Water-loss (intracellular) dehydration assessed using urinary tests: how well do they work? Diagnostic accuracy in older people1–3

• Published in: The American journal of clinical nutrition Vol. 104; no. 1; pp. 121 - 131
• Main Authors: Hooper, Lee, Bunn, Diane K, Abdelhamid, Asmaa, Gillings, Rachel, Jennings, Amy, Maas, Katie, Millar, Sophie, Twomlow, Elizabeth, Hunter, Paul R, Shepstone, Lee, Potter, John F, Fairweather-Tait, Susan J
• Format: Journal Article
• Language: English
• Published: Elsevier Inc 01.07.2016
• Subjects: aged; dehydration; osmolar concentration; sensitivity and specificity; specific gravity; urinalysis; specific gravity; urinalysis; osmolar concentration; NU-AGE; MMSE; NNUH; ROC; aged; USG; dehydration; DRIE; sensitivity and specificity
• ISSN: 0002-9165; 1938-3207
• DOI: 10.3945/ajcn.115.119925

Background: Water-loss dehydration (hypertonic, hyperosmotic, or intracellular dehydration) is due to insufficient fluid intake and is distinct from hypovolemia due to excess fluid losses. Water-loss dehydration is associated with poor health outcomes such as disability and mortality in older people. Urine specific gravity (USG), urine color, and urine osmolality have been widely advocated for screening for dehydration in older adults. Objective: We assessed the diagnostic accuracy of urinary measures to screen for water-loss dehydration in older people. Design: This was a diagnostic accuracy study of people aged ≥65 y taking part in the DRIE (Dehydration Recognition In our Elders; living in long-term care) or NU-AGE (Dietary Strategies for Healthy Ageing in Europe; living in the community) studies. The reference standard was serum osmolality, and index tests included USG, urine color, urine osmolality, urine cloudiness, additional dipstick measures, ability to provide a urine sample, and the volume of a random urine sample. Minimum useful diagnostic accuracy was set at sensitivity and specificity ≥70% or a receiver operating characteristic plot area under the curve ≥0.70. Results: DRIE participants (women: 67%; mean age: 86 y; n = 162) had more limited cognitive and functional abilities than did NU-AGE participants (women: 64%; mean age: 70 y; n = 151). Nineteen percent of DRIE participants and 22% of NU-AGE participants were dehydrated (serum osmolality >300 mOsm/kg). Neither USG nor any other potential urinary tests were usefully diagnostic for water-loss dehydration. Conclusions: Although USG, urine color, and urinary osmolality have been widely advocated for screening for dehydration in older adults, we show, in the largest study to date to our knowledge, that their diagnostic accuracy is too low to be useful, and these measures should not be used to indicate hydration status in older people (either alone or as part of a wider tranche of tests). There is a need to develop simple, inexpensive, and noninvasive tools for the assessment of dehydration in older people. The DRIE study was registered at www.researchregister.org.uk as 122273. The NU-AGE trial was registered at clinicialtrials.gov as NCT01754012.