Selection of sources of evidence and characteristics of the studies

The number of retrieved publications and the selection is illustrated in the PRISMA flow diagram (Figure 2). Overall, 134 studies (published in 141 scientific references, due to duplicate publications on the same study, and instances of multiple studies in a single publication) published between 1961 and 2022 were selected. The studies included were conducted in thirty different countries around the world (Figure 3). The participant characteristics of the human intervention studies included in this scoping review are displayed in Table 1.

Fig. 2. PRISMA-ScR flow diagram of the publication search and screening process. IAFNS, Institute for the Advancement of Food and Nutrition Sciences; FSANZ, Food Standards Australia New Zealand; EFSA, European Food Safety Agency.

Fig. 3. World map depiction to highlight countries with human clinical investigations of dietary pectin (n = 134) included in the scoping review. Colours indicate the number of studies completed in each country.

Table 1. Characteristics of the study participants, pectin daily dose and intervention duration from included studies (n = 134)

* Includes non-healthy participants in ‘healthy and non-healthy’ studies (n = 3), and ‘non-healthy’ studies (n = 49).

The majority of studies covered outcomes related to gut health (n = 50; reported in fifty-two publications) followed by glycaemic response and appetite (n = 47), fat metabolism (n = 34; reported in thirty-six publications), bioavailability of micronutrients (n = 20; reported in twenty-two publications) and other topics (e.g. radioactivity, heavy metals, cancer-related, bioavailability of drugs, infant growth parameters, cardiovascular health) (n = 21). Immune health-related studies (n = 8; reported in nine publications) are a more recent subject, with the first paper published on this topic in 2008 (Figure 4). The most common reference to pectin source was non-defined; however, other prevalent botanical origins included citrus and apple (Figure 5). Citrus was the most common botanical origin of pectin tested for fat metabolism and bioavailability of micronutrient studies and was widely used for gut health studies. Types of pectin included in the ‘other’ pectin category included carrot RG-I and other RG-I sources (e.g. potato).

Fig. 4. Publications (n = 141) timeline for human intervention studies with dietary pectin (n = 134) for the six most common categories of health-related effects. One reference can include several categories.

Fig. 5. Human intervention studies (n = 134) and pectin botanical origin by health topic.

The maximum reported dose was 50 g/d of HM pectin for 14 d^{(Reference Miettinen and Tarpila20)}. Seven other studies also tested daily doses above 30 g/d for 1–42 d^{(Reference Brenelli, Campos and Saad21–Reference Remer, Pietrzik and Manz29)} (Figure 6). The longest studies evaluated the consumption of 1 g/d of citrus pectin oligosaccharides for 127 d in infants^{(Reference Magne, Hachelaf and Suau30)} and 0.4–20 g/d of citrus pectin, hydrolysed citrus pectin and apple pectin for 168 d in healthy and non-healthy adults^{(Reference Havelund, Aalykke and Rasmussen31–Reference Palmer and Dixon35)} (Figure 7).

Fig. 6. Maximum daily dose of pectin evaluated in human intervention studies (n = 134) according to health topic. Average (red line) and median (green line) pectin daily dose.

Fig. 7. Study duration for human intervention studies (n = 134) with pectin (number of days) for each health topic category, including average (red line) and median (green line).

Among the ninety-one studies that were randomised controlled trials, thirty-three were conducted with parallel groups and fifty-nine in a crossover manner. Another forty-two studies were not randomised: thirty-one of them were conducted before 2000, and the more recent ones were pilot single-arm studies that explored new areas for potential health benefits related to pectin consumption (Figure 8).

Fig. 8. Heatmap visualisation for types of human study design by health topic. Multiple health topics may be covered in one study. *PP, post-prandial. **Markers of glucose regulation: fasting blood glucose, HOMA-IR, glycated haemoglobin.

Summary for gut health

Fifty studies (reported in fifty-two publications) evaluated markers related to gut health dating back to 1976. Of these studies, thirty-one were conducted before 2000. The botanical origin of the pectin used was not described in more than half of the studies, citrus was the second most frequently evaluated, followed by apple (n = 3) and sugarbeet (n = 2, Figure 5). The dose of pectin tested in studies ranged from 0.4 g/d to 50 g/d (Figure 6), with an average and median tested dose of 13.5 g/d and 12.6 g/d, respectively. The average duration of pectin intervention was 22 d, and the median 9 d (Figure 7). Most of the studies evaluated markers related to gut function (i.e. transit time, faecal output, digestive symptoms, gut barrier)^{(Reference Miettinen and Tarpila20,Reference Cummings, Southgate and Branch23–Reference Marthinsen and Fleming27,Reference Magne, Hachelaf and Suau30–Reference Kang, Tay and Guan32,Reference Atia, Girard-Pipau and Hébuterne36–Reference Xi, Xu and Tan68)} or gut fermentation (i.e. breath gas, faecal pH, SCFA, microbial enzymes)^{(Reference Cummings, Southgate and Branch23–Reference Fleming and Rodriguez25,Reference Atia, Girard-Pipau and Hébuterne36,Reference Fanaro, Jelinek and Stahl38–Reference Judd and Truswell40,Reference Mallett, Rowland and Bearne43,Reference Ross and Leklem48–Reference Spiller, Chernoff and Hill50,Reference Thorning, Bertolt and Nielsen54,Reference Vargo, Doyle and Floch55,Reference Zimmaro, Rolandelli and Koruda58,Reference An, Wilms and Smolinska60,Reference Bjorneklett, Fausa and Midtvedt61,Reference Nguyen, Welsh and Manion64,Reference Tadesse and Eastwood66,Reference Wanders, Mars and Borgonjen-van den Berg67,Reference Chinda, Nakaji and Fukuda69–Reference Wolever and Robb75)} (Figure 8).

