In addition to malnutrition and renal function impairment, anemia is associated with hyponatremia in the elderly.

Hyponatremia is the most common electrolyte abnormality among the elderly living in long-term care facilities. In this study, we investigate the associated factors of hyponatremia, and its association with anemia in the institutionalized elderly in Taiwan. A total of 414 participants aged 65 years and above were recruited from eight long-term care facilities in 2002-2003. Baseline characteristics, medical records, and biomarkers were obtained. Hyponatremia was defined as a serum Na-concentration<135 mmol/l. Relationships between hyponatremia and the demographic and laboratory characteristics were tested using multiple logistic and linear regression analyses. The prevalence of hyponatremia and anemia was 14.7% and 56.0%, respectively. Anemia, hypouricemia, and the placement of tubes (including nasogastric tube, tracheostomy tube, and Foley catheter) were significantly associated with hyponatremia after adjustment for potential confounders using multiple logistic regression analysis. The adjusted odds ratios (OR) and 95% confidence interval (95%CI) for these three factors were 3.28 (1.40-7.69), 4.98 (2.18-11.36), 9.15 (3.33-25.12), respectively. Multiple linear regression analyses also showed that serum Na concentration was significantly associated with hemoglobin, uric acid, and number of tubes. In conclusion, it was found that anemia, the placement of tubes, and hypouricemia were associated with hyponatremia in the institutionalized elderly. In those with the above conditions, serum Na concentration should be monitored.

Predictive equation of resting energy expenditure in obese adult Taiwanese.

OBJECTIVE: To examine the accuracy of present resting energy expenditure (REE) predictive equations among obese adult Taiwanese, and obtain new predictive equations for this population. METHOD: A total of 80 participants (men = 42, women = 38) aged 18-64 years (mean 41.5 ± 13.1 years) with body mass index (BMI) ≥30 kg/m(2) were recruited from obesity clinic of a tertiary hospital. Anthropometric data including weight, height, waist circumference (WC), hip circumference (HipC) were collected. Body fat and fat free mass (FFM) were measured using bioelectric impedance analysis. Indirect calorimeter (Vmax 29n) was used for measurement of REE. The commonly used predictive equations (Harris-Benedict equation, Bernstein equation, Mifflin equation, Owen equation, Schofield equation, Cunningham equation, Wang equation, and Liu equation) were tested for group means and individual prediction accuracy for our target population. New predictive equations were formulated using multiple linear regression analyses. RESULT: The mean BMI was 34.3 ± 4.6 kg/m(2). All of the commonly used predictive equations showed statistically significant differences with the measured REE values. As for individual prediction accuracy, these predictive equations showed poor performance and only Mifflin equation provided an individual accuracy over 40%. The new predictive equation recommended for obese adult Taiwanese is below: REE (kcal/d) = 11.6 × weight (kg) - 204.84 × gender - 4.8 × height (cm) + 1594.6 (gender: men = 1; women = 2). CONCLUSION: The previous commonly used predictive equations were less accurate in obese Taiwanese. For this specific population, the newly developed equation should be applied for better prediction of REE.