Epidemiologic Associations for Protection of Infants:  2015 Clinical Study

This post is a follow-up to a previous ACSH post that explored relationships between consumption of raw and pasteurized bovine (cow) milk among cohorts of children and adults and associated infectious and chronic disease outcomes. This post will focus on evidence from the study by Loss, Consumption of unprocessed cow’s milk protects infants from common respiratory infections, and on its interpretation in light of advances in knowledge over the past decade. 

Loss and colleagues designed a prospective birth cohort study (Protection against Allergy—Study in Rural Environments, or PASTURE) to estimate the effects of raw and processed cow's milk on common infections in infants, including cold or runny nose (rhinitis), fever, ear infection (otitis), cough, and diarrhea, during the first year of life. Acknowledging the complexity of protection against infectious and chronic diseases, they generated estimates of effects on health outcomes, controlling for 16 potentially confounding factors, including breastfeeding and farm effects, for the period from birth to 53 weeks of age. 

The background motivation for the Loss study, as provided by the authors, was their expectation that breast milk and cow's milk might exert similar anti-infective effects in human infants, given co-evolutionary adaptations since humans first domesticated cows approximately 10,000 years ago. Common antimicrobial and immunomodulatory effects from bioactive components present in both bovine and human milk were thought, even in 2015, to provide passive immunity to infants and prevent or attenuate infections. Indeed, the subsequent decade of research has further advanced knowledge of the bioactive components of breast milk and bovine milk, including the milk microbiota, which are powerful immunomodulators and direct and indirect competitors against bacterial and viral pathogens of concern. 

On the one hand, the authors suggested that pasteurization of cow's milk over the past century solved one problem (milk-borne infections), but also acknowledged the possibility that pasteurization might have unintended adverse consequences, including the loss of beneficial bioactive components denatured by heat. Industrial processing of cow's milk includes heating to 161-167°F for 15 to 30 seconds for traditional pasteurization and to 275°F for 5 seconds for ultra-high-temperature (UHT) processing, essentially sterilizing the milk and rendering it shelf-stable. 

The authors selected UHT milk feeding as the reference category to assess the significance of milk effects on health outcomes in this cohort of infants, given the strong impact of UHT thermal processing on heat-sensitive components of raw milk associated with protective health effects. For this reason (potential degradation, denaturation, or destruction of beneficial bioactive milk components by heat), the authors chose the most severe heat treatment, UHT, as the reference treatment to estimate effects for raw and traditionally pasteurized milks. In addition, UHT milk is commonly consumed in Europe (>90% of consumers in France and Italy, and near zero in Denmark), including 113 infants (11.5%) in the Loss study cohort who were fed UHT at least once in their first year of life. No reliable source of information about UHT milk consumption in the US was identified. An unconfirmed marketing estimate for the US suggests that UHT milk may account for ~31% of total milk sales.

The strongest protective effects on health outcomes from any cow's milk were observed for raw milk, with adjustments for 16 potential confounding factors, including farm effect and breastfeeding. Parents reported on the following health outcomes weekly.

• rhinitis (p=0.015) 
• respiratory infections (p=0.045) 
• otitis (p<0.001)
• fever (p=0.058) 

Early-life consumption of raw milk significantly reduced the risk of respiratory infections in infants by about 30%. The protective effects were attenuated with boiled raw milk, pasteurized milk, and UHT milk. Some protection against fever was also observed with boiled and pasteurized milks. Pasteurized milk provided no significant protection at the p=0.05 level for any disease metric except fever. The protective effects of raw breastmilk and raw cow milk were comparable, suggesting similar anti-infective or immunomodulatory properties of raw milks from human and bovine sources. 

Supporting evidence included the inverse association observed in children fed raw milk and high-sensitivity CRP (hsCRP) values [1]. Loss suggested that low hsCRP levels in infants fed raw milk implied a sustained anti-inflammatory or immunomodulatory effect of raw cow's milk, citing studies that support a contribution to chronic illness (obesity, respiratory impairment, asthma severity, and atherogenic lipid profiles) even with slightly elevated hsCRP values.

It is puzzling that Loss and colleagues did not directly report results for diarrhea, beyond stating that no ‘clear’ association between cow’s milk and diarrhea was observed. No details about feeding for the 319 infants with diarrhea were provided. An inquiry to Dr. Loss seeking more detailed feeding results for diarrheal illnesses is awaiting a response.

The Loss study is consistent with an extensive body of evidence supporting the protective effects of breastfeeding against diarrhea and other infections. Further, a meta-analysis reported that exclusive breastfeeding for 6 months reduced the odds of diarrhea by 43% among infants across 32 countries. Protection against diarrhea from raw human and cow’s milk would be consistent with the functions of bioactive factors, including ‘colonization resistance’ by the milk microbiota, which competes with pathogens and directly and indirectly modulates the immune system through bioactive components of raw milks. Further, I contributed to two studies documenting that both raw breastmilk and raw cow’s milk are now understood to ‘seed and feed’ the human gut and protect against infectious and chronic disease. Although neither raw breastmilk nor raw cow’s milk can be assumed to prevent diarrhea, data from multiple sources support raw milk as a healthful, though not risk-free, food.

The Loss study’s concluding paragraph acknowledged protective effects of raw milk against respiratory infections and fever but failed to note the absence of a ‘clear’ association between raw cow’s milk and diarrhea. Although the authors stated, “Once more, raw milk can confer life-threatening infectious disease,” this claim was not supported by their data. The authors cited three studies published between 2003 and 2009 on brucellosis and illness from E. coli O26:H, as well as other potential hazards. However, recent evidence challenges the common opinion that raw milk is inherently dangerous. Thus, the warning from Loss and colleagues about raw milk safety is inconsistent with the body of evidence presented in this large cohort study and merits further deliberation.

The major strengths of the Loss study include aspects of the study design.

• a large cohort of 983 infants representing 37,306 person-weeks of observation
• a prospective design controlling for breastfeeding, farm effects, and 14 other variables from birth to 53 weeks of age
• weekly recording of health outcomes in diaries to minimize recall bias
• weekly recording of feeding practices (raw ‘farm’ milk, boiled ‘farm’ milk; industrially processed (pasteurized, UHT) milk; breastmilk; formula)
• data on a biomarker of inflammation in blood samples from 602 of the 983 infants at age one year to address possible mechanisms of protection

Weaknesses include incomplete reporting on diarrhea, the lack of physician confirmation of parents’ weekly infection reports, and questions about the representativeness of the European cohort for children and for raw milk from other nations. 

[1] High-sensitivity C-reactive protein (hsCRP) is measured by blood testing as an indicator of the level of inflammation.