Most of the studies were conducted in healthy subjects (n = 2 in infants, n = 25 in adults, n = 1 in elderly >65 years). The two RCT parallel studies in infants have specifically evaluated the digestive tolerance and other gut-related effects of approximately 1 g/d of citrus pectin oligosaccharides (95% < 3.8 kDa) supplemented via the infant formula (in replacement of equivalent dose of maltodextrin) for 50–127 d^{(Reference Magne, Hachelaf and Suau30,Reference Fanaro, Jelinek and Stahl38)} . The study in elderly compared, in an RCT parallel design, the effect of 15 g/d of sugarbeet pectin versus maltodextrin on the gut microbiome and associated outcomes between elderly and younger adults^{(Reference An, Wilms and Smolinska60)}. Overall, there is limited impact of a dietary supplementation with pectin on the gut function (full transit time, frequency and consistency of stools) of healthy individuals, illustrating a good digestive tolerance until 50 g/d, which is the maximum reported tested dose in humans. At this level and similar to what is observed in general with dietary fibres, there is a transient increase of flatus gas correlated with higher production of SCFA and, thus, microbial fermentation of pectin^{(Reference Rabbani, Teka and Saha46,Reference Rabbani, Teka and Zaman47)} .

Several studies were also conducted in specific sensitive or non-healthy populations. Two RCT parallel studies evaluated pectin (non-defined origin) supplementation for 7 d on the improvement of digestive symptoms and intestinal permeability in infants and young children with recurrent diarrhoea^{(Reference Rabbani, Teka and Saha46,Reference Rabbani, Teka and Zaman47)} . Studies conducted in non-healthy adults (n = 2 RCT crossover, n = 4 RCT parallel and n = 5 non-randomised) included: gastro-oesophageal reflux disease (n = 2)^{(Reference Havelund, Aalykke and Rasmussen31)}, gastric surgery (n = 2)^{(Reference Bjorneklett, Fausa and Midtvedt61,Reference Jenkins, Gassull and Leeds76)} , other intestine surgery (n = 2)^{(Reference Sandberg, Ahderinne and Andersson49,Reference Thirlby and Kelly53)} , small intestine-related issues (n = 2)^{(Reference Kang, Tay and Guan32,Reference Atia, Girard-Pipau and Hébuterne36)} , ulcerative colitis (n = 1)^{(Reference Wei, Gong and Zhu56)}, hypercholesterolaemic adults (n = 2)^{(Reference Miettinen and Tarpila20,Reference Veldman, Nair and Vorster77)} . Finally, nine RCT studies (n = 7 parallel, n = 2 crossover) evaluated the thickening effect of an equivalent daily dose of 1.4–16 g of pectin (mostly not described origin) to enteral feeding for 1–30 d during hospitalisation of patients. Frequency and severity of gut symptoms usually occurring with this type of feeding were evaluated^{(Reference Khongcharoensombat, Khemtong and Lakananurak42,Reference Mao, Xiong and Hu44,Reference Nakamura, Inokuchi and Fukushima45,Reference Tabei, Tsuchida and Akashi52,Reference Adachi, Furuta and Aimi59,Reference Miyazawa, Tomomasa and Kaneko63,Reference Schultz, Ashby-Hughes and Taylor65,Reference Wanders, Mars and Borgonjen-van den Berg67)} . There was a general decrease of digestive symptoms observed in fragile populations such as hospitalised patients with enteral feeding, who experienced fewer reflux and diarrhoea events. These benefits can be seen with a minimum of 2.5 g/d of pectin and are certainly explained by a thickening effect of pectin in the enteral formula especially when the pH starts to decrease in the stomach. This effect was also observed in formula-fed infants presenting frequent reflux symptoms when their usual formula was switched to a pectin-thickened (with other thickeners) formula^{(Reference Dupont and Vandenplas78)}. Unfortunately, there is a general lack of description on the type of pectin that was tested, except that it is of citrus origin. However, LM pectin is known to react with ionised calcium, especially in acidic conditions; thus, we can hypothesise that, as reported by Tabei et al. (2018), LM pectin could be favoured for enteral feeding^{(Reference Tabei, Tsuchida and Akashi52)}.

Despite a significant amount of research conducted in vitro in human gut microbiome models^{(Reference Pascale, Gu and Larsen11)}, there is still a limited number of intervention studies that have evaluated the effects of pectin on the gut microbiota composition^{(Reference Magne, Hachelaf and Suau30,Reference Fanaro, Jelinek and Stahl38,Reference Mallett, Rowland and Bearne43,Reference Thorning, Bertolt and Nielsen54–Reference Wei, Gong and Zhu56,Reference An, Wilms and Smolinska60)} , and among them only three studies report composition obtained with 16S rRNA/DNA techniques^{(Reference Thorning, Bertolt and Nielsen54,Reference Wei, Gong and Zhu56,Reference An, Wilms and Smolinska60)} . In general, these studies reported limited effects, except for pectin-derived ingredients with the lowest molecular weight^{(Reference Magne, Hachelaf and Suau30,Reference Thorning, Bertolt and Nielsen54)} . However, several intervention studies demonstrate that pectin is slowly but completely fermented by the bacteria in the human gut microbiome, which can relate to higher SCFA production, especially acetate^{(Reference Cummings, Southgate and Branch23,Reference Fleming, Marthinsen and Kuhnlein24,Reference Spiller, Chernoff and Hill50,Reference Pomare, Branch and Cummings72)} . Because of the relatively poor reporting of the characteristics of the pectin, it is not possible to confirm yet if there is a visible structure–function relationship in human studies as it was found for in vitro studies conducted with human gut microbiome^{(Reference Pascale, Gu and Larsen11)}, nor what the minimum dose and duration of supplementation are to observe the increased production of SCFA.

Summary for glycaemic response & appetite

The effect of pectin on glycaemic response has been widely studied, with interventions related to glycaemic response and appetite being the second largest health topic and including forty-seven studies. The publication date of these studies ranged from 1977 to 2020, though only thirteen of these forty-seven studies were published after the year 2000. The pectin botanical origin tested in these studies consisted primarily of non-defined pectin sources (n = 34), as well as apple (n = 8) and citrus (n = 2) (Figure 5). The dose of pectin tested in studies related to glycaemic response ranged from 0.1 g/d to 40 g/d (Figure 6), with an average and median tested dose of 13.2 g/d and 14.5 g/d, respectively. The average duration of pectin intervention was 11 d, though the median was 1 d, as twenty-nine of the forty-seven studies were 1-d, single-intake studies (Figure 7).

