The effect of bovine dairy products and their components on the incidence and natural history of infection: a systematic literature review

Abstract

Background

Dairy products and their components may impact immune function, although the current evidence base has some research gaps. As part of a larger systematic literature review of dairy products/components (including probiotics, dairy proteins, and dairy fats) and immune function, we identified the available epidemiologic research on the impact of dairy products/components on incidence and natural history of infectious diseases.

Methods

PubMed and Embase databases were systematically searched through May 2022 to identify eligible studies using pre-defined Population, Intervention, Comparator, Outcomes, and Study design criteria. Herein, we focused on describing the impacts of dairy product/component on infectious disease outcomes, including the effect on leukocyte and cytokine response in humans. Risk of bias assessment was performed using the Academy of Nutrition and Dietetics Quality Criteria Checklist. The Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) guidelines were followed.

Results

Among 9,832 studies identified from the larger literature search, 133 relevant publications from 128 studies reported on dairy product/component and infectious disease outcomes. Few studies are available on the impact of non-fermented milk and traditional yogurt on infectious disease. Evidence was identified to suggest milk and yogurt drinks fermented with Lactobacillus strains reduce the risk and burden of common infectious diseases (CIDs), although the findings are mixed and difficult to reconcile due to heterogenous study populations, bacterial strains, and study methods. Few studies are available on the impact of dairy products/components on the natural history of infection, with the available findings indicating probiotics may both improve gastrointestinal symptoms among HIV-infected persons and help eradicate and alleviate the symptoms of Heliobacter (H.) pylori. The available evidence also suggests lactoferrin may reduce the virological burden of COVID-19 and hepatitis C virus. No consistent changes in leukocytes or cytokine production were observed for any type of dairy product or their components, but probiotics appeared to enhance natural killer cell levels/activity and the phagocytic process.

Conclusions

Dairy products, particularly those with added probiotics, may represent an easily accessible nutritional intervention to prevent and improve the course of infectious diseases. This review highlights the need for additional research in this potentially impactful area.

Prospero registration

CRD42022333780.

Introduction

Infectious diseases are responsible for substantial morbidity, mortality and economic impacts worldwide (e.g., [1, 2]). A new era of infectious disease transmission dynamics has been described, defined by outbreaks of emerging, re-emerging, and endemic pathogens that are being transmitted rapidly due to global connectivity [3]. Furthermore, an increased incidence of viruses such as influenza and respiratory syncytial virus has been described in the wake of the COVID-19 pandemic, due to the “immunity debt” created by the non-pharmaceutical interventions to control the spread of COVID-19 [4]. As the demographics in the United States shift toward an older population, the impact of these infectious diseases may also be enhanced due to age-related decrements in immune function [5]. Given the substantial burden of infectious diseases and shifting transmission dynamics, it is important to find easy and accessible measures for infection prevention. These measures could be particularly useful to combat “immunity debt” in the wake of another global pandemic and in communal settings of vulnerable populations, such as long-term care facilities for elderly persons and daycare and school settings.

An evolving field of research considers whether the consumption of specific food components has immunostimulatory effects [6, 7]. As an integral part of a healthy dietary pattern [8], dairy products and their components could represent an important and easily modifiable factor to prevent infection in both general and immunocompromised populations. The potential immune-modulating effect of dairy product components has been considered largely in animal and cell models and to a lesser extent in humans [6, 9]. The cumulative findings suggest a beneficial role of dairy products for immune-related outcomes overall; however, studies are heterogeneous in design and quality [10].

While reviews and meta-analyses have examined some dairy products/components in relation to immune-related outcomes, the full scope of this topic is undefined and the findings are described in the literature as insufficient and heterogenous [10,11,12,13,14,15,16]. Herein, we sought to identify the full scope of existing literature on this topic and to examine whether dairy products and/or their bioactive components are associated with any aspect of immune function in humans. During our systematic assessment of outcomes that have been examined in the published literature to date, the incidence and natural history of infectious diseases was identified as an outcome with available evidence sufficient to review. Given the increased focus on infectious disease transmission with the COVID-19 pandemic and the substantial morbidity associated with common infectious diseases (CIDs), this SLR examined the available evidence on the potential for dairy products/components to impact the risk of acquiring an infectious disease or affect the natural history of an infectious disease. To complement this assessment, we also evaluated the evidence for the effects of dairy products/components on leukocyte and cytokine response, as these cellular effects are an important component of the pathogenesis of infectious diseases.

The goal of this systematic review is to provide a central location to weigh the epidemiologic evidence on the potential for dairy products/components to impact infectious disease risk/natural history.

Methods

The study protocol was registered on Prospective Register of Systematic Reviews (PROSPERO, CRD42022333780) (www.crd.york.ac.uk/PROSPERO) prior to the start of this SLR. In the design, execution, and reporting of the current SLR, we followed all Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) guidelines; the PRISMA checklist is provided as Additional file 1 [17]. The overall SLR was conducted to 1) identify the available evidence on dairy products/components and immune-related outcomes and 2) assess the breadth of the available evidence on the identified outcomes. As the studies resulting from the overall literature review were incredibly heterogeneous in scope, we narrowed the focus of the current manuscript to outcomes related to infectious diseases. Herein, we summarize the evidence related to nutritional interventions with dairy products/components and 1) the incidence of infectious diseases, 2) the natural history of infectious diseases and 3) the impact on leukocyte and cytokine response. Other outcomes related to immune function will be presented in future publications.

Eligibility criteria

Pre-defined study population, intervention, comparator, outcomes, and study design (PICOS) criteria were used to assess study eligibility.

Study population

All studies of humans without dairy sensitivities were included; there were no restrictions on geographical location, sex, age, or health status.

Intervention

Exposures/interventions of interest included the following:

• 1. Cow’s milk products recommended by the United States Department of Agriculture (USDA) in their most recent 2020 guidance, i.e., milk (including milk powders), yogurt, and cheese [18], including those with added traditional and experimental probiotic strains;

• 2. Cow’s milk proteins (i.e., whey and casein proteins) and peptides;

• 3. The fat components of milk (i.e., milk phospholipids and the MFGM); and

• 4. Estimated dairy intake as measured by dietary recall instruments.

