Table 2.

Characteristics of articles comprising physical exercise programs

Study, country	Type of study and economic evaluation	Time horizon C: Cost H: Health outcome	Study population, Number of participants	IG: Intervention group CG: Control group	Outcome measures (measures of benefit)	Cost data; source	Perspective	Results
D’Amico [28] UK, 2016	RCT CEA CUA	C: 12 weeks H: 12 weeks	Community-dwelling persons with a clinical diagnosis of dementia according to ICD-10 and one or more significant BPSD symptoms defined by the NPI as well as a caregiver willing to participate in the exercise training ▪ 131 randomized dyads ▪ 113 completed (89%) ▪ economic analysis subsample 52 dyads	IG: Dyadic exercise regimen (individually tailored walking program, for 20–30 min at least five times per week for 12 weeks, in the first 6 weeks the intervention was facilitated by a registered exercise professional) CG: CAU	Primary ▪ NPI Secondary ▪ ZBI ▪ GHQ ▪ DEMQOL-Proxy ▪ QALY	CSRI completed by the caregiver; intervention cost	(I) Health & social care perspective (II) Societal perspective	Potentially cost-effective considering behavioral and psychological symptoms, not cost-effective considering QALY gains. Mean cost difference between the groups were not significantly different over 12 weeks. This holds for both perspectives. From the societal perspective, the ICER was £421 per incremental difference in NPI score.
Davis [33] Canada, 2013	RCT CEA	C: 6 months H: 6 months	Community dwelling women aged 70 to 80 years; MMSE score ≥ 24; Subjective memory complaints, scored ≥6/8 on the Lawton and Brody Instrumental Activities of Daily Living scale ▪ 86 randomized participants	Three-trial arms: IG: Resistance training IG: Aerobic training CG: Balance and toning	▪ Executive cognitive function of selective attention and conflict resolution - Stroop test	Questionnaire to assess total healthcare resource utilization completed by the PwMCI	Health care system perspective	At 6 months resistance training and aerobic training yielded in health care cost savings when compared to controls doing balance and toning exercises.
Pitkälä [20] Finland, 2013	RCT CEA	C: 24 months H: 12 months	Community-dwelling persons with a clinical AD diagnosis and their spousal caregivers; 96% were receiving AD medication ▪ 210 randomized dyads ▪ Cost data 24 months: n = 210 ▪ Outcome assessed at 12 months: n = 161	Three-trial arms: IG: 1-year group-based exercise provided during visits to day care centers twice a week (1 h training) IG: 1-year tailored home-based exercise for 1 h twice a week administered by physiotherapists CG: CAU and subjects received oral and written advice on nutrition and exercise methods	▪ Patients physical functioning: FIM ▪ Mobility: SPPB	Data on use and cost of health and social services were retrieved from central registries and medical records of both patients and caregivers. Cost of patients and caregivers were summed.	Health and social care perspective (not explicitly stated)	Deterioration in patient’s physical functioning was delayed with the tailored home-based exercise program. Both intervention groups showed a significant reduction in the number of falls. The home-based exercise was found to be administered without increasing the total health and social service costs. The group exercise may even decrease the use and costs of health and social services.

AD = Alzheimer’s disease, BPSD = Behavioral and psychological symptoms of dementia, CAU = Care as usual, CEA = Cost-effectiveness analysis, CSRI = Client Service Receipt Inventory, CUA = Cost-utility analysis, FIM = Functional Independence Measure, GHQ = General Health Questionnaire, ICD-10 = International statistical classification of diseases and related health problems, 10th revision, ICER = Incremental cost-effectiveness ratio, PwMCI = Person with mild cognitive impairment, MMSE = Mini-Mental State Examination, NPI = Neuropsychiatric Inventory; QALY = Quality-Adjusted Life Years, RCT = Randomized controlled trial, SPPB = Short Physical Performance Battery, ZBI = Zarit Burden Interview