Within this health topic, measured outcomes related to glycaemic response commonly included post-prandial glucose and insulin response^{(Reference Brenelli, Campos and Saad21,Reference Ranganathan, Champ and Pechard28,Reference Thorning, Bertolt and Nielsen54,Reference Jenkins, Wolever and Leeds71,Reference Jenkins, Leeds and Gassull79–Reference Williams, James and Evans96)} . Other measures included gastric emptying^{(Reference Atia, Girard-Pipau and Hébuterne36,Reference Nguyen, Welsh and Manion64,Reference Wanders, Mars and Borgonjen-van den Berg67,Reference Lawaetz, Blackburn and Bloom80,Reference Leeds, Ralphs and Ebied81,Reference Sandhu, El Samahi and Mena84,Reference Schwartz, Levine and Weinstock86–Reference Shimoyama, Kusano and Kawamura88,Reference Wanders, Feskens and Jonathan95,Reference Andersen, Holtug and Uhrenholt97–Reference Washington, Wilson and Greaves104)} , satiety^{(Reference Thorning, Bertolt and Nielsen54,Reference Wanders, Mars and Borgonjen-van den Berg67,Reference Lawaetz, Blackburn and Bloom80,Reference Schwab, Louheranta and Torronen85,Reference Schwartz, Levine and Singh87,Reference Wanders, Feskens and Jonathan95,Reference Di Lorenzo, Williams and Hajnal98,Reference Flood-Obbagy and Rolls105–Reference Tiwary, Ward and Jackson108)} , fasting glucose^{(Reference Thorning, Bertolt and Nielsen54,Reference Wanders, Mars and Borgonjen-van den Berg67,Reference Veldman, Nair and Vorster77,Reference Schwab, Louheranta and Torronen85,Reference Sheehan, Wei and Ulchaker109–Reference Bell, Hectorn and Reynolds112)} , other hormonal responses related to glycaemic response^{(Reference Remer, Pietrzik and Manz29,Reference Wanders, Mars and Borgonjen-van den Berg67,Reference Lawaetz, Blackburn and Bloom80,Reference Sandhu, El Samahi and Mena84,Reference Schwartz, Levine and Weinstock86–Reference Shimoyama, Kusano and Kawamura88,Reference Di Lorenzo, Williams and Hajnal98,Reference Sirtori, Triolo and Bosisio110,Reference Gardner, Schwartz and Krista113)} , and haemoglobin A1C (HbA1C)/Homeostatic Model Assessment for Insulin Resistance (HOMA-IR)^{(Reference Schwab, Louheranta and Torronen85,Reference Sheehan, Wei and Ulchaker109,Reference Sirtori, Triolo and Bosisio110,Reference Gardner, Schwartz and Krista113)} (Figure 8). Overall outcomes found pectin to be related with a reduction in post-prandial blood glucose and insulin peaks, increased satiety and delayed gastric emptying in both healthy and non-healthy individuals. Study design consisted mainly of RCT crossover (n = 32), followed by non-randomised studies (n = 11) and RCT parallel (n = 5). Furthermore, most of the RCT crossover studies were single intake (n = 24).

Fifteen studies within this health topic were conducted in adults with various health complications. Participants with short bowel syndrome and jejuno-colonic anastomosis, when consuming ∼10 g/d of citrus pectin, did not experience differences in gastric emptying time^{(Reference Atia, Girard-Pipau and Hébuterne36)}. Three studies of hypercholesterolaemic participants^{(Reference Veldman, Nair and Vorster77,Reference Sirtori, Triolo and Bosisio110,Reference Bell, Hectorn and Reynolds112)} found no effect of pectin consumption on fasting glucose levels, nor on gastric emptying. However, Sirtori et al. (2012) found hypercholesteraemic participants consuming bars with casein + apple pectin had significant reductions in post-prandial insulin levels^{(Reference Sirtori, Triolo and Bosisio110)}. In a study of adults with obesity, it was reported that consuming 15 g/d of pectin with a test meal significantly delayed gastric emptying time and also increased satiety^{(Reference Di Lorenzo, Williams and Hajnal98)}. Outcomes in those with diabetes found some improvement in glucose tolerance related to pectin consumption^{(Reference Monnier, Pham and Aguirre82,Reference Schwartz, Levine and Weinstock86,Reference Vaaler, Hanssen and Aagenaes114)} , though some found no significant change with pectin consumption related to glycaemic control^{(Reference Williams, James and Evans96,Reference Sheehan, Wei and Ulchaker109,Reference Gardner, Schwartz and Krista113)} . The type of test meals consumed with pectin in the trials may lead to different results, especially in ‘non-healthy’ individuals. In the study by Vaaler et al. (1980), pure pectin lowered post-prandial blood glucose in insulin-dependent diabetics; however, the same effect was not seen in a barley/citrus fibre mix which included pectin along with other fibre types^{(Reference Vaaler, Hanssen and Aagenaes114)}, highlighting the impact of the difference in structures and composition of fibres, especially in those with diabetes. Other health complications included dumping syndrome^{(Reference Jenkins, Gassull and Leeds76,Reference Leeds, Ralphs and Ebied81,Reference Andersen, Holtug and Uhrenholt97)} and post-bariatric surgery^{(Reference Speth, Jansen and Lamers93)}. For those experiencing dumping syndrome, in which after eating there is a rapid emptying of bowel contents into the colon alongside and rebound hypoglycaemia, pectin was generally related to prolonged gastric emptying and reduction in hypoglycaemia^{(Reference Leeds, Ralphs and Ebied81)}. Two studies also tested pectin in healthy adults through enteral feeding and found that pectin increased viscosity of enteral nutrition as well as was related to reduced glucose ingestion^{(Reference Shimoyama, Kusano and Kawamura88,Reference Fuse, Bamba and Hosoda115)} .

These findings on glycaemic response have also been summarised by EFSA (2010) which states the finding of a ‘cause and effect relationship has been established between the consumption of pectin and reduction of post-prandial glycemic responses’, with at least 10 g of pectin per meal in adults⁽⁴⁾. Possible mechanisms of action related to pectin and controlled glycaemic responses are likely due to delayed gastric emptying, reduced sugar diffusion and absorption, and slowed release of digestion-related gut hormones^{(Reference Goff, Repin and Fabek116,Reference Guo, Hou and Cui117)} . However, many of the studies in this review related to post-prandial glucose response were single-intake studies. Future work is needed to identify relationships between long-term regulation of glycaemic response and pectin intake.

