Relationship between prescription medication use and chronic constipation and chronic diarrhea in <scp>US</scp> adults: A propensity score matching study

Abstract This study aimed to clarify the relationship between all prescription medication with chronic diarrhea (CD) and chronic constipation (CC) in adults. Population data was extracted from the 2007–2008 and 2009–2010 National Health and Nutritional Examination Surveys (NHANES). CD and CC were defined according to the Bristol Stool Form Scale (BSFS), and frequent laxative users were also defined as having CC. Prescription medication was obtained from a questionnaire. Multivariable logistic regression was used to calculate odds ratios (ORs) and 95% confidence intervals (CIs); restricted cubic spline (RCS) dose response curve was used to estimate the relationship between prescription medication use and CD. 1:1 propensity score matching (PSM) was performed to test the validity of the results. No significant associations between prescription medication use and CC were found. RCS showed a nonlinear positive association between prescription medication use and CD. Multivariate logistic regression analysis showed that the adjusted ORs (95% CI) compared with reference in model 1, 2, and 3 were 1.575 (1.297–1.912), p &lt; 0.001, 1.384 (1.126–1.700), p = 0.002, and 1.350 (1.098–1.661), p = 0.004. The same results were found in the subgroup analyses by sex. In addition, similar conclusions were reached after analysis of PSM in the normal bowel habits and CD. This study indicated that prescription medication use may induce the occurrence of CD. This finding is of particular clinical significance for vulnerable populations, who require complex polypharmacy regimens.