Prenatal and maternal exposures to dairy products/components in relation to pediatric outcomes were considered. This review did not evaluate the impact of probiotic strains administered outside the context of dairy products (e.g., as isolated supplements or powders). Studies of bovine colostrum, non-bovine milks, hyperimmunized milk, and raw/unpasteurized milk were excluded. Studies where dairy products/components were administered through a feeding tube were included, but studies using a jejunal tube or other system bypassing the stomach were excluded. Likewise, studies involving the administration of a dairy product/component through a non-oral route (i.e., intranasal, topical, ophthalmic) were excluded. In this review, yogurts using the traditional starter cultures Lactobacillus (L.) bulgaricus and Streptococcus (S.) thermophilus are referred to as traditional yogurt, while probiotic yogurts are those with additional bacterial strains added.

Comparator

Studies with comparison group(s) of low or no dairy product/component consumption or studies comparing pre- and post-intervention outcomes were included. Studies that provided relevant data but did not calculate an effect estimate or conduct any statistical comparisons were excluded.

Outcomes

All studies with outcomes related to immune function were included, excluding outcomes related to milk allergies, milk sensitivities, or antibodies to milk proteins. This review summarizes the following outcomes reported in the included studies: 1) the incidence of infectious diseases, 2) the natural history of infectious diseases, 3) leukocyte response and 4) cytokine response. Data on some cytokines, i.e., the small proteins triggered by infection, were excluded from our review since previous SLRs have summarized the impact of dairy products/components on these biomarkers of inflammation, including adiponectin, c-reactive protein (CRP), homocysteine, interleukin (IL)-1, IL-6, IL-8, IL-15, IL-18, intercellular adhesion molecule (ICAM), monocyte chemoattract protein (MCP)-1/CCL2, serum amyloid A (SAA), tumor necrosis factor (TNF), and vascular cellular adhesion molecule (VCAM) [12, 13]. Given the pleiotropic nature of cytokines in general and interleukins specifically (i.e., they could have both inflammatory and non-inflammatory effects), all other cytokines were included. Viral and bacterial disease outcomes were included while fungal infections were excluded.

Study design

All observational studies and clinical trials were included. Studies not published in English, conference abstracts, meta-analyses, and case reports or case-series with ≤20 cases were excluded. If more than one article from the same study population was published, data from the publication with the longest follow-up or most relevant population and/or outcomes were evaluated.

Study identification, screening, and abstraction

The PubMed and Embase databases were used to identify relevant studies published through May 19, 2022, as described in Supplemental Table 1 (Additional file 3). These citations were combined using DistillerSR software [19], which was used to manage the citations during all levels of review and data extraction. One researcher reviewed the titles and abstracts using the PICOS guidance. If an abstract was considered of potential interest, the full-text article was considered by two independent reviewers, with conflicts between the two reviewers resolved by discussion or a senior reviewer. Once a study was identified as relevant at the full-text level, select data were abstracted into DistillerSR. Abstraction elements included study characteristics (e.g., study design, time period of recruitment, and location), population characteristics (e.g., age and health status), information on the intervention(s) and controls (e.g., exposure details, dose and duration), and results related to the relevant outcomes. The abstracted data were assessed by a second reviewer for quality control; conflicts were resolved by a senior reviewer.

Table 1 Characteristics of included studies, organized by outcome, dairy exposure, and study quality (N=45)

Tables 2, 3, 4, 5 and 6 and Supplemental Tables 3 and 4 summarize the effect measures and/or relevant statistical comparisons in the identified studies. If a comparison was not statistically significant, we indicated what specific outcome was measured, which groups were compared and that the difference was not statistically significant (NSS). If a comparison was statistically significant, we indicated what specific outcome was measured, which groups were compared, the direction of the change, and that the difference was statistically significant (SS) with a p-value. Hazard ratios (HRs), relative risks (RRs) and odds ratios (ORs) were reported with the 95 percent confidence interval (95% CI), where appropriate. In Supplemental Tables 3 and 4 (Additional file 3), hierarchies were created to be parsimonious in reporting the most meaningful results: comparisons between the experimental and control groups were chosen over comparisons within these groups; comparisons with the highest vs. the lowest dose were chosen (i.e., tertile 3 vs. tertile 1); and, the most adjusted comparison was chosen if multiple models were available.

Table 2 Clinical trials of dairy products/probiotics on the incidence, duration, and severity of acute infections (N=23)

Table 3 Studies of dairy proteins on incidence, duration, and severity of acute infections (N=7)

Table 4 Literature on the effects of dietary patterns involving dairy on acute infections (N=4)

Table 5 Studies of dairy products/probiotics on the natural history of chronic infections (N=5)

Table 6 Studies of dairy proteins on the natural history of infections (N=6)

Risk of bias assessments

Risk of bias (RoB) assessment was evaluated using the Academy of Nutrition and Dietetics Quality Criteria Checklist [66], which was specifically designed for nutritional studies. This checklist collects yes, no, not available (NA), or unclear responses to 10 validity questions to assess various domains where bias can arise in a study (e.g., inclusion and exclusion criteria, withdrawal, data collection, data analysis, and conflicts of interest). RoB assessment was conducted by one reviewer; the results were reviewed independently by a second reviewer for complete equality control. A senior reviewer resolved any conflicts and finalized the RoB results. Study quality was determined as positive quality, neutral quality, or negative quality, depending upon the scoring results from the domains (Supplemental Table 2, Additional file 3).

Data synthesis

Qualitative synthesis was conducted, as meta-analysis could not be performed due to the heterogeneous nature of the dairy exposures and reported outcomes. Results are summarized below by outcome, including the incidence, duration, and severity of infections (Tables 2, 3 and 4) and the natural history of infectious diseases (Tables 5 and 6). Within each of these outcomes, studies were summarized by the exposure/intervention, including whole dairy products, a particular probiotic added to a dairy product, dairy proteins, and dietary intake of dairy. Milk products were separated by traditional and fermented milk, and yogurt was separated by traditional and probiotic yogurts. This level of exposure and comparison group detail was considered to evaluate what component of the dairy product could be beneficial – the cumulative matrix of the whole dairy product, traditional yogurt ferments, particular probiotic strains delivered in dairy, and/or proteins. Studies were also summarized by bacterial strain, as beneficial effects may be strain specific. Additional effect modifiers that were considered in the qualitative synthesis included the age and health status of the population.

