Food and Behaviour Research

Donate Log In

Does milk cause constipation? A crossover dietary trial.

Crowley ET, Williams LT, Roberts TK, Dunstan RH, Jones PD. (2013) Nutrients.  5(1) 253-66. 

Web URL: View this and related abstracts via PubMed here. Free full text of this article is available online.


The aims of this study were to:

(1) determine whether replacement of cow's milk protein with soy resolves Chronic Functional Constipation (CFC); and

(2) investigate the effects of cow's milk β casein A1 and cow's milk β casein A2 on CFC.

Children diagnosed with CFC were recruited to one of two crossover trials: Trial 1 compared the effects of cow's milk and soy milk; Trial 2 compared the effects of cow's milk β casein A1 and cow's milk β casein A2. Resolution of constipation was defined as greater than eight bowel motions during a two week intervention.

Thirteen children (18 to 144 months) participated in Trial 1 (6 boys, 7 girls). Nine participants who completed the soy epoch all experienced resolution (p < 0.05).

Thirty-nine children (21 to 144 months) participated in Trial 2 (25 boys, 14 girls). Resolution of constipation was highest during the washout epoch, 81%; followed by cow's milk β casein A2, 79%; and cow's milk β casein A1, 57%; however, the proportions did not differ statistically.

The results of Trial 1 demonstrate an association between CFC and cow's milk consumption but Trial 2 failed to show an effect from type of casein. Some other component in cow's milk common to both A1 and A2 milk may be causing a problem in these susceptible children.


This two-part study was designed to test two related hypotheses, viz

1) that substitution of soy milk for cows' milk might help alleviate constipation in children, and

2) that constipation might also be alleviated by consumption of cows milk containing A2 but not A1 beta-casein (standard cows' milk contains both forms of this milk protein)

Findings supported the first hypothesis.  With respect to the second, although results were in the predicted direction (resolution of constipaton for A2 milk was 79%, versus 57% for A1 milk), the group difference was not statistically significant.

The authors concluded that some component of cows' milk other than A1 beta-casein may also contribute to constipation (as can many other factors, of course). However, the small number of children meant the sudy had only low 'statistical power' (i.e. power to detect a real difference if there is one).

The A1 hypothesis is based on the fact that digestion of A1 but not A2 milk gives rise to opioid peptides, and opioids are known to slow digestive transit time.

Findings from subsequent studies in both animals and humans have since confirmed the link between A1 beta-casein and slow transit time - so it is possible that Study 2 was too small and underpowered for the observed A1/A2 group difference to reach significance. See: