Table 2.

Characteristics of included studies.

RQ 1	Publication	Study Type if Applicable: Population	Relevant Findings	Level of Evidence 2	Consistency with ESPEN Guideline
1.1	Lacey et al. 2019 [1]	Expert opinion n = 12, experts of varying specialties	-Plasma osmolality (in mOsm/kg) is an objective surrogate marker of low-intake dehydration -Osmolality is underutilized in clinical practice	4	Yes
1.2	Lacey et al. 2019 [1]	See above	-When plasma osmolality cannot be assessed, the authors recommend to calculate plasma osmolarity (CO) (in mmol/L)	4	Yes
	Munk et al. 2021 [11]	Diagnostic accuracy study n = 90, older adults from emergency medical department	-32% with impending (≥295–300 mOsm/kg) and 11% with current dehydration (>300 mOsm/kg) -Significant association between CO and Osmolality -Using CO is superior to current clinical practice	2+	Yes
	Woijszel et al. 2020 [12]	Cohort study n = 358, hospitalized older adults	-58% with dehydration (CO > 295 mmol/L) -Dehydration was more frequent in patients with e.g., multimorbidity, polypharmacy, hypertension, diabetes, chronic kidney disease,	2+	Yes
	Mantantzis et al. 2020 [13]	Cohort study n = 1047, community-dwelling older adults	-33% with dehydration (CO > 296 mmol/L) -Higher CO was associated with older age, more morbidities, and greater decline in cognitive functioning and well-being over time	2+	Yes
1.3	Bunn and Hooper 2019 [14]	Diagnostic accuracy study n = 188, care home residents	-Commonly used clinical signs and symptoms of low-intake dehydration (49 tested) inadequately discriminated between persons with or without low-intake dehydration -The authors suggest to use a 2-stage screening process instead of clinical signs and symptoms -1: CO, 2: Serum osmolality measurement if CO > 295 mmol/L	2+	Yes
	Lacey et al. 2019 [1]	See above	-Clinical signs and symptoms are not reliable outside extremes -They should not be used alone to detect dehydration	4	Yes
	Johnson & Hahn 2018 [15]	Cohort study n = 60, nursing home residents	-51% showed renal fluid conservation consistent with dehydration -Clinical signs of mucous membranes and tongue furrows did not correlate with serum osmolality -Clinical signs might rather reflect physical status and age than dehydration	2−	Yes
	Akdeniz et al. 2018 [16]	Cohort study n = 40, hospitalized older adults	-Average fluid intake was 1747 mL/day -20% were dehydrated and 23% had impeding dehydration, despite comparably high fluid intake -The clinical relevance of low fluid intake as a marker of dehydration is questionable	2+	Yes
	Ekman et al. 2020 [17]	Cohort study n = 38, rehabilitating older adults after hip surgery	-Average fluid intake was 1008 mL/day -18% with dehydration (≥300 mOsm/kg), those showed correlations to presence of tongue furrows and poor skin turgor -21% with concentrated urine, those showed correlations to low fluid intake and a decreased body weight	2−	No
2	Cook et al. 2019 [18]	Literature review care home residents	-Care home residents have low fluid intake -No hydration intervention alone was effective in addressing dehydration	4	Yes
	Painter et al. 2017 [6]	Literature review dementia and aged care facilities	-Intake of texture-modified fluids was associated with lower energy and fluid intake -Lack of evidence that texture modified fluids improve fluid intake and negative consequences (i.e., aspiration, pneumonia)	4	Yes
	Jimoh et al. 2019 [19]	Cohort study n = 22, long-term care residents	-Drinks are mostly consumed between meals -Residents with sufficient fluid intake were offered beverages more frequently and drank more with medications and before breakfast -Offering drinks more frequently might improve fluid intake	2-	Yes
	Marra et al. 2016 [20]	Cohort study n = 247, long-term-care residents	-31% with impending (295–300 mOsm/kg) and 38% with dehydration (>300 mOsm/kg) -Average water intake was 1147 mL/day -Variance in water intake was influenced by, e.g., type of liquid beverage (thin vs. thick), type of ONS, which could be targeted by nutritional interventions	2+	yes
	Bak et al. 2018 [21]	Pre-post study Two wards, nursing home residents	-Drinking vessels which were lightweight, with large handles and volume of 200–300 mL were preferred -Introduction of new vessels at breakfast improved fluid intake	n.a.	Yes
	Wilson et al. 2019 [22]	Pre-post study Two care homes, Care home residents	-Increased choice of beverages and opportunities to drink was associated with increased range of consumed fluid and higher fluid intake	n.a.	Yes
3	Masot et al. [23]	Literature review Older adults at different care levels	-Recommendations do not consider physiology of ageing and health problems of older people -Authors recommend 1.5–2.0 L of fluid per day	4	Yes
4	Polhuis et al. [24]	Randomized Trial n = 20, elderly community-dwelling men	-The intake of equal volumes of wine and spirits compared to non-alcoholic wine and water resulted in higher urine output within 4 h, but not 24 h -Moderate amounts of stronger alcoholic beverages showed short diuretic effects	1++	Yes

¹ RQ1: How should low-intake dehydration be identified in older persons? 1.1 by osmolality, 1.2 by osmolarity, 1.3 by clinical signs and bioelectrical impedance (BIA). RQ2: What interventions may help to support older persons to drink well and prevent low-intake dehydration? RQ3: How much should older people drink each day? RQ4: What should older people drink each day? ² Level of evidence according to the Scottish Intercollegiate Guidelines Network (SIGN) 1++ (highest), 4 (lowest) [4]. SIGN do not have tools for pre-post studies, so for two papers SIGN were not applicable (n.a.). CO = calculated osmolarity, BMI = body mass index, n.a. = not applicable, ONS = oral nutritional supplements, RQ = research question.