Results

Article identification

Figure 1 displays the PRISMA flow diagram detailing study inclusion/exclusion at each stage of review. The database searches yielded 12,973 hits. After de-duplication across databases, 9,832 abstracts were screened, 389 of which were identified as potentially relevant and flagged for full-text review. After reviewing the full-text articles, 207 articles were excluded for the following reasons: 74 did not have any exposures of interest, 50 had no outcomes of interest, 34 had no primary data (i.e., meta-analyses, opinion pieces, or reviews), 16 were relevant reviews, 14 had no effect measures calculated or statistical testing, 10 had no relevant comparison group, 5 were an in vitro or in vivo study, 2 were case reports or case-series with less than 20 patients, 1 was a conference abstract, and 1 publication was excluded because its primary data was included in another publication. Sixteen relevant reviews were identified, and 5 additional studies were identified by examining their reference lists. Thus, 187 publications meeting the pre-defined PICOS criteria were included in the overall SLR.

Fig. 1

PRISMA flow diagram. *Four of these review articles were not searched for additional references because the topic of the article was bovine colostrum and/or hyperimmune milk. Source of flow diagram template: Page et al. 2021 [17]

Among the 187 publications, 133 relevant publications were identified, reporting on 128 unique studies; many of these studies reported on more than one outcome relevant to this SLR. Our review identified 34 studies specifically examining the incidence, duration, and/or severity of acute infectious diseases (Tables 2, 3, and 4) and 11 studies on the natural history of infectious diseases (Tables 5 and 6). Our review also identified 76 studies of leukocyte levels and measures of leukocyte activity, proliferation, cytotoxicity and phagocytosis (Supplemental Table 3, Additional file 3) and 47 studies of cytokine production (Supplemental Table 4, Additional file 3). Five instances were identified where two publications reported results from the same study [28, 55, 56, 67,68,69,70,71,72,73]; however, both publications were included because they provided unique data. Future publications will discuss other immune-related outcomes.

Characteristics of included studies on the incidence and natural history of infectious diseases (N=45)

Table 1 presents the characteristics of the 45 included studies on the incidence and natural history of infectious diseases. Of the 45 studies, 31 (69%) were given a positive quality rating, 12 (27%) a neutral rating, and 2 (4%) a negative rating. Thirty-eight (84%) were clinical trials, 5 were cohort studies, 1 was a case-control study, and 1 was cross-sectional. These studies were conducted in diverse locations, with 44% in Asia, 33% in Europe, 7% in North America (all US), 7% in Africa, 4% in South America, and 4% in Australia. The time period of study enrollment/follow-up ranged from 2000-2021. The most common exposure being investigated was a probiotic added to a milk or yogurt product (N=16), followed by whey proteins (N=13). The infectious disease under investigation varied significantly and included acute infections such as respiratory infections, influenza, and COVID-19 and chronic infections such as HIV and hepatitis C.

Incidence of acute infection (Tables 2, 3 and 4)

Whole dairy products (N=23, Table 2)

Twenty-three controlled, clinical trials were identified that administered whole dairy products in the experimental group and followed the study population prospectively for acute infectious disease incidence (Table 2). The comparison group varied widely between studies, with studies designed to examine the cumulative exposures associated with milk [20], fermented milk [21,22,23,24,25], traditional yogurt [26,27,28], and probiotic yogurt [28, 29] ingestion and other studies designed to examine the impact of adding a particular probiotic to the fermentation process of milk or yogurt drinks [30,31,32,33,34,35,36,37,38,39,40,41,42]. The infectious disease under consideration also varied and included broad categories of infections with a self-limited duration, including “winter infections” [21] and CIDs [24, 25, 31,32,33]. Some studies specifically evaluated respiratory tract infections (RTI) [34, 36, 38] and symptoms [41], upper respiratory infections (URTI) [20, 29] and symptoms [35], the common cold and influenza [26,27,28, 30, 39, 40], and norovirus gastroenteritis [22]. The method for evaluating disease incidence also varied from clinical verification of incident cases to disease definitions based solely on questionnaire data. The majority (N=15 or 65%) of these clinical trials were randomized and double-blinded [23,24,25, 30,31,32,33,34,35,36, 38,39,40,41,42]. The most common study locations were Japan and Italy (N=7 and N=6, respectively), with two studies conducted in North America (Table 1). Nine of these studies were conducted in healthy, adult populations [20, 21, 28,29,30, 37, 39,40,41], five were conducted among children attending daycare centers/schools [24, 25, 33, 38, 42], and five were conducted among elderly persons, both free-living [27, 32] and institutionalized [22, 23, 36]. Other potentially immune challenged populations were considered, including two studies of healthcare/shift workers [26, 31] and two studies of persons undergoing intense physical training [34, 35]. The vast majority (N=21, or 91%) of the studies were categorized as positive (N=15) or neutral (N=6) quality on RoB assessment (Table 1).

Only one study was identified that evaluated the impact of non-fermented milk (Tables 1 and 2). The average number of incident URTI cases was significantly lower (p<0.01) among Japanese adults consuming milk during weekly physical exercise for one year (compared to those consuming a sports drink), and a significant inverse correlation (p=0.03) between total dairy consumption and URTI severity was reported [20]. The study was limited, however, by a small sample size (N=13), a lack of blinding, and URTI diagnosis based on self-report [20].

Five clinical trials evaluated the cumulative impact of fermented milk, which included traditional ferments and ferments from various additional Lactobacillus strains (Tables 1 and 2). Mixed findings were reported. Two relatively large trials of Italian school children (N=146 and 268, respectively) reported a significantly lower proportion of clinically verified CID among children given 150 mL of milk fermented with L. paracasei CBA l74 daily for three months compared to children given a drink with a similar energy content, with significant incidence rate ratios (IRR) of 0.64 (95% CI=0.42-0.98) [24] and 0.36 (95% CI=0.29-0.44) [25]. A large trial (N=360) of healthy Italian adults given Actimel® (fermented milk with traditional ferments and L. casei DN-114 001) twice daily for three months found no difference in the cumulative incidence of clinically verified winter infections compared to no intervention [21], although the study was categorized as negative on RoB assessment due to the lack of details provided on the exposure, intervention and withdrawal. A study of elderly hospitalized Japanese patients found no reduction in the cumulative incidence of norovirus gastroenteritis associated with the daily ingestion of fermented milk with L. casei Shirota, compared to no treatment, during their hospital stay [22]. Despite mixed findings on disease incidence, the studies of fermented milk did, however, report consistent decreases in the duration of infections (winter infections, p=0.02 [21]; infections requiring antibiotics, p=0.05 [23]), the duration of symptoms ([fever associated with norovirus gastroenteritis, p=0.03 [22]), and measures of disease severity (CID, p=0.02 [24]; CID, p<0.001 [25]), although the data were collected largely in an open-label setting [21,22,23].

