Sankhya: The Indian Journal of Statistics

1996, Volume 58, Series B, Pt. 1, pp. 118--144



D. K. JAIN and PADAM SINGH ,Indian Council of Medical Research

SUMMARY.  The incidence of uterine cervix cancer varies to a great extent in the resident female population i seven cities i.e., Bangalore, Bombay, Madras, Delhi, Ahmeddabad, Poona and Nagpur in India. The relationship between incidence and age has been studied for this cancer in which incidence increases prograssively from young adult life into old age. Five year age-specific incidence rates between 30 and 69 years of age is used to estimate the parameters, in each city, in the mathematical model ${I_ x}\eq{bx^ k}$ where ${I_ x}$ is the incidence age at age x, and b (intercept) and k (slope value or power term) are parameters. The logarithms of the incidence rates are regressed on age groups. The slope value (k) lies between 2 or 3 for uterine cervix cancer. The findings suggests that the value of the k appears to be a characteristic of uterian cervix cancer rather than of geographical areas. The value of k might be a biological constant characteristic of the tissue in which the cancer is produced . Polynomial model of degree two significantly improves the fit of the data but this model does not seem to result in a unique set of biologically meaningful parameter estimates.
        Intercities variations in the incidence of the cervical cancer have been observed among women. Age-adjusted, truncated age-adjusted (35-64 years), age adjusted to world population, incidence rate and cumulative rate and risk in India are higher for this cancer compared with experience of most other countries. Average annual age-adjusted (world population) incidence rates in cities  varies between 19.6 and 43.9 per 100,000 women. In the absence of other causes of death, women have an estimates 2.10% to 4.53% risk of developing uterine cervix cancer between the ages of 30 and 69 years. Time trend analyses of 10 year data in Bangalore, Bombay and Madras and 4 years data in Delhi did not reveal statistically significant decrease or increase in the incidence of uterine cervix cancer for most of the age groups.

AMS (1980) subject classification.  Primary 62J05, 93E24; secondary 92C60.

Key words and phrases. Cancer, registry, population, age, incidence, rates, Armitage-Doll model, weighted least squares linear regression and polynomial regression analysis, time trends.

Full paper (PDF)