Summary for fat metabolism

There were thirty-four studies (reported in thirty-six publications) included in this review related to the health topic of fat metabolism, with publications dating from 1961 to 2022. Of these studies, twenty-five were published prior to 2000 (Figure 4). Pectin botanical origin tested in these studies was commonly citrus and non-defined sources (Figure 5). The dose of pectin tested in these studies ranged from 0.1 g/d to 50 g/d, with the average and median dose being 14.6 g/d and 15 g/d, respectively (Figure 6). The duration of the interventions was on average 28 d, though many were single-intake studies, and the median was 21 d (Figure 7).

Major outcomes studied in these publications were related to measures of blood cholesterol^{(Reference Brouns, Theuwissen and Adam9,Reference Miettinen and Tarpila20,Reference Challen, Branch and Cummings22,Reference Jenkins, Reynolds and Leeds26,Reference Remer, Pietrzik and Manz29,Reference Palmer and Dixon35,Reference Durrington, Manning and Bolton37,Reference Judd and Truswell40,Reference Kay and Stewart Truswell41,Reference Mao, Xiong and Hu44,Reference Stasse-Wolthuis, Albers and van Jeveren51,Reference Thorning, Bertolt and Nielsen54,Reference Vargo, Doyle and Floch55,Reference Veldman, Nair and Vorster74,Reference Veldman, Nair and Vorster77,Reference Schwab, Louheranta and Torronen85,Reference Sheehan, Wei and Ulchaker109–Reference Bell, Hectorn and Reynolds112,Reference Annunziata, Maisto and Schisano118–Reference Stasse-Wolthuis, Katan and Hermus125)} , blood triacylglycerol^{(Reference Brouns, Theuwissen and Adam9,Reference Miettinen and Tarpila20,Reference Challen, Branch and Cummings22,Reference Jenkins, Reynolds and Leeds26,Reference Durrington, Manning and Bolton37,Reference Judd and Truswell40,Reference Kay and Stewart Truswell41,Reference Thorning, Bertolt and Nielsen54,Reference Vargo, Doyle and Floch55,Reference Veldman, Nair and Vorster74,Reference Veldman, Nair and Vorster77,Reference Schwab, Louheranta and Torronen85,Reference Sheehan, Wei and Ulchaker109–Reference Bell, Hectorn and Reynolds112,Reference Cerda, Robbins and Burgin119,Reference Hillman, Peters and Fisher120,Reference Hoffmann, Linseisen and Riedl122–Reference Schwandt, Richter and Weisweiler124)} , fat absorption/excretion^{(Reference Miettinen and Tarpila20,Reference Ranganathan, Champ and Pechard28,Reference Atia, Girard-Pipau and Hébuterne36,Reference Judd and Truswell40,Reference Kay and Stewart Truswell41,Reference Ross and Leklem48,Reference Sandberg, Ahderinne and Andersson49,Reference Stasse-Wolthuis, Albers and van Jeveren51,Reference Fuse, Bamba and Hosoda115,Reference Bosaeus, Carlsson and Sandberg126,Reference Isaksson, Lundquist and Akesson127)} , bile acid metabolism^{(Reference Miettinen and Tarpila20,Reference Judd and Truswell40,Reference Kay and Stewart Truswell41,Reference Ross and Leklem48,Reference Stasse-Wolthuis, Albers and van Jeveren51,Reference Hillman, Peters and Fisher121,Reference Bosaeus, Carlsson and Sandberg126)} and blood pressure^{(Reference Lau, Liu and Paniagua34,Reference Stasse-Wolthuis, Albers and van Jeveren51,Reference Thorning, Bertolt and Nielsen54,Reference Veldman, Nair and Vorster77,Reference Schwab, Louheranta and Torronen85,Reference Sirtori, Triolo and Bosisio110–Reference Bell, Hectorn and Reynolds112)} (Figure 8). Most studies in this health topic were non-randomised study designs (n = 14), followed by RCT crossover (n = 11) and RCT parallel design (n = 9). Furthermore, fourteen studies tested pectin consumption in non-healthy adults and elderly populations, with obesity^{(Reference Annunziata, Maisto and Schisano118)}, diabetes^{(Reference Sheehan, Wei and Ulchaker109)}, gut-related symptoms^{(Reference Atia, Girard-Pipau and Hébuterne36,Reference Sandberg, Ahderinne and Andersson49,Reference Bosaeus, Carlsson and Sandberg126)} , hypercholesterolaemia/hypertension^{(Reference Miettinen and Tarpila20,Reference Lau, Liu and Paniagua34,Reference Veldman, Nair and Vorster74,Reference Veldman, Nair and Vorster77,Reference Sirtori, Triolo and Bosisio110,Reference Bell, Hectorn and Reynolds112,Reference Schwandt, Richter and Weisweiler124)} , pancreatic insufficiency^{(Reference Isaksson, Lundquist and Akesson127)} or severe stroke^{(Reference Mao, Xiong and Hu44)}.

Of the thirty-four studies related to fat metabolism, twenty-eight reported positive effects of pectin on lowering cholesterol, triacylglycerol or fat absorption. Brouns et al. (2012) examined the cholesterol-lowering properties of pectin derived from various botanical sources with differing DE and molecular compositions^{(Reference Brouns, Theuwissen and Adam9)}. Here, both citrus (DE 70%) and apple pectin (DE 70%) were most efficacious in lowering cholesterol, compared with citrus and apple pectin of low MW and DE, indicating pectin cholesterol-reduction properties to be dependent on molecular structure.

Several studies that measured cholesterol and triacylglycerol in relation to pectin consumption found that pectin was related to serum cholesterol reduction but had no effect on serum triacylglycerol^{(Reference Miettinen and Tarpila20,Reference Challen, Branch and Cummings22,Reference Ranganathan, Champ and Pechard28,Reference Remer, Pietrzik and Manz29,Reference Atia, Girard-Pipau and Hébuterne36,Reference Durrington, Manning and Bolton37,Reference Judd and Truswell40,Reference Kay and Stewart Truswell41,Reference Sandberg, Ahderinne and Andersson49,Reference Sheehan, Wei and Ulchaker109,Reference Annunziata, Maisto and Schisano118–Reference Hillman, Peters and Fisher120)} . These results were found conclusive by EFSA Scientific Opinion (2010), and it has since been stated that ‘a cause-and-effect relationship has been established between the consumption of pectin and maintenance of normal blood cholesterol concentrations’, with at least 6 g/d of pectin in one or more serving in adults⁽⁴⁾. This is likely due to the increased gut viscosity from pectin, which limits the reabsorption of bile acids, increases cholesterol elimination in faeces as bile acids, upregulates bile acid synthesis from cholesterol and decreases levels of circulating cholesterol^{(Reference Gunness and Gidley128)}.

