1. Introduction
A growing body of evidence over the past decade has demonstrated the importance of the gut microbiome in all aspects of physical and mental health. While it is still unclear what exactly constitutes a ‘healthy’ gut microbiome, certain bacterial groups have been strongly associated with better health outcomes.
Lactobacillus is one of the foremost genera considered to have probiotic properties [
1]. A probiotic is defined as a live microorganism which, when administered in adequate amounts, confers a health benefit on the host [
2]. The word ‘
psychobiotic’ is an expansion of this term and describes an organism that has been proven to be beneficial in relation to psychological functioning [
3]. There have been a wide variety of studies undertaken in recent years that have demonstrated the benefit of a
Lactobacillus probiotic, both mono- and multi-strain, for improving a range of health outcomes, including obesity [
4], diabetes [
5], liver disease [
6], cardiovascular disease [
7], gastrointestinal conditions [
8] and neuropsychiatric disorders, such as depression, anxiety and autism [
9].
A key present-day challenge involves identifying the most effective ways of maintaining a healthy gut microbiome and promoting the growth of probiotic bacteria. While commercial probiotic products are widely available, there are concerns in relation to regulation, quality control, efficacy and cost [
10]. Dietary intake is one of the main factors regulating gut microbiome composition and food-based interventions can be tailored to each individual to modify their bacterial profile [
11]. While unravelling the diet–microbiome relationship is a formidable task given the many confounding factors, attempts to do so have been made over the past decade. Gut microbiome profile has been shown to be distinctly different in those living in rural areas with a traditional diet in comparison to urban-based westernised populations [
12,
13,
14]. Even when one accounts for contributions of human genetic and geographical factors between populations, subsistence methods and diet significantly impact gut microbiota composition [
15]. It is hypothesised that a ‘microbiota insufficiency syndrome’ has resulted from modern lifestyle with its highly processed diets, overuse of antibiotics and increased sanitation and that the ‘industrialised’ microbiota may be a major contributing factor in the rise of many non-communicable chronic diseases in westernised societies [
16]. Even as one moves from looking at the early ancestral microbiota to more recent times, significant changes in lifestyle have continued until relatively recently. Ireland, as with many countries in the developed world, was a predominantly agrarian society up until the mid-late 20th century. In 1966, over 30% of the workforce were employed in agriculture, with this figure estimated at less than 5% in 2016 [
17]. Consumption of unpasteurised milk was a common part of the diet of those living on farms and epidemiological studies suggest that it may have played a protective role against the development of allergies and atopic diseases [
18].
Despite food safety concerns, the consumption of unpasteurised milk appears to be growing in popularity [
19,
20]. To our knowledge, there are no studies exploring the impact of unpasteurised milk intake on the gut microbiome. In this observational study, we investigated the effect of a dietary change involving the intake of unpasteurised milk on gut microbiota composition, metabolites and psychological status in 24 participants undertaking a residential, farm-based, 12-week cookery course. Our centre had previously published a study [
21] on the microbiota composition of unpasteurised milk taken from Irish cows, which would thus be representative of the expected microbiota composition of the raw milk that would be consumed by participants in our study. Given the reported high proportion of viable probiotic bacteria such as
Lactobacillus (and other LAB such as
Lactococcus and
Leuconostoc), along with the fact that Lactobacillus are considered intrinsically resistant to gastric acid [
22], we hypothesised that a dietary change involving raw milk consumption would alter the gut microbiome of participants with a potential differential increase in the relative abundance of these probiotic bacterial groups.
4. Discussion
In this observational study, we investigated the effect of a dietary change on the gut microbiome of participants who undertook a 12-week residential cookery course on an organic farm, where the majority of food consumed and used for cooking, was locally-sourced, seasonal and produced using organic methods. Of particular interest was the use of unpasteurised milk and dairy products obtained from a small herd of Jersey cows on the farm. Most participants had not been using any unpasteurised dairy prior to the course and all used these products to some degree throughout their stay. We found that the main change in terms of microbiome composition was a dramatic increase in the participants’ Lactobacilli between pre-course and post-course faecal samples. This increase was strongly associated with the participants’ intake of unpasteurised milk and dairy products. In addition, a positive change was noted in relation to microbiome metabolites with an increase in valerate and, to a lesser extent not quite reaching statistical significance, propionate.
