To the Editor.
—I am sympathetic to the apology by Conn for his "ert" placebo that was published in the June 1981 Archives (141:846-847). Unfortunately, the pH-lowering effect of sorbitol complicates interpretation of his encephalopathy trials even more than acknowledged in the editorial. The antimicrobial activity of the aminoglycoside antibiotics is notably diminished by increased acidity.1 Thus, the fact that "lactulose was found to be more effective than neomycin [sulfate] plus sorbitol" may not be as happy an observation as Conn sanguinely asserts. The decreased stool pH induced by sorbitol would have diminished the antibacterial effect of the neomycin, thus likely making it seem less effective than it would if used alone while, reciprocally, increasing the apparent effectiveness of lactulose together with neomycin.