Six studies measured blood pressure in relation to pectin intake. However, only one of the studies demonstrated a decrease in blood pressure after 16 g/d of sugarbeet pectin for 84 d^{(Reference Schwab, Louheranta and Torronen85)}, whereas the others evaluated lower doses (0.1–15 g/d) of supplementation and did not show any changes to blood pressure^{(Reference Stasse-Wolthuis, Albers and van Jeveren51,Reference Thorning, Bertolt and Nielsen54,Reference Sirtori, Triolo and Bosisio110–Reference Bell, Hectorn and Reynolds112)} .

Summary for bioavailability of micronutrients

Twenty human intervention studies (reported in twenty-two publications) investigating the effect of pectin on the bioavailability of micronutrients were included in this review. Most of these studies evaluated citrus pectin (n = 10), apple pectin (n = 6) or pectin from a non-defined origin (n = 4) (Figure 5). The doses of pectin tested in studies ranged from 0.1 g/d to 36 g/d (Figure 6), with an average and median tested dose of 14.2 g/d and 11.5 g/d, respectively. The average duration of pectin intervention was 12.4 d, though the median was 1 d, as eleven of the twenty-two studies were 1-d, single-intake studies (Figure 7). Most frequently evaluated outcomes were iron bioavailability (n = 9; ten publications)^{(Reference Sandberg, Ahderinne and Andersson49,Reference Bell, Hectorn and Reynolds112,Reference Monnier, Colette and Ribot129–Reference Grudeva-Popova and Sirakova136)} , calcium bioavailability (n = 4)^{(Reference Cummings, Southgate and Branch23,Reference Sandberg, Ahderinne and Andersson49,Reference Eliaz, Hotchkiss and Fishman133,Reference Grudeva-Popova and Sirakova136)} (Figure 8) and other minerals bioavailability: copper (n = 4 non-randomised studies)^{(Reference Sheehan, Wei and Ulchaker109,Reference Lei, Davis and Fang130,Reference Eliaz, Hotchkiss and Fishman133,Reference Grudeva-Popova and Sirakova136)} ; zinc (n = 4 non-randomised studies)^{(Reference Sandberg, Ahderinne and Andersson49,Reference Sheehan, Wei and Ulchaker109,Reference Lei, Davis and Fang130,Reference Drews, Kies and Fox137)} ; magnesium (n = 3 non-randomised and n = 1 RCT crossover studies)^{(Reference Sandberg, Ahderinne and Andersson49,Reference Sheehan, Wei and Ulchaker109,Reference Grudeva-Popova and Sirakova136,Reference Drews, Kies and Fox137)} ; sodium (n = 2 non-randomised studies)^{(Reference Sandberg, Ahderinne and Andersson49,Reference Grudeva-Popova and Sirakova136)} . Fifteen studies were conducted with healthy adults and five with non-healthy adults: idiopathic hemochromatosis (n = 1)^{(Reference Monnier, Colette and Ribot129,Reference Monnier, Colette and Aguirre131)} , ileostomy subjects (n = 1)^{(Reference Sandberg, Ahderinne and Andersson49)}, diabetes (n = 1)^{(Reference Sheehan, Wei and Ulchaker109)} and hypercholesterolaemia/hyperlipoproteinaemia (n = 2)^{(Reference Bell, Hectorn and Reynolds112,Reference Grudeva-Popova and Sirakova136)} .

One study found lower apparent iron absorption from the small intestine in six subjects with ileostomy^{(Reference Sandberg, Ahderinne and Andersson49)}, and one single-intake study in patients with idiopathic haemochromatosis exhibited a significant drop in fractional absorption of inorganic iron compared with cellulose^{(Reference Monnier, Colette and Ribot129,Reference Monnier, Colette and Aguirre131)} . On the contrary, seven other studies conducted in healthy or hypercholesterolaemic/lipoproteinaemic subjects found no effect on iron balance at daily doses between 0.25 and 15 g/d and up to 84 d of supplementation^{(Reference Bell, Hectorn and Reynolds112,Reference Lei, Davis and Fang130,Reference Cook, Noble and Morck132–Reference Grudeva-Popova and Sirakova136)} . Similarly, no significant effect on calcium balance was reported by the four studies that investigated it with a daily dose of pectin between 15 and 36 g/d and up to 84 d of supplementation^{(Reference Cummings, Southgate and Branch23,Reference Sandberg, Ahderinne and Andersson49,Reference Eliaz, Hotchkiss and Fishman133,Reference Grudeva-Popova and Sirakova136)} . There was also no effect on the balance of the other evaluated minerals mentioned above. In general, pectin dietary supplementation had no impact on mineral balance in humans. There is a possibility that pectin might bind with some minerals in foods (e.g. Ca²⁺) or in the small intestine and limit their local absorption, with a possible influence of the degree of methylation on this phenomenon^{(Reference Wang, Munk and Skibsted138)}, but this is not confirmed by human intervention studies^{(Reference Sandberg, Ahderinne and Andersson49,Reference Jaramillo, Molina and Briones135,Reference Drews, Kies and Fox137)} , even at high doses such as 36 g/d of HM pectin (72%) for 42 d^{(Reference Cummings, Southgate and Branch23)} or in sensitive population like preterm infants (0.085% in human milk fortifier for 28 d)^{(Reference Moya, Sisk and Walsh139)}. As pectin is completely fermented in the large intestine and metabolised into SCFA, this can enhance the absorption of minerals from the colon, as it has been demonstrated for other non-digestible fermentable carbohydrates and prebiotics^{(Reference Beggs, Bhullar and Dimke140)}. The potential of colonic absorption estimated at approximately 10% for calcium in humans might fully compensate for the potential reduction in the small intestine and explain the neutral impact on mineral balance.

