Orally delivered acetate and appetite effects

The effects of long-term supplementation with apple cider vinegar as a source of acetate on anthropometric measures in human subjects were reported in a recently published three-way parallel study⁽Reference Kondo, Kishi and Fushimi³⁸⁾. In this study, healthy obese Japanese volunteers were randomised to the control (n 58, 50 completed), low-dose (n 59, 54 completed) or high-dose (n 58, 51 completed) treatment, consuming a beverage containing 0, 15 or 30 ml apple cider vinegar, respectively (equivalent to 0, 12·5 and 25 mmol acetate) daily for 12 weeks. At the end of the intervention, those in both the low- and high-dose treatment groups had a significantly lower BW, BMI, % body fat, waist and hip circumference, waist:hip ratio, visceral fat area, subcutaneous fat area and TAG concentrations relative to control. Additionally total fat area and systolic blood pressure were significantly lower in the high-dose group relative to control⁽Reference Kondo, Kishi and Fushimi³⁸⁾.

This study therefore provides some compelling evidence that vinegar as a source of acetate may influence appetite. The acute effects of vinegar supplementation on appetite and satiety in human subjects have been investigated in a few studies (summarised in Table 1) that will be reviewed later⁽Reference Hlebowicz, Lindstedt and Björgell^35–Reference Kondo, Kishi and Fushimi³⁸⁾.

In an acute four-way crossover study, Ostman et al. provided human volunteers (n 12) bread (containing 50 g carbohydrate) soaked in 0 g (control), 18, 23 and 28 g vinegar (equivalent to 0, 18, 23 and 28 mmol acetic acid) for their breakfast following an overnight fast. They reported a dose–response increase in the area under curve (AUC) for satiety ratings (Fig. 2), with a significantly higher AUC following the 28 g dose relative to control⁽Reference Ostman, Granfeldt and Persson³⁷⁾.

Fig. 2. Postprandial satiety score area under curve (AUC) following ingestion of bread (containing 50 g carbohydrate) soaked in 0, 18, 23 and 28 g white wine vinegar, supplying 0, 18, 23 and 28 mmol acetate, respectively. A dose–response increase in the satiety AUC was observed. Values are mean with error bars representing the SEM. Taken from Ostman et al. ⁽Reference Ostman, Granfeldt and Persson³⁷⁾.

The same group also investigated the effects of vinegar and form of wheat grain (refined, milled or wholegrain). In this four-way crossover study, participants (n 15, 13 completed) consumed a test breakfast comprising white, milled grain or wholegrain bread soaked in 28 g vinegar (supplying 28 mmol acetate) compared to white bread without vinegar (control). The postprandial satiety score AUC was significantly higher following wholegrain bread soaked in vinegar relative to the other three treatments, with none of the other treatments altering satiety⁽Reference Hlebowicz, Lindstedt and Björgell³⁵⁾. However, as wholegrain bread without added vinegar was not investigated, it is not possible to conclude whether the increased satiety arose from the wholegrain structure of the bread alone or it was an additive effect of the grain structure with the vinegar.

A two-by-two crossover study by a different group investigated the additive effects of acetic acid (28 mmol) and cinnamon mixed into a glucose drink and milk rice pudding. This study reported the satiety incremental AUC did not significantly differ between treatments, although a main effect of acetic acid approaching significance (P=0·064) was found⁽Reference Mettler, Schwarz and Colombani³⁶⁾.

However, none of these previous studies investigated quantitative effects on appetite. In addition, subjective effects on appetite were investigated using a single rating bipolar scale⁽Reference Hlebowicz, Lindstedt and Björgell^35, Reference Ostman, Granfeldt and Persson³⁷⁾ or point scale⁽Reference Mettler, Schwarz and Colombani³⁶⁾, rather than a variety of appetite-related visual analogue scale (VAS) questionnaires.

In our own laboratories, we carried out an acute three-way crossover study with two main objectives: to investigate the acute effects of acetate supplementation on subjective and quantitative measures of appetite, and to investigate the influence of product palatability. Subjective appetite effects were assessed using a variety of appetite-related VAS questionnaires. Alongside a standard breakfast (jam sandwiches) normal weight unrestrained eaters (n 16) were provided a drink of sugar-free orange squash containing 25 g vinegar (supplying 25 mmol acetate) in a more palatable (Pal, vinegar divided across two drinks) and unpalatable (Unpal, vinegar in a more concentrated form in one drink) form compared to control (drink with no added vinegar). Palatability ratings for the breakfast were significantly lower for the Unpal treatment compared to both control and Pal.

