In many regions of the world, grain legumes, also referred to as pulses, are neglected staple foods that have the potential to directly impact food insecurity while helping individuals achieve healthy and balanced diets and to accelerate progress towards the Sustainable Development Goals of the WHO/FAO [1
]. While pulses are frequently promoted for their content of protein and micronutrients, they are also an excellent source of dietary fiber, a fact that is under appreciated [6
]. Since the seminal work of Burkitt and colleagues, there has been an awareness of the role of dietary fiber in maintaining the physiological function of the mammalian intestine [7
]. However, recognition of the importance of dietary fiber for human health and well-being has never risen to the establishment of fiber as an essential nutrient. Nonetheless, a recommended level of dietary fiber has been proposed (14 g/1000 kcal) and many systematic reviews and meta-analyses support the health benefits associated with cohorts that meet or exceed this recommendation [8
]. Unfortunately, the majority of individuals in developed countries such as the United States fail to meet the recommended level of intake for dietary fiber and the magnitude of the gap is large, an approximately 50–70% shortfall [6
]. This gap has been resistant to change, despite decades of public health interventions and development of fiber enriched food products. Recently, we have advocated more focus on pulse crops to not only close the dietary fiber gap, but to extend intake above recommended levels to those that have been reported to significantly reduced chronic disease mortality [14
]. However, to achieve such intakes, i.e., 50–80 g total dietary fiber per day, we argue that a better understanding of the effects of high fiber foods such as pulses is needed. We recently published a comparative analysis of the dietary fiber content of chickpea, common bean, dry pea, and lentil, and reported that differences existed among pulses that may be of importance to various sectors of the food systems industry [6
]; however, from the consumers’ perspective, none of the differences observed negate the conclusion that pulses are an abundant source of dietary fiber.
Of the many functions that dietary fiber may affect in mammalian species, there is considerable interest in its influence on the meta-organism, i.e., the gut-associated microbiome [16
]. The available literature on this topic is expanding rapidly, e.g., [21
]; however, we found no studies in which the effects of the four most predominant pulses have been compared. This is important for several reasons. First, there is no consensus on what component(s) of dietary fiber are essential for gut health (reviewed in [6
]), and consequently, various reports of dietary fiber composition [21
] cannot be used to predict gut health effects. Second, a considerable body of literature indicates that gut health effects are attributable to microorganisms that are obligate anaerobes. Thus, except for one report in which microbial analyses were performed on cecal content and stool [21
], analyses of the microbiome have been limited to fecal material which may not be directly representative of the profile of microorganisms that colonize the intestinal tract of pulse fed animals/humans and that are responsible for effects of the microbiome on health status. Third, it is widely recognized that results of microbiome analyses can vary widely among publications for a host of technical reasons unrelated to the underlying biology and biochemistry being investigated [29
]; this makes direct comparisons among pulses within the same study of considerable value. Finally, while investigation of cultivars of various pulses is very important, it is equally important to also move the field in a translational direction, i.e., the analysis of each pulse crop as a collection of cultivars that are agronomically important and as might be eaten by the consumer, an approach that agricultural economists refer to as a market basket. In view of these gaps in research, the study reported herein was designed to address three questions when pulses were fed at a similar level of dietary protein and fiber. Do the four pulses: (1) exert the same effect on specific microbial populations that have been implicated in health promotion/disease prevention, (2) impact intestinal function in the same manner, and (3) affect adipose tissue deposition in a mouse model of diet-induced obesity?
