Introduction
• Approximately 90% of consumers in China [1] and South-East Asia [2, 3] are regarded as「lactose intolerant (LI)」.
• Recent published [4] data show that the gastrointestinal (GI) discomfort ascribed to LI may instead be due to intolerance to the A1 variant of bovine β-casein.
• Data also show that ～70% of Chinese consumers may be「tolerant」to A1-protein-free milk, with LI a secondary effect of A1-protein-induced GI inflammation, which may be mediated at least in part by bioactive peptides released from A1 β-casein.
• A key objective of the「a2 Milk™ for Gut Comfort」HVN contestable project is to apply novel techniques for identifying biomarkers of dairy protein intolerance.
• For the first clinical trial (the「a2 Milk™ for Gut Comfort」(aMiGo) trial), breath and plasma were the matrices used to address this objective.

Hypotheses
1. That dairy protein intolerance can be separated from LI on the basis of acute symptoms following dairy consumption.
2. That both the breath volatile and plasma peptide profile will be altered (a) in dairy intolerant individuals, and (b) in individuals consuming a2 Milk™ relative to those consuming conventional milk.

Methods
Female participants were assigned to one of three groups on the basis of lactose challenge screening: dairy tolerant (n = 10), lactose-intolerant malabsorbers (n = 10) and non-lactose dairy-intolerant (NLDI; n = 20; self-report as intolerant, but not lactose intolerant).Participants were then randomised to ingest on 3 separate occasions 750 mL conventional milk, a2 Milk™, or lactose-free conventional milk. Self-reported symptoms were recorded by visual analog scale (VAS) half-hourly over 3 hours.

Results
Intestinal symptoms differ between participant groups
• Tolerant subjects experienced no notable symptoms in response to milk consumption.
• NLDI subjects could be clearly distinguished from LI subjects based on symptoms (Fig 3).Based on visual analogue scale (VAS) scoring of symptoms, NLDI subjects reported increased abdominal distension (a) and bloating (b) after drinking milk, which subsided by 120 minutes. LI subjects experienced these symptoms throughout the post-meal period, without relief. LI subjects experienced delayed onset (past 90 minutes) of fecal urgency (c) and abdominal cramps (d), which were not experienced by NLDI subjects.

Breath VOC profiles change over time, but do not differ between participant groups, or in response to different milks.

Conclusions
• NLDI can be defined by a distinct subset of GI symptoms experienced after dairy consumption, independent of lactose tolerance.
• Rather than lower GI symptoms associated with lactose malabsorption, NDLI is characterised by early-onset upper GI symptoms that resolve rapidly.
• Breath VOC profiles change over time following milk consumption, suggesting they are influenced by dietary factors.

Future Work
• Optimisation of the breath metabolomics method is underway to address inter individual variation in VOC concentrations, which may be influenced by factors such as toothpaste used before sampling, and pattern of breathing during sample collection.
• The optimised protocol will be applied during a second clinical trial, to further assess the suitability of VOC metabolomics for identifying a biomarker for dairy protein intolerance, or to distinguish GI responses to different milks.
• Analysis of plasma peptide profiles is currently underway.