerol meal, and solid line connecting solid circles indicates diacylglycerol meal. Significant difference between triacylglycerol and diacylglycerol meals by paired t test, *P < 0.05, **P < 0.01.

Serum analysis (Table 4) 

The change in serum triglyceride levels at the 4-h time point after loading with the DAG meal tended to be lower than those after loading with the TAG meal (P < 0.1). There were no significant differences between the change in serum glucose levels after loading with the DAG or TAG meal. The change in serum insulin levels were significantly lower at the 0.5-h time point after loading with the DAG meal than after loading with the TAG meal, and the amount of insulin secreted remained lower at the 0- to 4-h time points. The change in serum non-esterified fatty acid levels at the 0.5-h time point after loading with the DAG meal tended to be higher than those after loading with the TAG meal (P < 0.1). There were no significant differences between the change in serum total ketone body levels after loading with the DAG meal and that after loading with the TAG meal.

Table 4. Changes in serum analysis items⁎

	Parameter	TAG meal	DAG meal
	TAG (Δ mmol/L)
	0.5 h	0.04±0.07	0.03±0.08
	1 h	0.25±0.12	0.22±0.14
	2 h	0.89±0.28	0.83±0.34
	4 h	1.53±0.70	1.27±0.77†
	Glucose (Δ mmol/L)
	0.5 h	0.97±0.64	0.97±0.55
	1 h	0.35±1.23	0.26±1.11
	2 h	0.28±0.98	0.44±0.88
	4 h	−0.16±0.57	−0.04±0.42
	Insulin (Δ pmol/L)
	0.5 h	384.5±193.1	296.1±133.6‡
	1 h	433.5±250.5	396.3±272.5
	2 h	366.2±309.3	298.0±172.1
	4 h	167.5±140.7	133.6±119.0
	NEFA (Δ mmol/L)
	0.5 h	−0.10±0.07	−0.07±0.05†
	1 h	−0.21±0.08	−0.17±0.04
	2 h	−0.20±0.09	−0.16±0.07
	4 h	−0.12±0.09	−0.09±0.12
	Total ketone bodies (Δ μmol/L)
	0.5 h	−13±17	−12±16
	1 h	−13±15	−12±16
	2 h	−10±17	−6±18
	4 h	−6±17	−2±18

DAG, diacylglycerol oil; NEFA, non-esterified fatty acid; TAG, triacylglycerol oil

⁎ Values are expressed as mean ± standard deviation. Differences in the Δ value between TAG and DAG meals by t test: † P < 0.1, ‡ P < 0.05.

Discussion 

In this study, we compared the postprandial energy balance after loading with a DAG meal with that after loading with a TAG meal. The postprandial ΔEE from baseline at the 3-h time point after loading with the DAG meal tended to be higher than those after loading with the TAG meal (P < 0.1; Fig. 1A). The cumulative ΔEE after loading with the DAG meal was also higher than that after loading with the TAG meal, with a 74-kJ difference at the 5-h time point, which was not significant (Fig. 1B). The ΔRQ after loading with the DAG meal was significantly lower than that after loading with the TAG meal, and the postprandial insulin increase 0.5-h after loading with the DAG meal was significantly lower than that after loading with the TAG meal (Table 4). These results suggest that DIT tends to be higher after loading with the DAG meal and that the DAG meal has higher lipid oxidation activity than the TAG meal (Fig. 2).

Although DIT accounts for only about 10% of total daily EE [9], some researchers have claimed that a long-term decrease in DIT has serious consequences for weight gain [11] and that DIT is decreased in obese individuals [10]. In this trial, postprandial EE tended to increase after loading with the DAG meal (P < 0.1; Fig. 1A), suggesting that the effect of DAG on DIT has a role in the mechanism underlying the body fat–reducing effect of DAG [4], [5]. When considering the results of a trial by Nagao et al. [5] on long-term ingestion of the DAG oil in Japanese subjects who had the same body mass index as in the present trial, this effect on DIT alone cannot theoretically explain the body weight–reducing effect of DAG. In their trial, decreases in body weight after a 16-wk ingestion period of the DAG oil differed from that of the TAG oil by 1.5 kg [5]. Because human adipose tissue is composed of 80% fat and 20% water and other components, EE after ingestion of the DAG oil was computationally higher by 37.7 kJ (9 kcal) × 1500 (grams of body fat) × 0.8 ÷ 120 (days) = 377 kJ/d [14]. In contrast, in the present trial, there was a 74-kJ difference between the cumulative ΔEE at the 5-h time point after loading with the DAG or TAG meal (Fig. 1B), a substantial difference from the results of the trial by Nagao et al. In addition, three-fold more DAG was administered in the present study than in the trial of Nagao et al., so the effect of a single DAG loading alone cannot explain the decrease in body weight associated with long-term ingestion of DAG, suggesting some additive or synergistic effect of repeated administrations.