Contrary to the studies that have evaluated mineral balances, intervention studies aiming to evaluate the potential impact of pectin-derived ingredients on vitamins and one type of flavonoid (i.e. quercetin), are all single-intake studies, except one that was short-term (7 d), and all were randomised crossover (Figure 8). Two single-intake studies found no significant impact of pectin on the mean serum concentration of α-tocopherol (vitamin E) when it was consumed with pectin (0.1 g of apple pectin or 8.9 g of HM citrus pectin)^{(Reference Hoffmann, Linseisen and Riedl122,Reference Horwitt, Elliott and Kanjananggulpan141,Reference Riedl, Linseisen and Hoffmann142)} . One study reported a reduction of urinary excretion of ascorbic acid (vitamin C) in healthy subjects with intake of citrus pectin for 7 d, and the reduction tended to be more important with 14.2 g/d than 4.2 g/d^{(Reference Keltz, Kies and Fox143)}. Another study demonstrated that a single intake of 15 g of apple pectin could increase serum vitamin A levels versus an intake of vitamin A without fibre^{(Reference Kasper, Rabast and Fassl144)}. On the contrary, the mean plasma increase of β-carotene for the next 48 h after a single intake of 12 g of citrus pectin mixed with food was lower than with the same meal without pectin^{(Reference Rock and Swendseid145)}. Another single-intake study in six healthy adults showed a lower relative increase in plasma concentrations (24-h area under the curve) of β-carotene, lycopene, and lutein with almost 9 g of HM citrus pectin or other fibres such as guar gum, alginate, cellulose and wheat bran than without dietary fibre^{(Reference Riedl, Linseisen and Hoffmann142)}. The possible interaction between pectin and bile salts may create a transient physical barrier that limits lipase to access the lipid surface and interferes with the formation of micelles that are needed for the absorption of hydrophobic compounds such as β-carotenes^{(Reference Cervantes-Paz, Ornelas-Paz and Ruiz-Cruz146)}, but interestingly, this does not seem to happen for the vitamins A and E, which are also hydrophobic. Pectin seemed to impact vitamin A bioavailability positively and had no effect on vitamin E, as mentioned above.

In an experiment, LM citrus pectin was used to coat folic acid-fortified rice and, thus, protect the folic acid during the washing and cooking of the rice. Despite a lower bioavailability of the folic acid with the pectin coating (5% w/w, equivalent to 10 g), the coating helped to protect folic acid during washing and cooking, still providing a slightly higher folic acid intake than without coating^{(Reference De Ambrosis, Vishnumohan and Paterson147)}. Finally, one study in nineteen healthy adults reported higher 24-h urinary excretion of quercetin and its metabolites after a single intake taken via a drink containing 2 g of HM apple pectin versus the same drink without added pectin. In another study reported in the same publication and conducted in a subgroup of six subjects, a dose–effect relationship (between 0.6 and 2 g) was suggested, and a greater effect of HM than LM apple pectin was found^{(Reference Nishijima, Takida and Saito148)}. It is difficult to conclude the effects of pectin on various vitamins and polyphenols from human studies, but the structure–function relationship is also thought to play a role there, as suggested by in vitro studies^{(Reference Cervantes-Paz, Ornelas-Paz and Ruiz-Cruz146)}. In addition, it is important to consider that the composition of food matrices and the conditions of intake will also influence the role of pectin on the bioavailability of micronutrients. A recent in vitro study with Caco-2 cells showed that the reduced bioaccessibility of ferulic acid and naringenin (a citrus flavonoid) seen with an HM apple pectin was completely counteracted by the concomitant intake of sucrose or olive oil^{(Reference Kruger, Sus and Frank149)}.

Summary for immune health

In total, eight studies (reported in nine publications) were included in the scoping review related to immune health outcomes. Studies that evaluated immune health-related outcomes were the recent studies ranging from 2008 to 2022 (Figure 4). Three of the studies used RG-I from carrots^{(Reference Lutter, Teitsma-Jansen and Floris150–Reference McKay, Teitsma-Jansen and Floris152)} and one RG-I from peels of Korean citrus hallabong^{(Reference Lee, Kim and Kwak111)} and were performed with very low doses (0.1–1.5 g/d, Figures 5 and 6). The maximum tested dose was generally lower (mean 6.5 g/d, median 5.4 g/d) than for other health outcomes, and the duration of dietary supplementation had a mean of 31 d and a median of 28 d of intervention (Figures 6 and 7). All studies were RCT parallel studies^{(Reference Mao, Xiong and Hu44,Reference Nakamura, Inokuchi and Fukushima45,Reference Miyazawa, Tomomasa and Kaneko63,Reference Wanders, Mars and Borgonjen-van den Berg67,Reference Sirtori, Triolo and Bosisio110,Reference Lee, Kim and Kwak111,Reference Lutter, Teitsma-Jansen and Floris150–Reference McKay, Teitsma-Jansen and Floris152)} (Figure 8). Four studies related to enteral feeding evaluated the frequency of infections and general health status with pectin (5–16 g/d)^{(Reference Nakamura, Inokuchi and Fukushima45,Reference Xi, Xu and Tan68,Reference Lutter, Teitsma-Jansen and Floris150,Reference McKay, Teitsma-Jansen and Floris152)} . Major immune-related health outcomes studies included immunostimulatory markers such as IL-6, soluble intracellular cell adhesion molecule-1 (siCAM), high-sensitivity C-reactive protein (Hs-CRP), IL-10, IL-1β and TNF-α (n = 3)^{(Reference Sirtori, Triolo and Bosisio110,Reference Lee, Kim and Kwak111,Reference McKay, Oranje and Helin151)} ; lymphocyte count (n = 1)^{(Reference Mao, Xiong and Hu44)}; and cases of gastroesophageal reflux disease (GERD) in children with cerebral palsy (n = 1)^{(Reference Miyazawa, Tomomasa and Kaneko63)}. General outcomes found pectin related to enhanced immune response and reduced disease symptoms. For instance, carrot-derived RG-I accelerated an innate immune response and reduced symptoms of an acute viral infection with rhinovirus^{(Reference Lutter, Teitsma-Jansen and Floris150)}. In another study, pectin was used in a therapy for children with cerebral palsy to test effectiveness in decreasing vomiting and chronic respiratory symptoms^{(Reference Miyazawa, Tomomasa and Kaneko63)}. Here, it was observed that children on the high-pectin diet (2:1 enteral formula to pectin liquid) had significantly decreased reflux episodes and vomiting and decreased cough and respiratory symptoms.