While administration of probiotics in the form of conventional pharmaceutical agents such as tablets or capsules is a common method, the majority of probiotics commercially available are in the form of food-based delivery systems, which use probiotic bacteria in their production or add these bacteria during the manufacturing process, for example, cheese, yoghurt or fermented drinks [
35]. There are several problems associated with pharmaceutical and commercially-produced probiotic formulations. Firstly, the probiotic potential of bacteria is species and strain-specific but efficacy is often generalised across products in the current unregulated commercial probiotic market [
36]. Secondly, there are many aspects of the manufacturing process of such products, which can alter the delivery of viable functional probiotic bacteria [
37]. Because probiotic products are generally categorised as food supplements, they are subject to less stringent regulatory criteria and quality control processes with regard to microorganism specification, their numbers and functional properties [
10]. Thirdly, there is a cost consideration when it comes to commercial probiotic products, which may place daily probiotic supplements out of the reach of many.
An alternative to consuming commercially-produced probiotic supplements for the maintenance of a healthy gut microbiome is to alter one’s diet. It is increasingly accepted that the ‘Western-diet’, characterised by highly-processed, low-fibre, high-sugar, high-fat foodstuffs has negative implications for health [
38], which may be mediated by an unfavourable impact on the gut microbiome [
39]. In contrast, adherence to a Mediterranean-style diet (characterised by high-level consumption of olive oil, fruit, nuts, vegetables, and cereals with moderate intake of fish and poultry) has been strongly associated with better physical [
40] and mental [
41] health outcomes, which again may be related to a beneficial impact on the gut microbiome and metabolome [
42]. Gut microbiome composition can be rapidly and significantly altered by introducing dietary change [
43] with the impact of food choices on the microbiome being highly individualised [
11]. In this study, the key change in relation to dietary intake during the 12-week residential course was an increase in dairy products, which in this context were unpasteurised. This was a major change for our subjects, the vast majority of whom did not consume unpasteurised milk or dairy products prior to the course.
Cow’s milk is produced on a massive scale worldwide and has long played an important role in human nutrition [
44]. Cow’s milk harbours a rich microbiota and typically contains a significant population of lactic acid bacteria (LAB) that includes
Lactococcus (8.2 × 10
1–1.4 × 10
4 CFU/mL),
Streptococcus (1.41 × 10
1–1.5 × 10
4 CFU/mL),
Lactobacillus (1.0 × 10
2–3.2 × 10
4 CFU/mL),
Leuconostoc (9.8 × 10
1–2.5 × 10
3 CFU/mL) and
Enterococcus spp. (2.57 × 10
1–1.58 × 10
3 CFU/mL) [
45]. Other organisms present in substantial proportions are
Pseudomonas and
Acinetobacter, so-called psychrotrophs which can flourish during cold storage conditions and typically cause milk spoilage [
46]. Pasteurisation of milk gained widespread popularity in the early 1900s when cow’s milk was linked to the spread of disease epidemics such as tuberculosis, diphtheria, typhoid fever, scarlet fever, anthrax and cholera [
47]. A recent Irish study, using molecular, culture-independent techniques, compared the microbial content of raw and pasteurised cow’s milk [
21]. Authors reported that, although the bacterial diversity of the raw and pasteurised milk was similar, raw milk contained mostly viable cells whereas the cell population in pasteurised milk were predominantly nonviable. Thus, while pasteurised milk appeared to have a somewhat similar microbiome composition to that of the raw milk, any potential probiotic LAB would have been in a nonviable state. In this study,
Pseudomonas and
Acinetobacter, two major genera found in unpasteurised milk, were not detected by 16S rRNA analysis of the microbiomes of the participants, either pre or post treatment. This may be due to a selective filtering effect of the human immune system or physiological barriers such as gastric acid, which is known to act as such a filter [
48,
49].