Mixed results were also reported on the benefits of traditional and probiotic yogurt (Tables 1 and 2). Two open-label trials in Japan examined the impact of daily ingestion of a traditional yogurt (Meiji Probio Yogurt R1® with L. delbrueckii ssp. bulgaricus) on the incidence of the common cold and influenza. One trial randomized female healthcare workers to the yogurt drink (N=479) or no supplementation (N=482) for 16 weeks [26]. There was no statistically significant difference in the cumulative self-report of a physician diagnosis of influenza or common cold between the yogurt and control groups, nor were there any statistically significant differences in influenza or common cold cumulative incidence in Kaplan-Meir analysis. The second trial randomized elderly Japanese persons to Meiji Probio Yogurt R1® or milk in two separate study locations for 8 (Fungata) and 12 weeks (Arita); a reduced odds of cold or influenza was observed when the two sites were meta-analyzed (OR=0.39, p=0.02) [27]. In a crossover trial of 30 US adults, no statistically significant difference was reported between the number of influenza and cold cases the month prior to the study compared to during treatment (treatment included traditional yogurt smoothie, a yogurt smoothie with added Bifidobacterium (B.) animalis subsp. lactis [BB-12] pre-fermentation, and a yogurt smoothie with BB-12 added post-fermentation); subjects consuming traditional yogurt smoothie and BB-12 added pre-fermentation experienced significantly fewer number of days with cold/flu symptoms (p<0.05 and p<0.01, respectively) [28]. Finally, in a study comparing probiotic yogurt supplemented with L. paracasei N1115 to no yogurt for 12 weeks among Chinese adults, a reduced risk of URTI events based on self-reported symptoms was found (RR=0.55, 95% CI=0.31-0.97) [29], similar in magnitude to the risk reductions for CID observed with L. paracasei fermented milk described above [24, 25].

The remaining thirteen clinical trials were designed to evaluate whether the addition of probiotic(s) to milk or yogurt drinks influenced the risk of infectious diseases [30,31,32,33,34,35,36,37,38,39,40,41,42]. One study investigated a probiotic from the genus Bifidobacterium [30], one study investigated a probiotic from the genus the genus Lactococcus [39], and the remainder of the studies investigated a probiotic from the genus Lactobacillus, with seven investigating L. casei [31, 32, 34,35,36,37, 74], one investigating L. johnsonii [23], one investigating L. rhamnosus GG [38], and three investigating a combination of Lactobacillus strains [40,41,42] (Tables 1 and 2). These studies are summarized below by bacterial genus and strain.

In the one identified trial of the probiotic species Bifidobacterium, 136 Chinese adults were randomized to 250 g of a yogurt supplemented with B. animalis subsp. lactis B1-04 (Qingrun®) or a control yogurt daily for three months (Tables 1 and 2). The supplemented yogurt was associated with a statistically significant approximately 60-70% reduction in the incidence of common cold and influenza-like illness (OR=0.38 [95% CI=0.17-0.81] and 0.32 [95% CI=0.11-0.97], respectively). The supplemented yogurt was also associated with a statistically significant reduction in URTI duration (p<0.0001) and severity (p<0.0001) [30], similar to the findings from the study of BB-12 supplemented yogurt smoothies [28].

Three large, double-blind clinical trials randomized study participants (N=1,000 German shift works, N=972 elderly Germans, and N=638 US children) to 200 mL of a dairy drink (Europe: Actimel®; US: DanActive®) fermented with L. casei DN-114 or a non-fermented dairy drink and followed them for the incidence of CID for three months, with mixed findings (Tables 1 and 2). The study of shift workers reported no difference in the cumulative number of all CIDs by Poisson regression, but a statistically significant reduced odds of CID (OR=0.70, 95% CI=0.54-0.90); a significant reduction in the cumulative duration of fever was also found (p=0.02), but no difference was found for other measures of disease severity/duration [31]. Similarly, the study of US children reported a reduced incidence of CID associated with DanActive® (IRR=0.81, 95% CI=0.65-0.99), but no impact on symptom duration or severity [33]. In contrast, the study of elderly Germans found no difference in the cumulative number of CIDs or the mean CID rate, but statistically significant improvements in measures of disease duration [32]. The authors cited the low number of observed events as a possible explanation for the lack of an association between the fermented drink and CID incidence in this study. An additional study of Actimel® was conducted in 47 male Italian cadets; 300 mL was given daily for one month during commando training and no difference was reported in the cumulative number of persons with RTI, the incidence of RTI or the duration of symptoms [34].

Three clinical trials of milk fermented with L. casei Shirota (Yakult®), compared to a non-fermented milk, reported trends toward an improvement in respiratory tract illnesses and symptoms (Tables 1 and 2). The consumption of 80 g of Yakult® 30 days prior to running a marathon was associated with a reduced, but not statistically significant (p=0.08), number of persons reporting upper respiratory symptoms after a marathon in Brazil [35]. A large (N=773) clinical trial of clinically verified RTI among elderly persons in Belgian nursing homes found no difference in the number of participants with at least one day of RTI symptoms, but a trend toward a reduced odds of developing a severe RTI in logistic regression modeling (OR=0.592, 95% CI=0.335-1.049) [36]. No difference in the duration of symptoms was found in either of these Yakult® trials [35, 36]. Finally, among a healthy population of adults in Japan, the daily consumption of Yakult® for 12 weeks was associated with a statistically significant reduction in the cumulative proportion of patients with an incident URTI (p=0.002) and cold (p=0.005) event, but not an influenza event; statistically significant reductions in the duration of each URTI episode (p=0.002) and the cumulative days with URTI symptoms (p=0.001) was also observed [37]. The authors suggested the age of the study participants may explain the conflicting results from the Yakult® studies, i.e., the older patients may be less responsive to the immune modulating effects of the yogurt drink [37].

