Premium
P3‐346: STUDY RECRUITMENT AND BASELINE CHARACTERISTICS OF A RANDOMIZED CONTROLLED TRIAL: THE FINNISH GERIATRIC INTERVENTION STUDY TO PREVENT COGNITIVE IMPAIRMENT AND DISABILITY (FINGER)
Author(s) -
Ngandu Tiia,
Lehtisalo Jenni,
Solomon Alina,
Ahtiluoto Satu,
Laatikainen Tiina,
Lindström Jaana,
Peltonen Markku,
Antikainen Riitta,
Hänninen Tuomo,
Jula Antti,
Mangialasche Francesca,
Paajanen Teemu,
Pajala Satu,
Rauramaa Rainer,
Strandberg Timo,
Tuomilehto Jaakko,
Soininen Hilkka,
Kivipelto Miia
Publication year - 2014
Publication title -
alzheimer's and dementia
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 6.713
H-Index - 118
eISSN - 1552-5279
pISSN - 1552-5260
DOI - 10.1016/j.jalz.2014.05.1439
Subject(s) - randomized controlled trial , medicine , dementia , observational study , physical therapy , intervention (counseling) , cognition , psychological intervention , population , disease , psychiatry , environmental health
mean age of caregivers (range 48.6-77.9 years), share of female caregivers (0.50-1.00), and share of spousal caregivers (0.00-1.00) in stepwise regression through backward elimination to predict the level of depression in dementia family caregivers. Results: The best-fit linear regression model explained about 17% of the original variation in dementia family caregiver depression (coefficient of determination r 2 1⁄4 0.310 and adjusted r 2 1⁄40.172, ANOVA, p 1⁄4 0.017). This model included 13 PCNM vectors as explanatory variables. Age, female share, and spousal share were excluded. Standardized model residuals were normally distributed (Kolmogorov-Smirnov test, p1⁄4 0.869) and their absolute values were all<2.2 indicating no outliers. Conclusions: Spatial patterns were stronger explanators of dementia family caregiver depression than caregiver’s age, gender, and relationships to care recipient. The model was able to explain high levels of depression detected in Japan, Korea, and Taiwan, but failed to predict variation observed among locations in the USA and Europe. A larger dataset is required to construct a more generalizable model.