The increase in the proportion and number of older people in developed countries
has resulted in research investigating risk factors for adverse health outcomes,
including mortality. However, research has been limited by the range of risk
factors included in regression models. This is partly because traditional
statistical methods and software packages allow a restricted number of variables
and combinations of variables. This article describes ongoing research to
overcome these limitations through the CoRGA program, which combines Cox
regression with a genetic algorithm for the variable selection process. CoRGA
was used to try and identify the best combination of risk factors for 4-year
all-cause mortality. The combination of 10 risk factors identified by CoRGA
included both known and new risk factors for mortality in older people. Further
research is seeking to develop the program further and to identify further risk
factors for all-cause mortality in older people.
1
Grundy E
.
Progress report: demography and gerontology: mortality trends among the
oldest old
. Ageing and the Society1997; 17;
713-725
.
2.
2
Bassuk S S
,
Wypij D
,
Berkman L F
.
Cognitive impairment and mortality in the community-dwelling elderly
. American Journal of Epidemiology2000; 151;
676-687
.
3.
3
Fried L P
,
Kronmal R A
,
Newman A B
.
21 of 27 variables predicted mortality in older adults
. Journal of the American Medical Association (JAMA)1998; 279:
585-592
.
4.
4
Oman D
,
Reed D
.
Religion and mortality among community-dwelling elderly
. American Journal of Public Health1998; 88 (10);
1469-1475
.
5.
5
Altman D G
. Practical Statistics for Medical Research.
London: Chapman and Hall/CRC
, 1991.
6.
6
Bath P A
. Data mining in health and medical information. In
Cronin B
ed. Annual Review of Information Science and Technology vol. 38.
In press.
7.
7
Cox D R
.
Regression models and life tables
. Journal of the Royal Statistical Society Series B (Methodological)1972; 34 (2);
187-220
.
8.
8
Collet D
. Modelling Survival Data in Medical Research.
London: Chapman and Hall/CRC
, 1994.
9.
9
Bland M
. An Introduction to Medical Statistics321.
Oxford: Oxford University Press
, 1995.
10.
10
Akaike H
.
A new look at the statistical model identification
. IEEE Transactions on Automatic Control1974; 19;
716-723
.
11.
11
Hosmer D W
,
Lemeshow S
. Applied Survival Analysis: Regression Modeling of Time to Event
Data.
New York: Wiley
, 1999.
12.
12
Goldberg D E
. Genetic Algorithms in Search, Optimization and Machine Learning.
Reading, MA: Addison-Wesley
, 1989.
13.
13
Peña-Rayes C A
,
Sipper M
. Evolutionary Computation in Medicine: An Overview.
Amsterdam: Elsevier
, 1999.
14.
14
Withley D
. A genetic algorithm tutorial. In Computer Science
Technical Report.
Colorado State University
, 1993.
15.
15
Mitchell T M
. Machine Learning.
New York: McGraw-Hill
, 1997.
16.
16
Narayanan M N
,
Lucas S B
.
A genetic algorithm to improve neural network to predict a
patient’s responses to Warfarin
. Methods of Information in Medicine1993; 32;
558
.
17.
17
Jefferson M F
,
Pendleton N
,
Lucas C P
,
Lucas S B
,
Horan M A
.
Evolution of artificial neural network architecture: prediction of
depression after mania
. Methods of Information in Medicine1998; 37;
220-225
.
18.
18
Bath P A
,
Pendleton N
,
Morgan K
,
Clague J E
,
Horan M A
,
Lucas S B
.
New approach to risk determination: development of risk profile for new
falls among community-dwelling older people by use of genetic algorithm
neural network (GANN)
. Journal of Gerontology: Medical Science2000; 55A (1);
17-21
.
19.
19
Wallet B C
,
Marchette D J
,
Solka J L
,
Wegman E J
.
A genetic algorithm for best subset selection in linear regression
. In Proceedings of the 28th Symposium on the Interface,
1996. 1999.
20.
20
Vinterbo S
,
Ohno-Machado L
.
A genetic algorithm to select variables in logistic regression: example in
the domain of myocardial infarction
. Journal of the American Medical Informatics Association1999;
984-988
.
21.
21
Stacey A
,
Kildea D
.
Genetic algorithm search for large logistic regression models with
significant variables
. In Proceedings of the 22nd International Conference on
Information Technology Interfaces, ITI, Pula,
Croatia, 2000.
22.
22
Colley D A
. An Introduction to Genetic Algorithms for Scientists and
Engineers.
London: World Scientific Publishing
, 1999.
23.
23
Morgan K
.
The Nottingham Longitudinal Study of Activity and Ageing: a methodological overview
. Age and Ageing1998; 27;
511
.
24.
24 Ahmad R. PhD thesis, in preparation.
25.
25
Dyer A R
,
Stamler J
,
Greenland P
.
Association of weight change and weight variability with cardiovascular and
all cause mortality
. Chicago Western Electric Company Study2000; 177 (4);
324-333
.
26.
26
Korten A E
,
Jorm A F
,
Leteneur L
,
Jacomb P A
,
Henderson A S
,
Christensen R B
.
Health cognitive and psychological factors as predictors of mortality in an
elderly community sample
. Journal of Epidemiology and Community Health1999; 53;
838
.
27.
27
Ho S C
,
Woo J
,
Sham A
.
Risk factor change in older persons. A perspective from Hong Kong: weight
change and mortality
. American Journal of Epidemiology1994; 150;
978-986
.
28.
28
Idler E L
,
Benyamini Y
.
Self-rated health and mortality: a review of twenty-seven community studies
. Journal of Health Social Behaviour1997; 3;
21-37
.
29.
29
Benyamini Y
,
Idler E L
.
Community studies reporting association between self-rated health and mortality
. Research on Ageing1999; 21 (3);
392-401
.
30.
30
Bath P A
.
Social isolation and mortality: gender differences in loneliness and living alone
. Gerontologist2000; 40 (special issue 1);
578
.
31.
31
Carmelli D
,
Swan G E
.
The relationship of type behaviour and its components to all-cause
mortality in an elderly subgroup of men from the Western Collaborative Group Study
. Journal of Psychosomatic Research1995; 40 (12);
475-483
.
32.
32
Xiang A
,
Lapuerta P
,
Ryutov A
,
Buckley J
,
Azen S
.
Comparison of the performance of neural network methods and Cox regression
for censored data
. Computational Statistics and Data Analysis2000; 34;
243-257
.