Vinegar treatment significantly lowered subjective appetite VAS ratings for hunger (P=0·045) and the desire to eat (P=0·036) and increased ratings for fullness (P<0·0001). Mean ad libitum EI of a large pre-weighed homogenous pasta meal provided 3 h postprandially was significantly influenced by vinegar treatment (P=0·022). The mean intake was lowest with Unpal treatment, followed by Pal, with the highest mean intake occurring with the control treatment, which was significantly higher than with Unpal treatment. Furthermore, a pooled correlation analysis found significant correlations between breakfast palatability ratings and various appetite measures including 24 h VAS AUC and ad libitum EI (J Darzi, GS Frost and MD Robertson, unpublished results). Our findings therefore indicate that while vinegar as a source of acetate may reduce appetite, this effect may be due, at least in part, to palatability effects of the vinegar-containing test products.

Orally delivered propionate and appetite effects

Two acute crossover studies from the same group have examined effects on appetite of bread baked with added Na propionate when consumed as part of a mixed meal for breakfast. In the first study, participants (n 11) attended on six occasions and were provided breakfast made using wholemeal bread (control) or the same bread with added Na propionate (low and high dose), sourdough, lactic acid or Ca lactate⁽Reference Liljeberg, Lonner and Bjorck⁴¹⁾. In the second study, participants (n 12) attended on three occasions and were provided breakfast prepared using wholemeal bread (control) or the same bread with added Na propionate (high dose only) or lactic acid⁽Reference Liljeberg and Bjorck⁴⁰⁾. The actual dose of Na propionate was not reported by the authors, but is calculated as 15 and 45 mmol for low- and high-dose breads, respectively. In both studies, the addition of Na propionate was reported to increase the postprandial satiety rating AUC in a dose–response manner (Fig. 3(A)), with a significantly higher-satiety AUC reported following the high dose of Na propionate bread relative to control⁽Reference Liljeberg and Bjorck^40, Reference Liljeberg, Lonner and Bjorck⁴¹⁾. However, at the same time, the acceptability rating score was reduced by the addition of Na propionate in a dose–response manner (Fig. 3(B))⁽Reference Liljeberg, Lonner and Bjorck⁴¹⁾.

Fig. 3. (A) Postprandial satiety score area under curve (AUC) and (B) acceptability score following ingestion of a mixed breakfast made with wholemeal bread (control) or bread baked with added Na propionate in a low (approximately 15 mmol) and high (approximately 45 mmol) dose. A dose–response increase in satiety was found, which was accompanied by a dose–response decrease in the acceptability score. *Mean values for the high-dose bread differed significantly from control (P<0·05). Values are mean with error bars representing the SEM. Adapted from Liljeberg et al. ⁽Reference Liljeberg, Lonner and Bjorck⁴¹⁾.

The provision of Na propionate is also associated with increased nausea. When volunteers (n 10) ingested 3 g Na propionate (equivalent to 31 mmol) combined with 30 g sunflower oil (as a source of PUFA) mixed into a pasta meal at breakfast, postprandial nausea ratings were significantly increased⁽Reference Frost, Brynes and Dhillo³⁹⁾. In this study, the postprandial incremental AUC for the appetite gut hormone GLP-1 was significantly increased by Na propionate and PUFA treatment, although subjective appetite ratings and EI of an ad libitum test meal served 4 h postprandially did not differ from the control (pasta with no added Na propionate/PUFA)⁽Reference Frost, Brynes and Dhillo³⁹⁾. However, as PUFA was delivered alongside the Na propionate, it was not possible to determine whether the observed effects were attributed to the added PUFA, propionate, or both and if the higher energy content of the Na propionate/PUFA pasta test meals had an effect.

In our own laboratories we carried out an acute two-way crossover study to investigate the effects of providing a palatable propionate-rich sourdough bread compared to an equally palatable and visually identical non-sourdough control bread on subjective and quantitative measures of appetite. Prior sensory evaluation of the bread products determined the sourdough and control breads to be equally acceptable. Healthy, normal weight unrestrained eaters (n 20) were provided a breakfast comprising jam sandwiches made using the palatable sourdough bread (supplying a total of 6·0 mmol propionate) or the non-sourdough control bread. In this study, we found that the palatable propionate-rich sourdough bread did not influence appetite, with no significant differences between treatments on postprandial appetite VAS ratings, ad libitum EI 3 h postprandially or on 24 h intake. Our findings therefore indicate that when provided in a palatable form, propionate does not appear to influence appetite⁽Reference Darzi, Frost and Robertson⁴²⁾. This implies that previous reports of Na propionate ingestion being associated with increased satiety⁽Reference Liljeberg and Bjorck^40, Reference Liljeberg, Lonner and Bjorck⁴¹⁾ may be a test product palatability effect, at least in part.