The consumption of raw milk is growing in popularity, although there is some debate in relation to its purported benefits and concern about the potential dangers of contracting milk-borne illnesses if the raw milk is contaminated with human pathogens [
50]. There is a strong suggestion from epidemiological literature that the consumption of unpasteurised cow’s milk or yoghurt by children living on farms or rural areas has a protective effect against the development of asthma, allergies and atopy, a finding that seems to be independent of other farm-related exposures [
18]. In addition, raw milk is anecdotally reported to be beneficial for people with lactose intolerance [
51]. This is thought to be due to the fact that raw milk contains high counts of LAB that produce lactase enzymes, which would otherwise be destroyed during pasteurisation. However, there is little research evidence to support these anecdotal claims and, in fact, one recent pilot randomised controlled trial (RCT) involving 16 adults with lactose malabsorption, failed to find any benefit of raw milk over pasteurised milk for gastrointestinal symptoms [
52]. Despite this, in a survey of raw-milk consumers [
53], over one-third of responders claimed to experience gastrointestinal discomfort from drinking pasteurised milk but no discomfort after drinking raw milk, although the vast majority of these people did not have a diagnosis of lactose intolerance. Another proposed benefit of raw milk is that it contains higher quantities of vitamins. A meta-analysis [
54] reported that pasteurisation reduced the concentrations of Vitamin E, Vitamin B12, Vitamin B2, Vitamin C and folate. Of these vitamins, B2 is of most importance as bovine milk contributes significantly to the recommended daily intake whereas in the case of all the others, milk is not typically an important source. In relation to the human gut microbiome, we are unaware of any studies specifically examining the effect of raw milk consumption. However, a few studies have investigated the impact of pasteurised milk on the human microbiome. One cross-sectional study reported a differential oral microbiome based on high versus low (pasteurised) milk intake [
55]. Another investigated the impact of whole milk supplementation on the gut microbiota and cardiometabolic biomarkers between lactose malabsorbers (LM) and absorbers (LA) [
56]. Authors found that whole milk supplementation significantly altered the intestinal microbiota composition in LM, resulting in an increase in the phylum Actinobacteria along with increases in several genera;
Bifidobacterium,
Anaerostipes and
Blautia. These changes occurred only in LM and not LA, suggesting that it was the increased lactose substrate reaching the colon that preferentially enhanced the growth of some microorganisms. In addition to pasteurisation, milk can be altered by skimming, which is currently a widespread procedure. Prior to the course, 10/24 of our participants reported consuming skimmed or semi-skimmed milk, while post-course, 23/24 participants consumed whole milk, reflecting the unpasteurised milk intake. Skimmed milk contains less fat than whole milk and thus also less fat-soluble vitamins such as A and E. However, regular unfortified milk is not a major contributor to a person’s recommended daily allowance of these vitamins [
57] and despite the variable amounts in different milk types there does not appear to be a significant difference in their bioavailability [
58]. Other micronutrients such as calcium, sodium and choline do not differ between skimmed and whole milk [
59]. Therefore, we considered the skimmed versus whole milk type to be of limited consequence.