In the present study, the ΔRQ after loading with the DAG meal was lower than that after loading with the TAG meal (Fig. 2), suggesting that a higher percentage of lipids was used after loading with the DAG meal, which is in agreement with the results of a trial by Kamphuis et al. [12] who used a respiration chamber. We noted a non-significant but potential effect of DAG on DIT, whereas Kamphuis et al. observed no difference in the 24-h total EE or DIT after loading with the DAG or TAG meal [12]. The discrepancy might be due to differences between the two trial protocols. They compared DAG with TAG loading on EE for 1 d, EE during sleep, and DIT after three meals plus between-meal snacks, whereas we compared DIT after loading with a single meal. Moreover, they administered a mean test oil loading of 33 g divided across meals (three meals plus between-meal snacks) in 1 d, whereas we loaded the subjects with 30 g of the test oil in a single meal. Therefore, we used a three-fold higher dose per meal compared with their trial. In addition, their study differed from ours in the subjects (female versus male) and analysis methods for DIT (they calculated DIT by subtracting resting EE from sleeping EE), possibly leading to different DIT results.

In the present study, our test diet contained a large amount of carbohydrate (52.1% as energy) and it stimulated insulin secretion in the DAG and TAG meals (Table 4). Although it promotes sugar use, insulin acts negatively in the degradation of body fat and the β-oxidation of fat in the liver and muscles, thus stimulating the storage of excess energy as fat in adipocytes [15]. Therefore, the significantly lower postprandial insulin levels after loading with the DAG meal than that after loading with the TAG meal in this trial might indicate a smaller stimulus in the direction of fat storage in postprandial energy metabolism. This low insulin level might be involved in the lower RQ after loading with the DAG meal in this trial (Fig. 2) and in the inhibitory effect of DAG on body fat accumulation [4], [5]. The effects of DAG on insulin secretion require further study such as the relation with gastrointestinal hormones.

Insulin is the most potent known regulator of flow of blood lipids into and out of adipose tissue through the mediation of lipoprotein lipase and hormone-sensitive lipase [16]. In the present study, postprandial serum non-esterified fatty acid levels decreased after loading with TAG and DAG meals (Table 4). In contrast, postprandial serum insulin level increased sharply after loading with the TAG and DAG meals (Table 4). Insulin is considered to be a potent key factor for suppression of postprandial non-esterified fatty acid defluxion from adipose tissue in our study. Serum TAG levels after loading with the DAG meal tended to be lower than those after loading with the TAG meal (P < 0.1; Table 4), suggesting that these low TAG levels are regulated at sites different from those of lipoprotein lipase- or insulin-mediated incorporation of lipids into adipocytes. In an experiment with rats, Kondo et al. [17] suggested that DAG oil metabolites resulting from lipase hydrolysis in the digestive tract are not easily used as substrates for DAG acyltransferase in the small intestinal epithelial cells compared with the TAG oil metabolites, suggesting that the DAG oil metabolites are not easily resynthesized into TAG. Therefore, lipid metabolism in the small intestinal epithelial cells might be important for DAG to inhibit postprandial hyperlipidemia.

In conclusion, a meal containing 30 g of DAG tended to produce higher postprandial ΔEE values than one containing 30 g of TAG, suggesting a higher DIT. Serum insulin level and postprandial RQ after loading with a DAG meal were significantly lower than those after loading with a TAG meal, suggesting higher postprandial lipid oxidation activity with the DAG meal.