Oral SCFA and appetite summary

In summary, initial studies indicate that oral SCFA may enhance satiety⁽Reference Hlebowicz, Lindstedt and Björgell^35–Reference Liljeberg, Lonner and Bjorck⁴¹⁾. However, our findings suggest that product palatability may explain the appetite effects of oral SCFA rather than being a physiological effect of SCFA (J Darzi, GS Frost and MD Robertson, unpublished results).

Studies investigating the link between sensory properties of food and appetite suggest that sensory properties influence food choice and quantities consumed⁽Reference Sørensen, Møller and Flint⁴³⁾. A pooled analysis of data from studies in which investigators manipulated food palatability concluded there was a strong linear relationship between the change in rated palatability and difference in food intake following food manipulation (Fig. 4)⁽Reference Yeomans⁴⁴⁾. This provides a rationale for our hypothesis that oral SCFA palatability explains effects on appetite and satiety, most likely via mechanisms mediated at the cephalic phase.

Fig. 4. Relationship between change in palatability ratings and food intake when palatability is manipulated. Each data point represents a published study. Taken from Yeomans⁽Reference Yeomans⁴⁴⁾.

Inulin-type fructans and effects on appetite

Initial rodent model studies give evidence that inulin-type fructans may enhance satiety⁽Reference Cani, Knauf and Iglesias^49–Reference Cani, Neyrinck and Maton⁵⁴⁾. Chronic OF supplementation in rodents significantly decreases EI⁽Reference Cani, Dewever and Delzenne^50, Reference Cani, Hoste and Guiot^51, Reference Cani, Neyrinck and Maton⁵⁴⁾, suppresses BW gain⁽Reference Cani, Knauf and Iglesias^49, Reference Cani, Hoste and Guiot^51, Reference Cani, Neyrinck and Maton⁵⁴⁾ (although not in all cases⁽Reference Cani, Dewever and Delzenne^50, Reference Kok, Morgan and Williams⁵³⁾), suppresses fat mass gain⁽Reference Cani, Knauf and Iglesias^49–Reference Cani, Hoste and Guiot^51, Reference Cani, Neyrinck and Maton⁵⁴⁾, increases proximal colon GLP-1 (7–36) amide concentrations⁽Reference Cani, Knauf and Iglesias^49–Reference Cani, Hoste and Guiot⁵¹,Reference Kok, Morgan and Williams^53, Reference Cani, Neyrinck and Maton⁵⁴⁾, increases portal plasma GLP-1 (7–36)⁽Reference Cani, Dewever and Delzenne^50, Reference Cani, Hoste and Guiot^51, Reference Cani, Neyrinck and Maton⁵⁴⁾ and gastric-inhibitory polypeptide⁽Reference Kok, Morgan and Williams⁵³⁾ concentrations, reduces plasma ghrelin concentrations⁽Reference Cani, Dewever and Delzenne⁵⁰⁾ and increases the caecum mass⁽Reference Cani, Dewever and Delzenne^50, Reference Cani, Hoste and Guiot^51, Reference Kok, Morgan and Williams⁵³⁾. Most studies examined OF, with few investigating higher DP inulin-type fructans (e.g. inulin HP)⁽Reference Cani, Dewever and Delzenne^50, Reference Daubioul, Rousseau and Demeure⁵²⁾. Supplementing rats diets with Synergy 1 (mixture of low- and high-DP inulin-type fructans) resulted in significantly suppressed BW gain⁽Reference Daubioul, Rousseau and Demeure⁵²⁾ and fat mass gain⁽Reference Cani, Dewever and Delzenne⁵⁰⁾, reduced EI⁽Reference Cani, Dewever and Delzenne^50, Reference Daubioul, Rousseau and Demeure⁵²⁾, reduced proximal colon GLP-1 (7–36) concentrations⁽Reference Cani, Dewever and Delzenne⁵⁰⁾ and reduced caecum mass⁽Reference Cani, Dewever and Delzenne⁵⁰⁾. Furthermore, supplementation with inulin HP significantly suppressed BW gain and increased caecum mass, but did not influence proximal colon GLP-1 (7–36) concentrations or fat mass gain⁽Reference Cani, Dewever and Delzenne⁵⁰⁾.