Summary of other health-related outcomes

Twenty-one studies were categorised as ‘other topics’. Of these studies, the dose of pectin under investigation ranged from 0.6 g/d to 36 g/d (Figure 6), with an average and median tested dose of 12.4 g/d and 10 g/d, respectively. The average duration of pectin intervention was 35 d, and the median was 21 d (Figure 7). The following health-related outcomes in this category include detoxification from radioactivity (n = 4)^{(Reference Bandazhevskaya, Nesterenko and Babenko153–Reference Nesterenko, Nesterenko and Babenko156)}, heavy metals (n = 3)^{(Reference Eliaz, Hotchkiss and Fishman133,Reference Tekutskaya157,Reference Zhao, Liang and Fan158)} , cancer-related physiological markers (n = 2)^{(Reference Keizman, Frenkel and Peer33,Reference Azemar, Hildenbrand and Haering159)} , bioavailability of drugs (n = 2)^{(Reference Issy, Lanchote and De Carvalho160,Reference Kasper, Zilly and Fassl161)} , and infant growth parameters (n = 2)^{(Reference Moya, Sisk and Walsh139,Reference Doherty and Jackson162)} and another eight studies investigated various markers related to metabolic and cardiovascular health^{(Reference Challen, Branch and Cummings22,Reference Cummings, Southgate and Branch23,Reference Lau, Liu and Paniagua34,Reference Annunziata, Maisto and Schisano118,Reference Andrews, Fernandes and Brownmiller163–Reference Sommer and Kasper166)} (Figure 8).

Three out of four studies showed a reduction of ¹³⁷Cs in populations previously exposed to radioactivity (Figure 8) when they were given apple pectin (3–10 g/d) for 14–28 d^{(Reference Bandazhevskaya, Nesterenko and Babenko153–Reference Nesterenko, Nesterenko and Babenko156)}. The authors hypothesised that pectin, by binding caesium in the gastrointestinal (GI) tract, can prevent its absorption into the systemic circulation and increase its faecal excretion^{(Reference Hill, Schlager and Vogel154)}; thus, it might be more efficient in case of exposure to contaminated foods than to reduce the contamination level of people exposed via other means^{(Reference Kinoshita, Tsubokura and Katsuragi155)}. Similarly, three single-arm studies also showed the potential of citrus-derived pectin (15 g/d of hydrolysed pectin or 3.6 g/d of HM pectin) to reduce blood levels of toxic elements such as heavy metals in less than 20 d^{(Reference Cook, Noble and Morck132,Reference Tekutskaya157,Reference Zhao, Liang and Fan158)} . Authors proposed that the RG-II part of pectic polysaccharides can chelate very specifically some cations like Pb²⁺ but not essential ones like Mg²⁺, Zn²⁺, Fe²⁺ and Fe^{3+(Reference Eliaz, Hotchkiss and Fishman133,Reference Tahiri, Tressol and Doco167)} . However, there is still debate whether this mechanism can only happen in the gut and contribute to preventing exposure from contaminated foods^{(Reference Tahiri, Tressol and Doco167)} or if it can also help in case of previous exposure that would require a partial absorption of pectin molecule before its fermentation by the gut microbiota^{(Reference Eliaz, Hotchkiss and Fishman133)}.

Two single-arm studies suggested that approximately 15 g/d of hydrolysed citrus pectin could improve quality of life in patients with cancer and might limit progression of prostate-specific antigen level in blood^{(Reference Keizman, Frenkel and Peer33,Reference Azemar, Hildenbrand and Haering159)} . Two single-intake, crossover studies evaluated the bioavailability of drugs (digoxin and valproic acid) given with pectin (5 g of non-defined pectin and 14 g of HM citrus pectin, respectively) and did not find a reduced appearance of the molecules in the blood, but possibly a delayed arrival^{(Reference Issy, Lanchote and De Carvalho160,Reference Kasper, Zilly and Fassl161)} .

Two RCT parallel trials have evaluated the use of pectin in nutritional solutions dedicated to preterm infants^{(Reference Moya, Sisk and Walsh139)} or severely malnourished young children in India^{(Reference Doherty and Jackson162)}. While the first one showed no effect or slight improvement of the growth parameters after 28 d, the second showed that using pectin at 10 g/kg/d in replacement of arachis oil was not relevant to accelerate weight gain in these conditions.

The possible protection of dietary fibres in the context of cardiovascular health does not seem to be related to platelet function or haemostasis, as suggested by a study where HM pectin was given at a daily dose of 36 g for 42 d (wheat bran was also tested with no significant effect)^{(Reference Challen, Branch and Cummings22)}. Similarly, in an RCT parallel study, dietary supplementation with 9.6 g hydrolysed citrus pectin for 168 d did not influence various collagen markers possibly involved in cardiac fibrosis^{(Reference Lau, Liu and Paniagua34)}. An RCT parallel study did not show any additional benefit of 0.6 g/d pectin for 56 d to reduce further serum level of trimethylamine-N-oxide (TMAO) compared with grape pomace polyphenols alone^{(Reference Annunziata, Maisto and Schisano118)}. Another RCT parallel study evaluated a 12-week supplementation with 8 g hydrolysed citrus pectin without showing any significant effect on knee osteoarthritis severity and pain management^{(Reference Andrews, Fernandes and Brownmiller163)}. Another three single-intake studies have evaluated the intake of 2 g of apple pectin on flushing symptoms after niacin intake^{(Reference Moriarty, Backes and Dutton165)} and on the release of pancreatic GI hormones and endogenous methanol production after the consumption of 10–15 g of pectin (non-described)^{(Reference Lindinger, Taucher and Jordan164,Reference Sommer and Kasper166)} .

Structure–function relationships

A remaining gap throughout this review of the literature was evident regarding the structure–function relationship of pectin. Pectin structural characteristics such as DE, MW and other structural regions (homogalacturonan, xylogalacturonan, RG-I, RG-II) largely influence the functional properties of pectin after ingestion. These compositions dictate the health-related outcomes of pectin consumption. For instance, gut microbiota access and ability to metabolise these pectin structures depends on traits such as DE, the composition of neutral sugars and the degree of structure branching (i.e. the ratio between the linear homogalacturonan and the branched RG-I and -II). Therefore, gut microbiome composition changes, production of SCFA and other metabolite production is dependent on these structural characteristics^{(Reference Pascale, Gu and Larsen11)}.