In a study of Japanese children attending daycare who were randomized to L. rhamnosus GG fermented milk (N=282) or non-fermented milk (N=289) three times daily for seven months, a reduced odds of RTI was observed with treatment, but it failed to meet statistical significance (age-adjusted OR=0.75, 95% CI=0.53-1.09) [38]. Another Japanese study randomized healthy adults to a yogurt drink with Lactococcus lactis ssp. lactis JCM5808 daily for 12 weeks or a non-fermented yogurt drink and found no difference in the cumulative incidence of influenza or cold cases, but a statistically significant decrease in the duration of some symptoms (cough and feverishness, p<0.001 for each and severe sore throat, p=0.01) [39].

Three additional trials compared dairy products with combinations of added Lactobacillus species to a non-fermented control group with mixed findings (Tables 1 and 2). Daily consumption of a yogurt drink fermented with L. paracasei, L. casei 431 and L. fermentium PCC for 12 weeks was associated with a statistically significant reduction in the proportion of Chinese adults with URTI (p=0.002) and influenza-like illness with a fever (p=0.03), as well as a reduction in URTI duration (p<0.001) and various measures of severity [40]. A study of 200 mL daily milk ingestion with L. rhamnosus IMC 501 and L. paracasei IMC 502 for four weeks found no association with the self-report of respiratory symptoms, but was limited by a small size (N=10) and categorized as “negative” in RoB assessment (Table 1) [41]. In a study of low socioeconomic status (SES) Brazilian children, 140 children were randomized to milk fermented with L. casei CRL431 and L. acidophilus or non-fermented milk and no difference was found between the number of patients with URTI, gastroenteritis, varicella, or pneumonia [42].

Dairy proteins (N=7, Table 3)

Seven studies investigated the potential impact of dairy protein supplements on the incidence of acute infectious diseases (Tables 1 and 3). One study evaluated hydrolyzed whey protein [43] and another study evaluated a combination of the immunoglobulin rich fraction from whey protein and bovine lactoferrin [44]; no studies of casein protein were identified. The remaining five studies evaluated bovine lactoferrin [45,46,47,48,49]. The studies investigated a range of outcomes, including respiratory tract infections [44,45,46], bacteremia [43], and sepsis [47,48,49]. The studies were placebo-controlled, randomized, double-blind clinical trials, except for one retrospective cohort [43]. Three of the studies were conducted in adult populations [43,44,45] and four were conducted in infants [46,47,48, 75]. All of the studies were categorized as positive (N=6) or neutral (N=1) quality on RoB assessment (Table 1).

Three studies of whey protein supplements suggested that this milk protein may reduce the incidence and burden of common infectious diseases, although the evidence base is small and the trials were diverse in the age of the study population, dosing schedule and outcome. A trial of 105 Australian adults with recurrent colds reported that taking a combination of the immunoglobulin rich fraction from whey protein and lactoferrin for three months significantly reduced self-reported cold events (p<0.001) and symptoms over that time (p<0.05), compared to placebo, but had no statistically significant impact on cold duration [44]. A trial of 209 Japanese adults reported no significant differences in the prevalence or number of episodes of infectious diseases (the majority of which were summer colds) between participants given placebo, 200 mg or 600 mg of bovine lactoferrin daily for 12 weeks, but reported significant reductions in the duration of all infectious diseases (p=0.05 and 0.01 for 200 mg and 600 mg dosing, respectively), with a significant dose-response trend (p=0.01). The duration of common colds was also significantly shorter (p=0.04) among participants given 600 mg lactoferrin, but a similar pattern was not observed for the other infectious diseases, including gastroenteritis, cold sores and styes [45]. Finally, 52 infants were given either regular cow milk based formula or formula with added bovine lactoferrin for one year in a US trial; while the lactoferrin supplemented formula was associated with a significant (p<0.05) reduction in the average number of lower respiratory tract infections (LRTI) per infant-year, similar associations were not observed for URTI or acute otitis media (AOM) and no differences in disease duration were found [46].

Although the evidence base is small (N=4), studies suggest whey protein supplements reduce the risk of bacteremia/sepsis in adults and infants. A retrospective cohort of 76 Japanese adults receiving a liver transplant found that patients administered hydrolyzed whey peptides as part of their enteral nutrition post-transplant had a statistically significant reduction (p=0.002) in the occurrence of bacteriemia, compared to patients that received standard enteral nutrition [43]. Three studies enrolled low birth weight/pre-term hospitalized infants, provided 100-250 mg of bovine lactoferrin daily, and followed them for sepsis. The largest study of infants randomized Italian participants to 100 mg bovine lactoferrin (N=153), 100 mg bovine lactoferrin with L. rhamnosus GG (N=151), or placebo (N=168) for the first 30 days of life; a multivariate logistic regression analysis for late-onset sepsis reported ORs of 0.32 (95% CI=0.14-0.77) and 0.21 (95% CI=0.08-0.55) for bovine lactoferrin and bovine lactoferrin/L. rhamnosus GG treatment, respectively [49]. These late-onset sepsis events included bacterial and fungal infections, and a statistically significant reduction in the risk of sepsis was reported for bacterial episodes alone (no p-values were reported). An Indian trial (N=130) administered either bovine lactoferrin (with weight-based dosing) or placebo for the first 30 days of life and found a significant reduction in all sepsis events (p=0.001) and sepsis-attributable mortality (0.03) [47]. Similarly, in a trial of low-birthweight or pre-term infants in Turkey, a significant reduction (p=0.01) in the rate of sepsis was found in the treatment group (N=25, 200 mg bovine lactoferrin daily), compared to placebo (N=22) [48].

Dietary patterns involving dairy (N=4, Table 4)

Four studies reported associations between infectious diseases and dairy exposures measured by responses on food frequency questionnaires (Tables 1 and 4). Two large prospective cohort studies (one conducted in Iran and the other in France) measured the association between estimates of dairy, milk, yogurt, and cheese intake in the years prior to the pandemic and the seroprevalence of COVID-19 with multivariate logistic regression models. Weak, statistically significant positive associations were observed for estimates of high-fat dairy (OR=1.40, 95% CI=1.09-1.92), high-fat milk (OR=1.54, 95% CI=1.20-1.97) and yogurt (OR=1.40, 95% CI=1.04-1.89) intake in the Iranian cohort [50], and for total dairy (OR=1.19, 95% CI=1.06-1.33), milk (OR=1.15, 95% CI=1.03-1.27) and yogurt (OR=1.12, 95% CI=1.00-1.25) intake in the French cohort [51]. A statistically significant 50% reduction in the odds of COVID-19 seropositivity (OR=0.51, 95% CI=0.37-0.69) was observed with low-fat dairy products in the Iranian cohort [50]. In addition, a case-control study of Australian children reported cheese and full-cream milk consumption had significant protective effects (p=0.003 and 0.04, respectively) on Campylobacter jejuni infection [52] and a cohort study of Bulgarian adults reported frequent yogurt consumption had a significant protective effect (p=0.05) on Heliobacter (H.) pylori cytotoxin-associated gene A (CagA) seropositivity [53].