Following on from these promising data from animal models, a few human studies have investigated acute (Table 2) and more have investigated chronic (Table 3) effects of supplementing with inulin-type fructans on appetite and/or anthropometrics. A review of these studies now follows.

Table 2. Summary of human clinical studies investigating acute effects of inulin-type fructan supplementation on appetite

E, energy; EI, energy intake; OF, oligofructose; AUC, area under curve.

Table 3. Summary of human clinical studies investigating chronic effects of inulin-type fructan supplementation on appetite

GERD, gastrooesophageal reflux disease; ↑ higher; ↓, lower; GLP, glucagon-like peptide; AUC, area under curve; OF, oligofructose; PYY, peptide YY; CCK, cholecystokinin; GIP, gastric-inhibitory polypeptide; BW, body weight.

Three different studies have been published that investigated acute effects of ingesting inulin-type fructans on subsequent appetite (Table 2). Peters et al. provided volunteers (n 21) with a meal replacement bar containing 8 g OF or a non-OF control to be eaten in the free-living setting at breakfast and lunch on day 1 and breakfast only on day 2. No significant effects of treatment were found on satiety ratings or EI at an ad libitum buffet meal provided 4 h postprandially⁽Reference Peters, Boers and Haddeman⁵⁵⁾.

By contrast, Archer and colleagues reported 24 h EI was significantly lowered by supplementation with 24 g inulin HP when supplied to volunteers (n 33) as a fat replacer in a sausage patty, relative to a full-fat patty (control). Satiety ratings were not significantly altered⁽Reference Archer, Johnson and Devereux⁵⁶⁾. However, control and inulin-containing sausage patties were not matched for energy in this study.

Similarly, consumption of yoghurt containing 6 g inulin-type fructans (type used not specified) significantly decreased subsequent ad libitum EI in a controlled setting relative to ‘no preload’ (control) in healthy volunteers (n 38)⁽Reference Perrigue, Monsivais and Drewnowski⁵⁷⁾. However, the overall EI including the preload and the ad libitum intake was more than the ‘no preload’ condition.

A larger number of chronic supplementation studies have been published investigating effects on appetite (Table 3), again with variable results and study limitations, as reviewed below.

Let us first take a look at the chronic effects of OF supplementation on subjective and quantitative appetite measures. In a two-way crossover study, Cani and co-workers reported that daily supplementation of healthy volunteers (n 10) with 16 g OF for 2 weeks significantly lowered ad libitum EI at breakfast and lunch and the mean daily EI relative to control (16 g maltodextrin)⁽Reference Cani, Joly and Horsmans⁵⁸⁾. Relative to control, OF supplementation also significantly increased satiety ratings after breakfast and dinner, and reduced hunger and prospective consumption ratings following dinner⁽Reference Cani, Joly and Horsmans⁵⁸⁾. However, as preload test meals were provided ad libitum rather than being a fixed, standardised preload, the postprandial appetite ratings are difficult to interpret.

Similarly, a recent parallel study reported supplementation of volunteers (n 51) with 14 g Jerusalem Artichoke concentrate (containing OF) daily for 12 weeks reduced rated hunger⁽Reference Antal, Regöly-Mérei and Biró⁵⁹⁾. However, it was unclear whether effects on hunger were significant.

By contrast, in a two-way parallel study, Parnell and Reimer reported daily supplementation of healthy overweight volunteers (n 48) for 12 weeks with 21 g OF did not significantly alter satiety ratings and EI relative to control (7·89 g maltodextrin), although the PYY AUC was significantly higher⁽Reference Parnell and Reimer⁶⁰⁾. However, this parallel study was not sufficiently powered to assess subjective appetite effects⁽Reference Flint, Raben and Blundell⁶¹⁾.

Effects of OF supplementation on gut peptide production have also been investigated by Piche and colleagues in patients with gastrooesophageal reflux disease. In a two-way crossover study, participants (n 9) ingested 19·8 g OF or sucrose (control) for 7 d. The postprandial plasma GLP-1 AUC was significantly higher in response to a meal tolerance test following OF supplementation relative to control; however both PYY and cholecystokinin were not significantly influenced by treatment⁽Reference Piche, des Varannes and Sacher-Huvelin⁶²⁾.