Foods are considered dietary staples when they are eaten in large amounts throughout the day as a primary source of calories and protein. These foods are also affordable and accessible to the populations in which they serve as dietary staples. Grain legumes, also referred to as pulses, have been dietary staples since the dawn of agriculture [15
]. However, in recent years a growing trend has been to replace them with animal derived foods and protein isolates [64
]. Consequently, in many developed countries, pulses are now consumed infrequently by most of the population [2
]. There are four pulse crops that are predominant in the world’s food supply: common bean, chickpea, dry pea, and lentil [15
]. While most attention is given to pulses because they supply a large amount of protein in the absence of lipid, they are also an exceptional source of dietary fiber, containing 2–3 times more total fiber per 100 kcal edible portion than other foods commonly promoted as rich sources of dietary fiber [6
Many barriers have obstructed progress in understanding how dietary fiber impacts human health. Prominent among these issues has been the lack of an internationally accepted consensus definition of dietary fiber and of an analytical approach that measures what is contained in that definition [68
]. Consequently, the publication of a consensus definition and method in 2009 and 2011, respectively, was a landmark development in the field [69
]. The definition and method divide dietary fiber components into three major categories: insoluble dietary fiber, soluble dietary fiber, and oligosaccharides. The variation among foods in these components as well as the investigation of human consumption in terms of these dietary fiber categories has been used to reflect differences in dietary fiber quality, but their association with health outcomes is controversial, and most recommendations continue to be framed in terms of total dietary fiber intake [6
]. Nonetheless, given our recent work indicating that common bean, chickpea, dry pea, and lentil have similar levels of total dietary fiber [25
], we decided that a comparative investigation of these pulse would provide an opportunity to compare the effects of similar amounts of dietary fiber from botanically similar but distinct sources. As discussed in the following sections, what is unknown is whether the bioactivity of pulses is equivalent.
A number of recent studies have reported the effects of the consumption of various pulse crops on the fecal microbiome in preclinical models and numerous beneficial changes have been reported, although comparative studies among pulses have not been undertaken [19
]. A major limitation of efforts to study the gut-associated microbiome in fecal specimens is that the majority of the commensal microbial species that have important effects are obligate anaerobes and the stability of these populations in feces is variable and appears to be short [73
]. Recognizing this situation, and the report of differences between the cecal and fecal microbiome in mice fed chickpea [21
], the bacterial analyses reported herein were made on the content of the cecum that was excised intact and immediately frozen in liquid nitrogen to preserve the anaerobic status of its content. The cecum is a segment of the gut that supports a diverse population of commensal microorganisms. Because the cecum sits at the intersection of the small and large intestine, its microbial content reflects the impact of dietary constituents that escape digestion in the small intestine. Consistent with expectations of a high fiber diet, rich in insoluble and soluble fermentable carbohydrates, the overall content of bacteria in the cecum was similarly increased by all pulses evaluated. Differential effects were also observed for the specific bacterium Akkermansia munciniphila.
While lentil and common bean induced, respectively, a 49-fold and 25-fold increase in this bacterium relative to the high-fat control diet, chickpea and dry pea had no effect. The relevance of A. muciniphila
is that colonization of the gut with this bacterium has been reported to be inversely associated with obesity, diabetes, and inflammation [74
]. The differential effects of pulse consumption on the content of A. muciniphila
is consistent with a selective effect limited to lentil and common bean. Given the emerging evidence of A. muciniphila’s
health benefits, the identification of the beneficial prebiotic components of pulse crops is essential for at least two reasons. First, for lentil and common bean, in which the effect on A. muciniphila
ranged between a 25- and 49-fold increase (Figure 2
B), knowledge of prebiotic components could not only lead to the identification of cultivars within each crop with even stronger colony promoting activity, but could also prevent the loss of those compounds during food processing and the development of new food products. We also observed that pulses had differential effects on relative levels of bacteria in the phyla Bacteriodetes (increased) and Firmicutes (decreased). Given the controversial nature of the literature indicating whether an increase in the Bacteroidetes to Firmicutes ratio is consistent with health benefits [52
], the importance of this observed is unclear. Nonetheless, the rank order of increase in comparison to the high-fat control diet was lentil (5-fold) > common bean (3.9-fold) > chickpea (2.9-fold) > dry pea (2.3-fold). For dry pea and chickpea, there are over 10,000 genetic variants of these pulses in various breeding programs and germplasm collections around the world [85
]. Knowledge of the prebiotics that account for health-beneficial effects would provide an opportunity to guide development of cultivars of these crops that have more favorable gut health properties than those studied in this experiment.
