ABSTRACT
Objective: To analyze the concordance and agreement between bioimpedance spectroscopy (BIS) and anthropometry for the diagnosis of protein energy wasting (PEW) in chronic peritoneal dialysis patients. Methods: Prospective, multi-center, observational study using multifrequency bioimpedance device (Body Composition Monitor -BCM ® - Fresenius Medical Care) and anthropometry for the diagnosis of PEW as recommended by the International Society of Renal Nutrition and Metabolism (ISRNM). Cohen's kappa was the main test used to analyze concordance and a Bland-Altmann curve was built to evaluate the agreement between both methods. Results: We included 137 patients from three PD clinics. The mean age of the study population was 57.7 ± 14.9, 47.8% had diabetes, and 52.2% were male. We calculated the scores for PEW diagnosis at 3 and 6 months after the first collection (T3 and T6) and on average 40% of the study population were diagnosed with PEW. The concordance in the diagnosis of PEW was only moderate between anthropometry and BIS at both T3 and T6. The main factor responsible for our results was a low to moderate correlation for muscle mass in kilograms, with an r-squared (R2) of 0.35. The agreement was poor, with a difference of more than 10 kg of muscle mass on average and with more than a quarter of all cases beyond the limits of agreements. Conclusion: Current diagnosis of PEW may differ depending on the tools used to measure muscle mass in peritoneal dialysis patients.
ABSTRACT
OBJECTIVES: Constipation is a multifactorial gastrointestinal disorder commonly found in hemodialysis (HD) patients. In this multicenter cross-sectional study, we aimed to evaluate the prevalence and factors associated with constipation, including the frequency of dietary fiber intake. METHODS: HD patients from 4 dialysis clinics were invited. Participants answered a questionnaire which included Roma IV criteria to assess constipation status, use of medications, and lifestyle habits. A food frequency questionnaire with 7 response options on the main dietary fiber sources (fruits, vegetables, legumes, whole grains, and seeds) was also applied. Answers were transformed into a score to estimate the weekly intake frequency, and every score point corresponded to one time per week. Demographical and laboratory data were obtained from medical records. Univariate analysis was used to compare participants according to constipation status, and variables with P < .20 were included in the regression analysis model. RESULTS: Three hundred five HD patients were included (male: 51%; age: 52.2 ± 14.7 years old; HD vintage: 46 (19-82) months). Ninety-three participants had constipation (30.5%). Median (interquartile) food frequency questionnaire scores were as follows: fruits: 6 (2-14); vegetables: 6 (3-10); legumes: 3 (1-7); whole grain: 0 (0-1); and seeds: 0 (0-0). In univariate analysis, participants with constipation were significantly older, had lower literacy, higher prevalence of diabetes, and lower total beverage intake. The logistic regression analysis model also included body mass index, wheelchair need, sedentarism, fruits score, and seeds score (all with P < .20 in the univariate analysis). The independent predictors of constipation were diabetes (odds ratio = 1.96, 95% confidence interval 1.07-3.6, P = .03) and fruits intake score (odds ratio = 0.95, 95% confidence interval 0.91-0.99, P = .04). CONCLUSIONS: Almost one-third of the participants had constipation. The independent determinants of constipation were diabetes and lower frequency of fruit intake. Nutritional counseling to increase fiber intake can potentially decrease the prevalence of constipation in this population.