Let us now take a look at the effects of mixed inulin-type fructans. In a two-way parallel chronic study, Cani and co-workers asked volunteers to supplement their usual diet with 16 g Synergy-1 (n 5) or 16 g Maltodextrin (control) (n 5) daily for 12 weeks. At the end of the supplementation period volunteers attended a meal tolerance test challenge. The authors reported postprandial subjective appetite ratings did not differ between the two treatments during the meal tolerance test, although the postprandial GLP-1 and PYY response was significantly higher with Synergy-1 treatment⁽Reference Cani, Lecourt and Dewulf⁶³⁾. However, this study was underpowered and the meal tolerance test was supplied ad libitum making interpretation of postprandial data difficult.

In a two-way crossover study Whelan and colleagues asked healthy volunteers (n 11) to consume a standard enteral formula (control) or a modified enteral formula with added pea fibre, inulin and OF for 2 weeks as their sole nutrition source⁽Reference Whelan, Efthymiou and Judd⁶⁴⁾. Significantly increased daily mean and minimum fullness and minimum satiety ratings were observed when the modified enteral formula was consumed relative to control⁽Reference Whelan, Efthymiou and Judd⁶⁴⁾. However, as pea fibre was also included in the active treatment, observed results may not have been solely due to the inulin-type fructans.

Less compelling is anthropometric and food intake data from chronic intervention studies, although these were not the main outcomes of interest for all studies⁽Reference Alles, de Roos and Bakx^65–Reference Luo, Van Yperselle and Rizkalla⁷²⁾ except for two⁽Reference Antal, Regöly-Mérei and Biró^59, Reference Parnell and Reimer⁶⁰⁾. One study reported daily supplementation with 14 g Jerusalem Artichoke concentrate (containing OF) for 12 weeks significantly reduced BMI and % body fat levels⁽Reference Antal, Regöly-Mérei and Biró⁵⁹⁾, while another reported daily supplementation with 21 g OF significantly reduced BW, fat mass and central fat mass⁽Reference Parnell and Reimer⁶⁰⁾. However, none of the other studies found a significant influence of inulin-type fructan supplementation on BW post-intervention⁽Reference Antal, Regöly-Mérei and Biró^59, Reference Alles, de Roos and Bakx^65, Reference Giacco, Clemente and Luongo^66, Reference Forcheron and Beylot^68–Reference Luo, Van Yperselle and Rizkalla⁷²⁾ or on food intake during the intervention⁽Reference Alles, de Roos and Bakx^65–Reference Forcheron and Beylot⁶⁸⁾.

In our own laboratories, we carried out an acute three-way crossover study to investigate the effects on subjective and quantitative measures of appetite of providing inulin HP or another NDC l-rhamnose as a component of breakfast and lunch compared to a visually identical control (no NDC) following a 6 d run-in period. l-rhamnose was investigated as previously published data suggest propionate production is favoured during l-rhamnose fermentation, both in vitro ⁽Reference Fernandes, Rao and Wolever⁷³⁾ and in vivo ⁽Reference Vogt, Pencharz and Wolever⁷⁴⁾. Propionate is the most potent ligand for the SCFA receptor FFA3 and appears to promote leptin expression⁽Reference Xiong, Miyamoto and Shibata^26, Reference Covington, Briscoe and Brown²⁷⁾ and the production of PYY⁽Reference Karaki, Mitsui and Hayashi^28–Reference Cuche and Malbert³⁴⁾ as discussed earlier, providing a rationale l-rhamnose may enhance appetite. Inulin HP was investigated due to the lack of previous data regarding effects on appetite.

During an acute meal challenge, healthy, normal weight unrestrained eaters (n 13) ingested 22·4 g inulin HP or 25·5 g l-rhamnose as a split dose at a standard breakfast and 180 min later at a standard lunch. We found neither treatment influenced appetite, with no significant differences between treatments for postprandial appetite VAS ratings, ad libitum EI 7 h postprandially, intake during the 24 h period following the breakfast preload nor on the mean daily intake during the run-in period (J Darzi, GS Frost and MD Robertson, unpublished results).

Colonic SCFA and appetite summary

While a number of studies have investigated the influence of supplementation with inulin-type fructans on markers of appetite in human subjects, the results to date are contradictory making it difficult to draw firm conclusions. This is in part due to variable dosages and differing choices of inulin-type fructan supplements.