Isn't this what you'd expect sterically anyway? The A-value of OH is smaller than that of CH3, so you'd expect the CH3 to be anti- to the OH on the adjacent carbon, right?
But I suspect hydrogen bonding is probably pretty important here, although without knowing the solvent it's pretty hard to talk about hydrogen bonding.
I agree with spirochete too, there's probably some stereoelectronic (lone pair--->CO sigma*) action going on here too.
Taken together, I'm struggling to find anything that would actually favour the anti conformation in this case!