An obvious limitation of this study is the inherent potential for confounding given that, in addition to a change in diet, study participants experienced a change in environment. Disentangling the impact of diet and geographical environment on the gut microbiome, however, is a very difficult task. Several large scale studies have attempted to explore the differences in microbiome composition between industrialised Western urban dwellers and those living in traditional rural communities in South America and Africa, such as the Hadza hunter-gatherers of Tanzania [
60], rural Papua New Guineans [
61], children from the rural African village of Burkina Faso [
14] and communities from Malawi and Amazonian Amerindians [
12]. Although a rural setting will likely contribute to gut microbiome differences, these farming environments are intrinsically linked to variation in diet and it is difficult to separate the impact of the farm environment itself and the farm-related dietary patterns. If a move to a rural farming environment were to account for the changes in microbiome seen in our study, one could postulate that the changes would be consistent with the microbiome composition in rural dwellers from the above studies. This was not the case. While rural dwellers from PNG did have higher abundance of
Lactobacillus than their urban counterparts [
61], those from the other rural farming communities did not [
14,
60]. Obviously, the rural locations in the above studies were at the extreme end in relation to geographical location and traditional lifestyle and poorly comparable to the developed farm environment in which our participants were based. In a study more closely resembling our location, authors compared the microbiome of infants from farming and non-farming families in Wisconsin, United States, and again no differences in
Lactobacillus or other LAB abundance were seen [
62]. Furthermore, the changes in bacterial taxa in the microbiome of our subjects were consistent with those species found in unpasteurised milk, supporting our conclusion that this specific dietary change was driving the microbiome differences between pre- and post-course time points.
In this study we found that, during the 12-week course, the levels of the faecal SCFA valerate increased with a trend towards an increase in propionate. Straight-chain SCFAs (acetate, butyrate, propionate and valerate) are produced by the gut microbiota during the fermentation of partially and nondigestible polysaccharides whereas branched-chain SCFAs (isobutyrate and isovalerate) result from the metabolism of proteins [
63]. SCFAs are thought to play a major role in the maintenance of gut and immune homeostasis [
64] as well as in the gut-brain axis response to stress [
65]. SCFA production can be stimulated by increasing dietary fibre intake [
66] or protein consumption [
67]. However, in our study, participants intake of fibre or protein did not change, and thus, it is proposed that increased valerate and propionate levels may have been secondary to increased abundance of
Lactobacilli, which, along with other LAB, are known producers of SCFA [
68]. Propionate has anti-inflammatory properties and has been shown to be of potential benefit across a range of disorders, including hypertension and cardiovascular disease [
69], obesity [
70] and hypercholesterolemia [
71]. Valerate is a less well-known SCFA with limited research to date into its therapeutic potential. However, a recent study revealed that it also appears to have an immunomodulatory effect [
72]. Interestingly, supplementation with
Lactobacillus acidophilus increased the concentration of valerate in the caecum of chickens infected with
Clostridium perfringens while reducing the infection-associated gut dysbiosis [
73]. Valerate may also hold some translatable therapeutic value in the context of
Clostridium difficile infection (CDI). Valerate was shown to be significantly reduced in the faecal samples of patients with recurrent CDI and recovered following successful treatment with FMT [
74].
Changes in the functionality of the microbiome were assessed in the context of a recent study which facilitates analysis of the neuroactive potential of a microbiome sample [
34]. Authors achieved this using a gut–brain-module (GBM) framework, which targets microbial pathways known to be involved in microbiota–gut–brain communication and have made this GBM catalogue available for use by other researchers (
https://raeslab.org/software/gbms.html). When applying our predictive metagenomic data to this GBM catalogue, we found an increase in the functional richness of the microbiome profile, as determined by the number of GBMs present, following the 12-week course (
Figure 4). Such a consistent general increase in GBMs without a significant increase in microbial alpha diversity goes somewhat against the intuition that a more diverse microbial ecosystem will necessarily display a higher functional diversity. More strikingly, the functional alpha diversity did not change during the course. GBMs represent a specific subset of microbiome function and are calculated using the values of specific KEGG Orthologues. A shift in microbial functions that specifically potentially impact the host brain without a corresponding general shift in microbial function detectible on the alpha diversity level shines light on the possibility that many more such specific shifts can occur undetected using current bioinformatics tools. Because of this, we call for a move away from general diversity and towards informed interrogation of specific functional changes in the microbiome as a readout.
