Lederle, F.A., Busch, D.L., Mattox, K.M., West, M.J., & Aske, D.M. (1990). Cost-effective treatment of constipation in the elderly: A randomized double-blind comparison of sorbitol and lactulose. American Journal of Medicine, 89, 597–601.

To evaluate the use of sorbitol as an inexpensive alternative to lactulose for treating constipation in older adults.

Lactulose and 70% sorbitol (0–60 ml daily) were given for four weeks. During the treatment period, patients were instructed to begin taking 30 ml of the study laxative at bedtime, thereafter adjusting the dose as needed from 0–60 ml. Patients were instructed to maintain high dietary fiber and avoid sources of free fructose such as apples and pears.

The length of the wash-in and washout periods was based on previous studies, which showed up to three to four days are required for lactulose to take effect and the carryover effect after cessation of lactulose is about six to seven days. The study began with a two-week lead-in period, during which patients received lactulose in a single-blind fashion. This was followed sequentially by washout period A (two weeks), treatment period A (four weeks), washout period B (two weeks), and treatment period B (four weeks). The purpose of the lead-in period was to ensure (a) the patient tolerated lactulose, (b) the patient understood how to fill out the diary, and (c) the conditions preceding the two treatment periods were similar. At the end of washout period A, patients were randomly assigned to receive one of the two study laxatives in treatment period A, with the other laxative being used in treatment period B.

The primary endpoints of the study were average number of bowel movements per week and the average number of days per week on which bowel movements occurred.

• The study reported on a sample of 40 ambulatory men. However, only 30 patients were accounted for (see Setting).
• Mean patient age was 72 years (range 65–86).
• Patients were included in the study if they had a history of constipation for a duration of at least one year. The study was designed to recruit ambulatory patients with the most severe constipation who were not taking narcotics.
• Twenty-one patients took bulk-forming agents throughout the study (psyllium in all cases).

• Minneapolis Veterans Affairs Medical Center in Minnesota from September 1988 to July 1989
• Five patients were residents of the Minnesota Veterans Home and the other 25 were outpatients.

This was a randomized, double-blind, crossover study.

• A patient diary was used to record bowel consistency (hard, soft, or loose), daily dose of study laxative, use of any other laxative or enema, and occurrence and severity of seven symptom categories (bloating, cramping, excessive flatulence, nausea, diarrhea, fecal incontinence, and other).
• Statistical analysis: t-test method for crossover trials (two-sided alpha)

• Thirty patients completed the study. Nine withdrew prior to randomization because of lactulose intolerance, and one patient dropped out after randomization.
• No significant differences existed between sorbitol and lactulose in any outcome measured except nausea, which increased with lactulose use (p < 0.05).
• Mean number of bowel movements per week was 6.71 with sorbitol and 7.02 with lactulose (95% confidence interval [CI] of the difference [–0.43, 1.06]).
• Mean number of days per week with bowel movements was 5.23 with sorbitol and 5.31 with lactulose (95% CI of the difference [–0.32, 0.48]).

The results supported the hypothesis that sorbitol and lactulose have no clinically or statistically significant difference in laxative effect. Sorbitol can be recommended as a cost-effective alternative to lactulose for the treatment of constipation in older men.

• The study had a short duration and was limited to ambulatory older men. Caution should be used in extrapolating these findings to women, younger patients, or those who are bedridden.
• The withdrawal of patients because they did not tolerate lactulose may have biased the study.
• The sample size was smaller than the 32 patients required to provide sufficient power for statistical significance of the results.