Predictors of turnover of lower gastrointestinal symptoms in diabetes mellitus

The definitive version is available at www.blackwell-synergy.comBackground: Both the natural history of GI symptoms and factors influencing symptom turnover in diabetes mellitus are unknown. We aimed to determine the natural history of GI symptoms in diabetes mellitus over a 3-yr period. METHODS: Subjects with predominantly Type 2 diabetes on the mailing list of Diabetes Australia were recruited and completed a validated questionnaire (n = 892); of these, 64% completed the questionnaire again 3 yr later. Subjects were classified as never having GI symptoms (abdominal pain, constipation, diarrhea, fecal incontinence), fluctuating (only on one survey), or persistent symptoms. RESULTS: Symptoms more often fluctuated than persisted, but the prevalence at recruitment and 3 yr later was similar. Predictors of symptom turnover varied. In a multivariate model, abnormal sweating and diabetic foot problems predicted symptom turnover for abdominal pain (OR = 2.01, 95% CI = 1.07–3.76 and 2.54, 1.23–5.24, respectively), and paresthesiae ("pins and needles") and foot problems predicted fecal incontinence (OR = 2.24, 1.06–4.75 and 2.38, 1.04–5.45, respectively), but not constipation or diarrhea. Symptom turnover in constipation was associated with depression (OR = 2.73, CI = 1.89–3.94) and neuroticism (OR = 0.57, CI = 0.36–0.89), and depression was associated with abdominal pain (OR = 1.56, CI = 1.02–2.38), but psychological factors were not related to the other symptoms. The odds for symptoms persisting were not generally related to type or duration of diabetes, or self-rated glycemic control. CONCLUSION: The natural history of GI symptoms in diabetes has been described for the first time; symptoms may either persist or fluctuate, but the prevalence is constant because symptom onset is balanced by disappearance. Glycemic control does not seem to predict symptom changeNicholas J. Talley, Stuart Howell, Michael P. Jones, Michael Horowit