Since differences in microbial populations were observed when total dietary fiber levels were similar, it suggests an effect due to some aspect of dietary fiber quality. Because small molecules are known to be associated with dietary fiber, the small molecule profile of the pulses that were fed was evaluated and differences in composition were observed in classes of chemical compounds that have limited bioavailability in mammals and that microbes are known to metabolize [86
]. For example, lentil was a distinct source of 3-fucosyllactose, which is an oligosaccharide identified in human milk and that has distinct beneficial effects on the human intestine. Of equal interest was the dominant presence of gulonic acid, a known substrate for bacteria in the Bifidobacterium
genus. It is also likely that different carbohydrate moieties included within the framework of total dietary fiber vary among pulses in a manner that supports the propagation of different microbial colonies within the gut.
There is a growing awareness of the value of increasing pulse consumption [1
], in part because this will close the dietary fiber gap and decrease the risk for chronic diseases. However, as we recently reviewed [88
], consumers have concerns about food tolerance that need to be addressed. The level of pulse consumption studied was intentionally high, i.e., approximately twice as high as median intakes of pulse consuming subgroups in the US and Canada. However, the dietary level was actually similar to dietary amounts consumed in developing countries where pulses are dietary staples. Moreover, the level incorporated into the diet (approximately 18g total dietary fiber/1000 kcal) would achieve levels of dietary intake in humans shown to reduce cardiometabolic disease related mortality by over 50% in the AARP cohort [14
]. We examined several markers of intestinal function in either the ileum and/or the ascending, transverse, and descending segments of the colon. Crypt height, the rate of cell proliferation and its location within the crypt, and the amount of alcian blue positive mucin within goblet cells lining the intestinal lumen were assessed. The limited changes observed that were not statistically significant after adjustment for false discovery. This finding is consistent with our previous report that cooked pulses are well tolerated because anti-nutrients are completely inactivated by heat [89
]. Our findings of normal gut physiology based on crypt size, mucus synthesis, and cell proliferation are consistent with the fact that pulse consumption at 2–3 times the intake observed in countries such as the United States and Canada is well tolerated. It is noteworthy that some laboratories have reported effects on the same intestinal morphometric parameters and have interpreted those effects as consistent with improved gut integrity. Our findings do not rebut those findings. Rather, our extensive experience in morphometric analyses recognized that differences in technical approaches can render different outcomes. In our judgement, the combined data of a number of publications [19
] confirm that pulses are well tolerated and underscore the value of functional rather than morphometric assessment of gut integrity as an objective approach to moving the field forward.
Pulse consumption has been reported to improve blood lipid profiles associated with cardiovascular disease in clinical studies and in preclinical experiments in the model of obesity used in the work reported herein [42
]. One possible contributor to pulse-mediated beneficial effects on circulating lipids involves changes in bile acid metabolism [89
]. Since it has been shown that a shift in the ratio of Bacterioidetes to Firmicutes alters intestinal bile salt hydrolase activity in the ileum, and that such changes alter the expression of the bile acid receptor FXR [48
], ileal transcript levels were assessed. Interestingly, all pulses induced expression of 1.5- to 1.8-fold relative to the obesogenic control diet. Whether this is enough to induce changes in bile acid metabolism in the liver associated with health benefits and/or results in changes in lipid metabolism in adipose tissue will require further detailed investigation. However, the fact that pulse consumption reduced lipid accumulation in subcutaneous and visceral adipose depots (Table 2
) is consistent with mediation by alterations in bile acid metabolism and receptor driven signaling with FXR being a candidate transcription factor.
This study has several limitations. They include (1) the inability to fully match chemical composition across dietary formulations when a whole food (pulse) approach is used; we argue that this approach is essential for understanding how the consumer may benefit from pulse crop consumption; (2) pulses were cooked and then immediately freeze dried; the potential impact of this approach on qualitative changes in carbohydrate constituents is unknown, but this approach is being used to make commercially available pulse powders that are being used as ingredients in the design of new food products; and (3) content of specific bacterial populations was determined by qPCR; while not as much information is obtained about overall microbial ecology as attained via high-throughput techniques, qPCR is a rapid, specific, and cost effect method [50