One GBM changed significantly after post-hoc correction; ‘GBM026; Nitric oxide synthesis II (nitrite reductase)’. Several studies have demonstrated the ability of various
Lactobacillus species to synthesise nitric oxide by nitrate reductase activity [
75,
76]. Nitric oxide is a complex and widespread signalling molecule that participates in virtually every organ system of the body. It is thought to play a role in the stress response and mood regulation [
77] and may represent one mechanism by which
Lactobacilli exert psychobiotic effects. The authors believe another GBM warrants discussing here, although its increase did not satisfy significance after post-hoc correction; ‘GBM004, Kynurenine synthesis’. This module was never detected in participants pre-course but was present in very high levels in 6 out of 24 participants post-course. This can be explained by the fact that the Kynurenine synthesis module requires two enzymatic steps. One of these was found in
Lactobacillus, but the other one was not specific to a single microbe in this data set, but rather spread over several microbes and was only found in the six participants positive for MBG004. This finding conforms well with literature regarding the emergent biosynthetic capacity of the microbiome [
78,
79].
Although we found no direct correlation between
Lactobacillus abundance and psychological measures, it is notable that stress and anxiety levels reduced significantly in those with higher baseline scores on the PSS and HADS-A. This is consistent with probiotic interventional trials in healthy populations, whereby an impact is often only seen in those with higher anxiety or depression scores at baseline [
80,
81]. Of course, there are many possible confounding factors when it comes to interpreting this reduction. Participants in this course had varying reasons for completing the course; for some, the purpose was to enhance or change their career options and, thus, possibly associated with stress; for others it was simply for leisure and viewed more as a holiday incorporating cookery classes. The change in environment and daily activity, the purpose of participation in the course and interaction with new people may all have contributed to psychological status. However, given the increasing evidence that the gut microbiome is an important node in gut–brain communication and that certain psychobiotics have anxiolytic effects, it is plausible to consider the possibility that the improvement in stress and anxiety may have been partially related to the increase in
Lactobacillus.
Lactobacillus rhamnosus (JB-1) has been shown to reduce anxiety behaviours in mice as well as altering central levels of gamma-aminobutyric acid [
82], a key neurotransmitter in anxiety regulation. Several species of
Lactobacillus have demonstrated the ability to reduce anxiety and stress levels in healthy subjects [
83,
84,
85] as well as in patients with chronic fatigue syndrome [
86] or laryngeal cancer [
87].
There are several limitations to our study. Firstly, this was an observational study. While of course an RCT would be preferable, there are many challenges inherent in designing RCTs involving dietary interventions. It can be difficult to define appropriate control groups and effective blinding of participants and investigators is often extremely difficult [
88]. In particular, it can be challenging to accomplish a high level of adherence with whole food, or dietary pattern, interventions. A major strength of our study in this regard was that our participants were based on-site for the entire duration of the study making it possible to ensure a consistency across individual diets, which would be difficult to achieve outside a residential setting. The potential confounding effect of the farm environment as an independent modulator of microbiome composition is addressed earlier in the discussion. Secondly, our sample size was quite small. However, previously published studies investigating the diet–microbiome relationship have involved participant numbers of ten or less [
43,
89] and have generally been of much shorter duration [
11]. Another factor that may limit the generalizability of our study was that participants undertaking this course were interested in food and cooking. Thus, they were likely to have good nutritional knowledge and possibly healthier than average diets at baseline. A specific limitation in this regard was an absence of any information on the use of non-nutritive sweeteners (NNS). These are being increasingly used due to the concern about the negative health impact of high-sugar diets and have been shown to significantly, and generally negatively, impact the gut microbiome [
90]. Finally, given the limitations of 16S rRNA gene sequencing we were unable to characterise organisms beyond the genus level. More accurate taxonomic classification would have been useful had shotgun metagenomic sequencing been performed. Despite these limitations, this is, to our knowledge, the first study to report on the potential impact of unpasteurised milk and dairy products on the human gut microbiome.