Age-dependent logistic regression model and its application.

Y Y Liu

Home/Archive/PubMed

PubMed · 2059320

Age-dependent logistic regression model and its application.

Y Y Liu

View original source ↗Export metadata ↓

Abstract

In the present paper we introduce the theory and algorithm of the unconditional and conditional age-dependent logistic regression model, which combines logistic regression analysis of case-control study with survival analysis of cases in the data, thus facilitating simultaneous comparison analysis between cases and controls and among cases with different ages of disease onset under study. In age-dependent logistic regression analysis, estimated compound relative risk (CRR) and compound attributable risk (CAR) comprise the variance contributions of risk factors to disease occurrence and the time of disease onset, thereby the role played by various risk factors in etiology and etiopathology can be objectively evaluated. The current logistic regression model is only a particular case of age-dependent logistic regression theory neglecting the variations in onset age of diseases.

Explore related subjects

Age Factors Aged Esophageal Neoplasms Humans Logistic Models Lung Neoplasms Male Middle Aged Multivariate Analysis Risk Factors Smoking Comparative Study Journal Article

Keep this discovery

Explore connections, maps & timelines

BibTeX RIS

Y Y Liu. 1991. Age-dependent logistic regression model and its application.. https://pubmed.ncbi.nlm.nih.gov/2059320/

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related citations

PubMed1992

[Advantages and limitations of single-channel urodynamics in childhood].

The simultaneous recording of urodynamic parameters and x-ray imaging of the lower urinary tract is the ideal technique for a complex investigation of the lower urinary tract morphology and function. The urodynamic and micturating cystourethrography findings were reviewed in 57 children. The causes of the examinations were recurrent urinary tract infection, voiding dysfunction and urinary tract dilatation by ultrasound. 53/57 patients had a clear cystomanometry results. 4 did not cooperate. 15/53 had unstable bladder. 34/57 had vesico-ureteric reflux. Unilateral reflux was more frequently noted in children with an unstable bladder. In cases of a stable bladder, bilateral reflux occurred more frequently. The ureteric reflux in children with unstable bladder should resolve with therapy to decrease bladder activity. Urodynamics is complementary rather than competitive with radiologic investigations. It is necessary to make micturating cystourethrography with cystomanometry.

Age Factors↗

PubMed1992

Sex differences in osteoporosis in older adults with non-insulin-dependent diabetes mellitus.

OBJECTIVE: To describe the association of non-insulin-dependent diabetes mellitus (NIDDM) with bone mineral density (BMD). DESIGN: A survey of men and women from an established epidemiologic cohort who were separately screened for diabetes by oral glucose tolerance test between 1984 and 1987 and for osteopenia by BMD measured in 1988-1989. SETTING: A community-based population of older adults, Rancho Bernardo, Calif. PARTICIPANTS: The first 627 consecutively seen white men and women aged 55 to 88 years. MAIN OUTCOME MEASURES: Bone density measured by single photon absorptiometry at the ultradistal wrist and midradius and by dual x-ray absorptiometry at the femoral neck and lumbar spine. MAIN RESULTS: Among the 236 men and 391 women, whose average age was 72 years, 41 men and 39 women had NIDDM, 56 men and 110 women had impaired glucose tolerance, and 139 men and 242 women had normal glucose tolerance. Men with diabetes had BMD levels similar to those men with normal glucose tolerance, whereas women with diabetes had significantly higher BMD levels at all sites than women with normal glucose tolerance. The increased bone density in diabetic women was unexplained by age, obesity, cigarette smoking, alcohol intake, regular physical activity, and the use of diuretics and estrogen. The multiply adjusted mean BMD in women with NIDDM compared with normoglycemic women was 0.600 g/cm2 vs 0.548 g/cm2 at the midradius; 0.265 g/cm2 vs 0.230 g/cm2 at the ultradistal wrist; 0.654 g/cm2 vs 0.610 g/cm2 at the femoral neck; and 0.962 g/cm2 vs 0.859 g/cm2 at the spine. The sex differences were unexplained by survivor bias, prior obesity, or duration of diabetes. Differences were seen in women (but not men) whose diabetes was first detected at the screening evaluation, ie, before drug or dietary treatment. Similarly, in women (but not men) without diabetes increasing BMD levels at all four sites were associated with increasing postchallenge glucose levels independent of age and body mass index. CONCLUSIONS: Older women with NIDDM or hyperglycemia had better BMD than women with normal glucose tolerance, independent of differences in obesity and many other risk factors. No differences in bone density by diabetic status were observed in men. We hypothesize that the sex differences may be explained by the greater androgenicity reported in women with hyperglycemic and hyperinsulinemic conditions.

Age Factors↗

PubMed1992

Is NIDDM a risk factor for noise-induced hearing loss in an occupationally noise exposed cohort?

Little is known about what factors, other than chronic exposure to noise, predispose individuals to noise-induced hearing loss (NIHL). The current retrospective study was designed to identify risk factors for NIHL in a population of 229 men [age 55-68 (mean = 63 years)] employed at a metal assembly plant. All men had been chronically occupationally noise-exposed for approximately 30 years (> or = 89 dBA) with an average Ea noise emission level) of 104.5. The clinical examination included a pure-tone threshold audiometric evaluation, discrimination of speech in background noise [W-22 Max (> 60% indicating better hearing)], blood pressure measurement, evaluation of lifestyle (alcohol consumption, cigarette smoking, noisy hobbies) and occupational and military history. Severe NIHL was defined as > or = 65 db loss at 3, 4 or 6 kHz in at least one ear +/- 20 db threshold in the contralateral ear. History of non-insulin dependent diabetes mellitus (NIDDM) was reported by 16.4% of the 146 men with severe NIHL compared to 4.8% of the 83 men without severe NIHL (odds ratio = 3.9, C.I. 1.2-11.9, P = 0.05). Simultaneous evaluation of several potential risk factors using a multiple logistic regression indicates that the significant predictors of severe NIHL were diabetes (P < 0.05), Ea (P < 0.05) and age (P < 0.05). These results suggest that a person with NIDDM who is also occupationally noise-exposed is more likely to develop severe NIHL than those without NIDDM. Longitudinal studies are necessary to confirm the temporal relationship between NIDDM and NIHL and to determine the exact mechanisms that are involved with this increased risk of hearing loss.

Age Factors↗