Natural history of infectious diseases (N=12, Tables 5 and 6)

Whole dairy products (N=5, Table 5)

Five studies were identified that evaluated the effect of whole dairy products/probiotics on a wide variety of outcomes related to the natural history of chronic infectious diseases (Tables 1 and 5). Three studies investigated the effect of whole dairy products, including fermented milk in children [54], probiotic yogurt in adults [55, 56] and a probiotic delivered in yogurt [57], on the natural history of HIV, with mixed findings. In a clinical trial of 60 Vietnamese children administered 65 mL milk fermented with L. casei Shirota daily, plasma viral load was found to decrease after 8 weeks compared to baseline (p=0.004) [54]. In a retrospective cohort of HIV-infected adults in Tanzania visiting a network of community-based nutritional intervention sites, daily ingestion of yogurt containing L. rhamnosus GR-1 was found to improve symptoms (diarrhea, p=0.05 and fever, p=0.01) and quality of life (ability to work, p=0.01 and impact of GI symptoms on daily life activities), compared to a group of HIV infected adults visiting the sites for other nutritional interventions [55, 56]. Contrary to these findings, however, no difference in symptoms, physical energy levels and the ability to perform daily activities was reported between HIV-infected adults given 125 mL yogurt with L. rhamnosus GR-1 (N=55) daily for 4 weeks, compared to HIV-infected adults given the same regimen of traditional yogurt (N=56) in a randomized, placebo-controlled, double-blind clinical trial in Tanzania [57].

Two randomized, placebo-controlled and blinded trials evaluated H. pylori-infected adults, the results of which suggest probiotics may help eradicate and improve symptoms of H. pylori infections [58, 59]. In a trial in South Korea, a statistically significant decrease (p=0.04) in the urea breath test was found when baseline levels were compared to those observed after 8 weeks of daily 150 mL consumption of L. paracasei HP7 fermented milk; no differences were found, however, when the treatment arm (N=65) and control arm (N=63, 150 mL daily consumption of regular milk) were compared. Some measures of symptoms were improved in the treatment group (gastrointestinal symptoms, p=0.05 and the physical health domain score of the World Health Organization Quality of Life [WHOQOL]-BREF, p=0.03) [58]. The other clinical trial of H. pylori infected adults reported 360 mL daily consumption of L. johnsonii La1 fermented milk for three weeks was associated with decreased bacterial density in the antrum and corpus (p=0.02 and p=0.04, respectively), as well as a decrease in gastric inflammation scores (p=0.02 for the antrum) and activity of gastric inflammation scores (p=0.01 and p=0.02 for the antrum and corpus), compared to pre-intervention levels; similar differences were not found in the control group of regular milk consumption [59].

Dairy proteins (N=6, Table 6)

Six studies investigated the impact of bovine lactoferrin on the natural history of infections in adults, including COVID-19 (N=3) [60,61,62] and hepatitis C (N=3) [63,64,65] (Tables 1 and 6).

The three studies of COVID-19 patients recruited asymptomatic and mild/moderate COVID-19 patients during the early stage of the pandemic in 2020-2021 [60,61,62]. Although the evidence is limited by the small number of studies and recruited patients, the studies suggested lactoferrin reduced the time to SARS-CoV-2 seroconversion but did not affect symptom resolution. In a clinical trial in Italy, patients were given liposomal bovine lactoferrin 1 g orally or 16 mg intranasally daily (N=32), standard of care treatment (N=32) or no COVID-19 treatment (N=28); the patients receiving bovine lactoferrin had a statistically significant shorter mean time to achieving a SARS-CoV-2 RNA negative test (mean=14.25 days), compared to the standard of care treatment (mean=27.13 days, p<0.001) and no treatment groups (mean=32.61 days, p<0.001) [62]. A similar finding was observed in another Italian study; a retrospective cohort of 121 COVID-19 patients reported the median time to a SARS-CoV-2 RNA negative test was statistically significantly (p<0.001) shorter in those treated with bovine lactoferrin (median=15 days) compared to standard of care treatment (median=24 days) [61]. In a multivariate Cox regression model adjusting for other predictors of SARS-CoV-2 RNA negativization, a HR of 1.65 (95% CI=1.09-2.25) was reported for bovine lactoferrin. No statistically significant difference in the median time to symptom resolution was found in this study [61]. Similarly, a clinical trial in Egypt found no statistically significant difference in the number of participants reporting clinical symptoms of COVID-19 in the treatment groups (N=36) compared to the control group (N=18) after seven days of treatment with 200-400 mg bovine lactoferrin [60].

Three trials of bovine lactoferrin among chronic hepatitis C virus (HCV) patients were identified. The studies were randomized, double-blind, placebo-controlled trials conducted in Japan that evaluated the impact of 600-7200 mg of bovine lactoferrin for 12-24 weeks, with mixed findings [63,64,65]. A trial of 3600 mg bovine lactoferrin daily for 12 weeks (N=97), compared to placebo treatment (N=101), reported no difference in the virological response rate (≥50% decrease in serum HCV RNA) [64]. Likewise, a trial of 600 mg bovine lactoferrin daily with standard HCV treatment for 24 weeks (N=18), compared to standard HCV treatment alone (N=18), reported no difference in the sustained virological response rate (absence of serum HCV RNA) [63, 65]. In contrast, a statistically significant increase (p<0.05) in the virological response rate after 8 weeks of bovine lactoferrin monotherapy (N=42), compared to HCV standard therapy (N=55), was reported. Furthermore, among the patients responding at 8 weeks in this trial, a statistically significant increase in the sustained virological response rate was observed after 24 weeks of therapy (bovine lactoferrin + standard HCV therapy vs. standard HCV therapy). This trial also found a statistically significant reduction (p<0.05) in HCV RNA titers at 8 weeks compared to baseline among the patients in the intervention group [64].