Furthermore, while many current studies of pectin demonstrate post-prandial glucose reduction and maintenance of normal blood cholesterol, there is little investigation into the structure of pectin exuding these effects. Much of this gap is due to the lack of pectin type reported in publications. In this review, for instance, the majority of pectin botanical sources was ‘non-defined’. Such dependencies on pectin structure with health outcomes were demonstrated in a compelling study by Wanders et al. (2014), when comparing non-fibre control with gelled, bulking and viscous forms of pectin, which found that only gelled pectin (with high viscosity and water holding capacity) significantly improved satiety, as well as reduced glucose and insulin peaks^{(Reference Wanders, Feskens and Jonathan95)}. Such findings demonstrate the important physico-chemical properties of pectin when examining health outcomes.

Similarly, many of the studies discussed here related to fat metabolism did not specify pectin structure. A better understanding of how MW and DE influence gastric emptying times and gut viscosity could, therefore, allow for a more precise pectin structure target in terms of capability to lower total and LDL-cholesterol and triacylglycerol. For example, in previous animal studies, higher DE pectin had greater effects on decreasing triacylglycerol and increased bile acid secretion^{(Reference Dongowski and Lorenz168,Reference Trautwein, Kunath-Rau and Erbersdobler169)} . Much has come to light on the role of the structure of pectin and its ability to lower blood cholesterol. For instance, a study by Brouns et al. (2012) found that pectin of high MW and high DE (from citrus or apple) was more effective in terms of cholesterol reduction^{(Reference Brouns, Theuwissen and Adam9)}. Such results have been documented in animal studies demonstrating the relationship between the molecular structure of pectin and blood cholesterol^{(Reference Yamaguchi, Uchida and Watabe170,Reference Terpstra, Lapré and de Vries171)} . As found in this scoping review, human clinical studies have limited documentation on the types of pectin used, and therefore, results are less consistent.

While pectin has been demonstrated to have various immunomodulatory effects in in vitro and animal studies, including RG bioactivity and the ability to suppress inflammatory markers such as IL-1β and IL-6^{(Reference Chen, Liu and Zhang172,Reference Ishisono, Yabe and Kitaguchi173)} , only recently has this work been studied in humans. Interestingly, many of the immune studies in vitro do take into account the structure and source of pectin, as well as DE^{(Reference Salman, Bergman and Djaldetti174)}. Other structural characteristics such as RG-I regions also influence the functional properties of pectin. For example, McKay et al. (2021), analysed RG-I extracted from bell peppers and carrots and classified the composition and structure before testing in humans^{(Reference McKay, Oranje and Helin151)}. This prior analysis determined that the immunostimulatory effects were due to this specific type of RG-I backbone more so than the type of side chain and also confirmed safety and acceptability in humans. Lutter et al. (2021) and McKay et al. (2022) then tested this carrot-derived RG-I in a clinical trial and found that consumption of this pectin accelerated innate immune responses and decreased common cold symptoms^{(Reference Lutter, Teitsma-Jansen and Floris150,Reference McKay, Teitsma-Jansen and Floris152)} . Lee et al. (2016) similarly tested RG though derived from citrus pectin and also found improved immune function^{(Reference Lee, Kim and Kwak111)}. However, four of the nine studies in the immune response category listed the pectin source as ‘non-defined’. This illustrates the structure–function relationship of pectin being largely unexplored in humans, though extremely important in determining bioactivity potential. There is, therefore, a need for further investigation into pectin molecular weights, DE, RG-I, side chains and monosaccharide components in relation to health outcomes under controlled experiments in humans^{(Reference Beukema, Faas and de Vos10)}.

Possible side effects

The highest dose of pectin tested was 50 g/d for 14 d in adults^{(Reference Miettinen and Tarpila20)}, which did not report any adverse side effects. The longest studies evaluated pectin consumption at various doses for 168 d in adults who were healthy (n = 1) and non-healthy (n = 2 gut-related, n = 1 cancer and n = 1 hypertension), with no adverse side effects reported. In infants, the sustained consumption of 1 g/d of citrus pectin oligosaccharides for 127 d^{(Reference Magne, Hachelaf and Suau30)} also did not report any significant adverse reactions.

The most frequently reported side effects included GI symptoms such as flatulence or bloating, most classified as mild and stopped just after consumption or resolved by adaptation after a few days. These symptoms were typical for dietary fibre^{(Reference Fleming and Rodriguez25,Reference Marthinsen and Fleming27,Reference Hillman, Peters and Fisher39)} and do not prevent the use of pectin in sensitive populations such as hospitalised patients with enteral feeding. Interestingly, digestive symptoms, such as diarrhoea, were generally improved with pectin^{(Reference Mao, Xiong and Hu44,Reference Nakamura, Inokuchi and Fukushima45,Reference Maruyama, Goshi and Kashima62,Reference Schultz, Ashby-Hughes and Taylor65,Reference Xi, Xu and Tan68)} .

Similarly to what occurs after the consumption of fruits, an endogenous production of methanol is produced when HM pectin (DE of 75% such as in apples) is fermented by bacteria inhabiting the large intestine^{(Reference Lindinger, Taucher and Jordan164)}. However, the experimental model was very specific as study subjects had to consume ethanol to maintain a relatively high level of blood ethanol to annihilate methanol catabolism. As highlighted by EFSA⁽¹⁵⁾, no sign of liver toxicity was noticed in the study of Cummings et al. (1979) in which five subjects consumed 36 g/d of HM pectin (DE 72%) for 42 d, and on the contrary, a recent review highlighted a promising role of pectin to limit the progression of liver damage in the context of alcoholic and non-alcoholic liver disease as currently illustrated by pre-clinical data^{(Reference Hu, Cassard and Ciocan175)}.

Study limitations

As mentioned previously, many studies included in this scoping review neglected to include details of pectin botanical origin and structure. Without this information, it is difficult to draw conclusions between pectin intake and health-related functional outcomes, as many of these health effects are directly related to the physico-chemical properties and molecular composition of pectin. Also, some studies were performed with pure pectin products, whereas others used commercially available pectin solutions, which, most of the time, are standardised with sugar to provide a specific functionality in foods and beverages. This may lead to overestimating the tested doses as reported in the publications. Furthermore, due to the nature of a scoping review, the review presented here is an overview of the existing literature. Individual studies included were not subjected to critical quality assessment and risk of bias. A rigorous systemic review of specific health topics outlined here would make it possible to draw conclusions on the health effects of pectin in human intervention studies.