<原著>生理的年齢予測の重回帰分析(II)

概要

論文の詳細を見る
In this paper, a predictive equation for chronological age (AGE) from various physical performance test, anthropometric measurements, and internal medical indecies. Fifty-five healthy adult males, aged 32 to 59 years old, volunteered as subjects in this survey. Data for this study were collected during the course of another study of various indices on the degree of health and fitness and life-style. Information collected included the followings: vital capacity (VITAL), grip strength (GRIP), back strength (BACK), standing trunk flexion (FLEX), standing blinded balance (BALANCE), heart rates after stepping test by Margaria's method (HR1, HR2), heart rates at rest. (HR), estimated % fat from skinfolds thickness (FAT), and systolic and diastolic blood pressure (SBP, DBP) at rest. Forward stepwise regression analysis examined the reliability of the physiological, internal medical and anthropometric measurements as the indicator of age. At the each step, the statistics of degree of fitness, that is, Akaike's AIC, Mallows' Cp, Schwarz criterion and R* (adjusted for d.f. R) adding to multiple correlation coefficient R and SEE (standard error of estimates) were calculated. The results obtained were as follows. 1) The specified features were not shown from descriptive statistics of dependent and independent variables. Multi-colinearity was not shown from independent variables correlation coefficient matrix (11×11). 2) In terms of following variables order, all variables were entered into equation; BACK, VITAL, SBP, GRIP, HR, DBP, HR2, HR1, FAT, BALANCE, FLEX. The multiple correlation coefficient R, and standard error of estimates (SEE) were 0.6317, 7.37 (age of years), respectively. 3) The equations obtained at each step, which included mindependent variables at m step, in general, were evaluated by several statistics such as criteria descrived above. According to these criteria, following descrived equation which included BACK, VITAL. SBP and GRIP was selected the best equation. y=53.07-0.219×BACK-.00398×V ITAL+0.1573×SBP+0.253×GRIP. However, this equation was a problem that the regression coefficient of GRIP was positive. Therefore, another forward stepwise regression analysis was conducted after GRIP were deleted. Cosequently, following equation was the better one and its multiple correlation was 0.612. y=57.93-.171×BACK-.00327×VITAL+.1468×SBP 4) When VO2 (estimated maximal aerobic power by Margaria's method) in replace of HR1 and HR2 was added as the 10th independent variable, at the 1st BACK was entered. VO2 at the 2nd step, and VITAL at the 3rd step were respectively entered and the equation was descrived as follows: y=89.0-0.161×BACK-0.042×VO2-0.00278×VITAL (R=. 6408) However, the calculated VO2 was unequal to other variables as for the age changes because this variables were essentially designed to correlate to chronological age. 5) The specific problems of these equations was not obtained from residuals and the co-ordinates of eis (standardized residuals) and predicted age, so the high reliability was verified and the validity of this equation was verified by the results of preceding studies. (Journal of Health Science, Kyushu University, 9: 147-157, 1987)
九州大学の論文
1987-03-28

<原著>生理的年齢予測の重回帰分析(II)

スポンサーリンク

概要

著者

関連論文

スポンサーリンク