Leukocytes (N=76), Supplemental Table 3

Supplemental Table 3 describes studies measuring the influence of dairy products and/or their components on levels of white blood cells and their components (i.e., granulocytes including neutrophils, basophils and eosinophils; monocytes; and lymphocytes, including T-cells, B-cells and natural killer [NK] cells). Other in vitro measures of immune function were identified, including studies of neutrophil activity, lymphocyte proliferation/activation/transformation, NK cell activity/function/cytotoxicity, and the phagocytic activity, tumoricidal activity and oxidative burst capacity of leukocytes. Additional file 2 provides detailed summaries of the evidence. The studies of leukocytes and probiotics and dairy proteins are summarized in Figs. 2 and 3, respectively. No consistent changes in leukocyte levels and function were observed for any type of whole dairy product or their components. Probiotics and dairy proteins appear to enhance NK cell levels/activity and the phagocytic process in a larger proportion of studies with these outcomes (Figs. 2 and 3). Isolated responses were not consistent across populations, however, and the clinical relevance of these biomarkers of immune response is not clear.

Fig. 2

Studies of probiotics administered on leukocyte levels/proliferation/activity, by probiotic strain (N=28)

Fig. 3

Studies of milk proteins on leukocyte levels/proliferation/activity, by health status (N=19)

Cytokines (N=47), Supplemental Table 4

Supplemental Table 4 describes studies that evaluated the effect of dairy products or their components on cytokine production, including the ILs of interest, interferons (IFN), transforming growth factor (TGF)-β, and chemokines of interest. Additional file 2 provides detailed summaries of the evidence. Overall, the results of these studies were conflicting, with most studies reporting no significant impact of dairy products (including whole dairy products, probiotics specifically, and dairy proteins) on cytokine production. Interpretation of results is challenging as biological or clinical relevance was not considered.

Discussion

Based on this SLR, the cumulative available evidence suggests daily ingestion of dairy products fermented with probiotics from the genus Lactobacillus for ~1-3 months may reduce the risk of acquiring common infectious diseases (in particular URTI, cold and influenza) and improve the duration/severity of these diseases. The evidence base is suggestive, but findings are difficult to reconcile because of heterogeneity in the age/health status of the study population, the bacterial strains administered, and the statistics measured. Some studies that observed no difference in infectious disease incidence between treatment and control groups cited the low number of observed incident events in the study population [32, 39] or an overall inadequate sample size [26, 28] as possible explanations for their negative findings. Of note, a recent meta-analysis on the effect of probiotic fermented dairy products on the incidence of RTIs combined studies published through October 2020, thus addressing any issues with insufficient sample size/power [15]. In this meta-analysis, consumption of probiotic fermented dairy products had a significant protective effect against RTIs in the overall analysis (RR=0.81, 95% CI=0.74-0.89) and separately in children (RR=0.82, 95% CI=0.73-0.93), adults (RR=0.81, 95% CI=0.66-1.00) and elderly populations (RR=0.78, 95% CI=0.61-0.98). The benefit was restricted to Lactobacillus supplemented products (RR=0.81, 95% CI=0.74-0.90), although only two studies were available on Bifidobacterium. Disease-specific analyses showed benefits for URTI, pneumonia and the common cold, but marginal benefits for LRTI. A beneficial role for probiotic fermented dairy products in the prevention of acute infections is consistent with the results of clinical trials of probiotics given in power or pill form; in a recent meta-analysis of studies evaluating the impact of probiotics given in any form on the incidence of URTIs, a combined RR of 0.76 (95% CI=0.67-0.87) was reported for at least one URTI event, with low-certainty evidence [76].

Nearly all studies in our SLR that found no association between probiotic fermented products and the incidence of common infectious diseases reported improvements in some measure of disease severity/symptoms/duration [21, 22, 28, 32, 34, 38, 39], suggesting that even if the immune modulating effect of probiotics are not significant enough to prevent illness, they may still improve the course of disease. These findings are particularly relevant for older populations that are known to have age-related decrements in immune function and a higher incidence of infection, burden of disease, and more severe complications (e.g., with influenza [77]). The incorporation of fermented dairy products into residential elderly settings may be an easy and potentially impactful nutritional intervention to slow the spread and impact of infectious diseases in these settings. While the USDA Dietary Guidelines do not specifically provide a recommendation for fermented dairy products, adults may consider incorporating fermented products into the recommended 3 cups of dairy per day [8].

A small group of studies (N=3) evaluated whether lactoferrin monotherapy (200-600 mg daily) can reduce the incidence and burden of common infectious diseases, with mixed results [44,45,46]. Findings from the three studies of lactoferrin as an enteral supplement to pre-term infants to prevent sepsis were also mixed, which may be due to variability in iron saturation, the route of administration and the dosing schedule [47, 48, 75]. A recent SLR and meta-analysis of lactoferrin supplementation for late-onset sepsis in preterm infants reported a combined RR of 0.82 (95% CI=0.74-0.91) with low-quality evidence [14]. Thus, nutritional intervention with lactoferrin may be a promising strategy to boost human lactoferrin from mother’s milk and prevent infections in infants, although larger and more detailed analyses are required. This intervention may not be directly applicable to dairy products, however, as the concentration of lactoferrin is lower in bovine milk (around 25-75 mg in a glass of milk [78]) and dairy products are not recommended until six months of age.

Two cohorts reported COVID-19 seropositive patients were significantly more likely to report a higher intake of dairy products (in particular high-fat dairy products), compared to COVID seronegative patients [50, 51]. While control for confounding variables was attempted in these studies, residual confounding cannot be ruled out and would be consistent with the weak associations observed (OR=~1.1-1.5). Additional studies are recommended to reconcile these findings, with disease measurement based on clinically confirmed incident infections and measures of verified exposures linked more closely in time to disease incidence.

