Long Oral Presentation Nutrition Society of Australia 50th Annual Scientific Meeting 2026

Health effects of A2 versus conventional milk: a systematic review and meta-analysis, with assessment of A1/A2 β-casein composition and dose (147169)

Rizliya Visvanathan 1 2 , Nicole J Kellow 1 2 , Michael J Houghton 1 2 , Aimee L Dordevic 1 2
  1. Victorian Heart Institute , Monash University, Clayton, VIC, Australia
  2. Nutrition, Dietetics and Food, Monash University, Clayton, Victoria, Australia

A2 milk, containing only the A2 β-casein variant, is marketed as a more digestible and less pro-inflammatory alternative to conventional milk1. However, clinical evidence is inconsistent, and most trials do not consider the A1 exposure from comparator milk. Therefore, this work aimed to determine effects of A2 versus conventional milk on gastrointestinal, growth, and inflammatory/immune outcomes in adults or children of any health status, focusing on A1/A2 β-casein composition and dose.

A systematic search of 5 databases identified 29 publications representing 25 independent trials (n = 3,275 participants), with study durations ranging from 1–24 hours for acute studies and 3–120 days for chronic studies. For each trial, the A1/A2 β-casein composition and dose of the intervention and comparator milk were extracted. Gastrointestinal, growth, and inflammatory/immune outcomes were pooled using three-level random-effects meta-analyses where ≥4 studies (≥3 for growth) reported a common outcome. Risk of bias varied across studies, with 9 studies judged as low risk, 10 as high risk, and 8 as having some concerns.

Individual studies frequently reported gastrointestinal improvements with A2 milk, particularly in milk-sensitive populations; however, pooled analyses found no significant (p > 0.05) effects from acute or chronic feeding studies, with prediction intervals 2–4 times wider than confidence intervals. No significant differences were found between A2 and conventional milk for growth outcomes, including weight, height, head circumference, and BMI (all p > 0.11). Among the pooled inflammatory markers (CRP, IL-4, IgE, IgG, and IgG1) and GSH, five showed no significant differences. IL-4 showed a statistically significant pooled reduction (p = 0.014; SMD = −0.49; 95% CI −0.78 to −0.20); however, this was based on only four trials and the prediction interval crossed the null. Narrative synthesis suggested possible benefits for gastrointestinal symptoms and inflammatory/immune outcomes among milk-sensitive or lactose-intolerant individuals, although this was not consistent across all such studies. Importantly, only 10 of the 25 trials reported the comparator milk's A1 β-casein content, which ranged from 22% to 75%2,3. In some trials, the comparator milk contained mostly A1, while in others it already contained a high proportion of A2. Most of the A2 milk used in the trials was presumed to contain 100% A2 β-casein. Daily A1 exposure in comparator arms also varied more than five-fold (1.89–10.0 g/day). These findings highlight substantial variation in comparator milk composition, limiting interpretation of health outcomes and attribution of effects to the absence of A1.

Overall, individual study findings were not consistently supported by quantitative synthesis, with small evidence bases, wide prediction intervals, and variation in A1/A2 β-casein dose limiting certainty. Larger trials with laboratory-verified A1/A2 β-casein composition and dosing are needed, particularly in milk-sensitive populations.

  1. 1. Nogueira, ACR, Ferreira TO, Ribeiro FB et al. (2025) Int J Dairy Technol 78, 1-14.
  2. 2. Milan, AM, Shrestha A, Karlström HJ et al. (2020) Am J Clin Nutr 111, 149–160.
  3. 3. Ramakrishnan M, Mysore Saiprasad S & Savaiano DA (2024) Nutrients 16, 1-13.