Our SLR identified a wide variety of potential applications for dairy products/components to improve the natural history of infectious diseases, likely due to the antimicrobial nature of lactoferrin and probiotics, although the available evidence in each research area is small and further research is required. Clinical trials suggest virological/bacteriological burden is reduced with lactoferrin (for COVID-19 [61, 62] and HCV [63,64,65]) and with probiotic [for H. pylori [59] and HIV [54]) treatment; symptoms of these conditions were also reduced with nutritional intervention in some studies [55, 56, 58]. Among persons with HIV/AIDS, the gut-associated lymphoid tissue is a major site of HIV replication and, therefore, represents a vulnerability to these patients, including the development of opportunistic infections. Probiotics can reinforce mucosal barrier function in the gastrointestinal system and modulate immune responses in the intestinal epithelium to improve outcomes in persons with HIV/AIDS. Food based interventions, such as probiotic yogurt, could help delay the progression of HIV/AIDs, particularly in populations with limited access to anti-retroviral treatment. Our review also found that probiotics may be useful to patients with H. pylori infection, a common bacterium that colonizes the gastric epithelium and increases the risk for stomach cancer; in addition to supporting gut health in the context of antibiotic treatment for H. pylori elimination, the evidence suggests Lactobacillus strains may be bactericidal in the gut and have an independent suppressive effect on H. pylori [58, 59]. The significant reduction in the time to COVID-19 negativization in two Italian studies suggests that lactoferrin improves viral clearance [9, 62], but the relevance of this research to dairy products is unknown. This finding should be confirmed in future studies, as a shorter time to COVID-19 seroconversion could limit the spread of infection.

One proposed mechanism for these observed effects is a modulation of the immune system by dairy product components. No consistent changes in white blood cells or cytokine production were observed for any type of whole dairy product or their components, among healthy and immune compromised populations (Supplemental Table 3, Supplemental Table 4, Figs. 2 and 3). Probiotics appeared to enhance natural killer cell levels/activity and the phagocytic process in a larger proportion of studies with these outcomes [79,80,81,82,83,84], suggesting this mechanism could play a key role in the reduction and/or burden of infections. Isolated responses were not consistent across populations, however, and the clinical relevance of these biomarkers of immune response is not clear. Limitations of this group of studies include small sample sizes, varied methods for measuring these biomarkers, a short intervention duration, lack of adjustment for multiple comparisons, and lack of an appropriate control group. Future clinical trials should continue to quantify biomarkers of immune function concurrently with disease incidence as measured by discrete antibody titers.

The main strength of this SLR is that the scope was broad, with few restrictions on exposures, outcomes, or study population characteristics. As such, this SLR provides a comprehensive scoping of the available evidence. While other reviews and meta-analyses provided a summary of specific dairy products and particular components, this SLR summarized all dairy products, including traditional and fermented products, and dairy proteins. Furthermore, this SLR included data on the impact of dairy products/components on leukocytes and cytokines, to potentially connect the epidemiologic findings with a mechanism. The quality of the evidence base is relatively strong, with the 96% of the studies classified as “positive” or “neutral” based on RoB assessment. This SLR also identified largely randomized, double-blind, controlled trials which are considered to be one of the strongest forms of epidemiologic evidence. Another strength of this review is that we thoroughly evaluated the component of exposure under study to differentiate the impact of the full dairy matrix versus the impact of specific probiotics strains; this approach has not been used in the previous reviews of this topic and allows for a better understanding of which dairy component may be bioactive.

While this SLR suggests a beneficial role for dairy in the incidence and natural history of infection, the interpretation of these findings is limited by substantial heterogeneity in study features, including the exposure, exposure dose/duration, the probiotic strain, the statistics measured, the infectious disease outcome, and the age and comorbidities of the study population. There is also substantial heterogeneity in how disease incidence was measured, with some studies relying solely on symptom report from a questionnaire. The lack of standardized RTI diagnosis, especially in older adults [85], further complicates the interpretation of these studies. The probiotics evaluated in the included studies comprised a wide variety of species and strains, both naturally occurring and experimental. It is possible that probiotics’ immune-modulating effect is strain-specific and, thus, the positive or negative findings may be related to strain-specific variation. Due to this heterogeneity, quantitative synthesis was not considered as we did not have sufficient studies with similar population types and exposures, although our qualitative synthesis was consistent with broader meta-analyses that have been attempted in the various areas [14,15,16, 86]. The evidence base is also limited by a lack of adjustment for the numerous factors that modulate the risk of infection, including nutrition, sleep, exercise and vaccination status. Other factors that may influence the efficacy of probiotics include genetic factors or the individual composition of gut microbiota.

Another major limitation of our review is the search terms were designed to capture the existing literature on a broad topic (i.e., dairy products and immune function) and, therefore, may have lacked the detail required to identify the universe of studies on each of the identified outcomes. The exclusion of the search term “probiotic”, for example, may have limited our search. Nevertheless, our conclusions are similar to reviews that have restricted their exposure of interest to probiotics in general [14, 16] and probiotic fermented foods specifically [15, 86]. Furthermore, our search strategy was not designed to capture the universe of studies measuring exposures assessed through food frequency questionnaires and we may have only captured those with keywords available in the abstract.

By summarizing the existing literature on this topic and providing a critical qualitative appraisal, this review plays an important role in that it provides a roadmap for valuable future research. A consortium of multicenter, randomized, placebo-controlled trials may be beneficial, with a range of specified exposure durations/doses, focused probiotic strains/dairy proteins, and clinically relevant outcomes (i.e., disease incidence based on objective antibody titers, when available) that are investigated along with longitudinal leukocyte and cytokine levels. Studies should incorporate sufficient numbers of patients to power their studies appropriately, given the background rate of infectious disease incidence in the underlying populations. Additional trials on the impact of traditional yogurt and milk would also be helpful to understand whether these products can be impactful without probiotic supplementation.

Conclusions

This SLR identified a wide variety of potential applications for dairy products/components to improve infectious disease outcomes, with the strongest evidence available for a bioactive role for probiotics. The evidence base is diverse, with limited studies available on specific exposures and outcomes.

Probiotics delivered through dairy products represent a promising nutritional intervention for reducing the incidence and burden of CIDs (including reducing disease severity/symptoms/duration), although additional research is required. Adjuvant fermented dairy products could be an alternative program for preventing infection that is easy, acceptable and very impactful, given the substantial morbidity and economic burden associated with CIDs. Numerous potential antimicrobial applications of lactoferrin and probiotics were identified, including reducing the risk of sepsis, improving the symptomatic burden of HIV, reducing HCV burden, and improving the course of H. pylori, although the evidence base was small and the relevance of this research to dairy products is unknown. Coordinated research programs are recommended in each disease area where the chosen exposures and the dosing schedule are based on mechanistic research, outcomes are based on clinical measures and biomarkers are tracked longitudinally to potentially correlate with clinical outcomes.