A matched case control study of risk indicators of breast cancer in assam, India.

Breast cancer is the leading cause of cancer death among women worldwide especially in a developing country like India. It also occupies the highest place with relative proportion 17.5% in the Hospital Based Cancer Registry in progress in the Dr. B. Borooah Cancer Research Institute. Dr. B. Borooah Cancer Research Institute is the Regional Cancer Care Center for entire North East region of India. With this background a matched case control study of 100 cases of breast cancer and 100 controls was carried out to investigate the role of different Socio economic, Female Reproductive and Life style related factors and to understand the etiology of breast cancer in Assam. Controls are matched to the cases by age at diagnosis (±5 years), family income and place of residence with matching ratio 1:1. Data were collected using questionnaire and then conditional logistic regression analysis is used to estimate the odd ratios for several factors. Study revealed that breast cancer occurrence has statistical association with the factors chewing habits (p=0.003), number of children (p=0.080), age at marriage (p=0.014), age at first child birth (p=0.007), age at menarche (p=0.010).

Estimating Risk of Breast Cancer Occurrences at Different Ages: Application of Survival Techniques

Background: Awareness is the primary means to control breast cancer occurrence. The purpose of the present work is to study the risk of breast cancer occurrence in different age group, for the study area, Assam, India, by means of survival analysis techniques. Methods: Survival and hazard functions are key concepts in survival analysis for describing the distribution of event times. In the present research a new individialized model has been proposed for cumulative hazard function, taking gamma probability distribution as probability distribution of breast cancer occurrences. Kaplan Meier Survival method has been applied to find out the probability of diseases occurrence in the early menarche and late menarche group. The data used for implementation were collected from the Record Department of a prime local cancer institute, for the period 2010-2012. The information for the risk factor age at menarche were collected from the patients registered during August 2011 to February 2012. Results: The study reveals that in the study area, cumulative hazard of the women belonging to 35 to 50 years is higher than the early and late aged women. The cumulative hazard plot with shape parameter 0.5, 1 and 10 shows that cumulative risk for early aged women are greater than the late age women but when this values is increased from 10, the opposite trend is observed. Further, the median age of disease occurrence among early menarche group is 52 years and for late menarche it is 54 years. Conclusion: The model developed could successfully point out the age group for women lying at higher risk of breast cancer occurrence. Additionally the important risk factor, age at menarche, was effectively applied to supplement to this calculation. It is hoped that practical use of this method would enhance not only awareness but also early detection of the said disease.

Multiregional two-sex population model.

"The theory of stable population has been developed quite thoroughly in the demographic literature, but it deals with single-region populations that are assumed to be undisturbed by migration. Rogers...proposed the stable population model for the case of a multiregional population that experiences internal migration for single-sex. In this paper, considering a birth-function dependent on both sexes, the multiregional stable population model is extended for two-sex population."

Two regional (rural-urban) life tables for India.

"This paper proposes to construct, for India, a two-regional life table, considering two regions viz., rural and urban. Such a table provides information on the probabilities of survivorship and migration, as well as life expectancy by age, region and place of birth and in particular information on: (1) The probability that an individual living in a rural area at age x will be living in an urban area at age y. (2) Life expectancy at birth in a rural area, and the number of years on average this individual may live in an urban area. The aim here is to compare and contrast rural to urban and urban to rural migration probabilities by sex and age for the years 1971, 1981 and 1991, and provide the underlying reasons. Region-specific life expectancy is also compared for these years." (SUMMARY IN ITA AND FRE)

Household structure and family pattern of a traditional society.

"In this paper, [the] main aim is to examine the household structure and family units of a scheduled caste population--a traditional society of India--with the help of survey data. Data [are] from a socio-demographic survey [of a] scheduled caste population residing in the rural areas of Assam." Information is provided on types of households and family units; residence pattern and household structure; and household size.

Analytical models for number of births--a review.

PIP: This review describes analytical probability models for number of births in human populations. First the biological variables are defined: effective reproductive period; fecundability; fetal loss; non-susceptible period. The models are classified into 2 groups, depending on whether they are designed to conform to a Markov renewal process or not. Equations in each group are presented with increasing complexity, involving heterogeneity and taking into account dependence of factors on each other, such as fecundability on parity. Recently, statisticians have introduced into their models the use of contraceptive techniques and the possibility of switching between methods.^ieng

An alternative to multiregional two-sex model.

PIP: Factors such as female fertility, male fertility, female survival rate, and male survival rate affect the replacement of the present generation by the next generation. This realization led to the formulation of the two-sex population model to replace the one-sex model based on age composition of females alone. In 1975 Rogers introduced the study of one-sex multiregional stable population. In 1987 Nath extended the study of multiregional stable population to two-sex cases. An alternative approach of incorporating the age structure of both the sexes, considering age difference of parents and making use of sex ratio(s), making certain assumptions on female birth rates of the regions constituting the multiregional stable population, was employed to obtain a few female birth models that become stable. In the 4 models the effect on the female birth trajectory was studies using: 1) constant age difference between the parents when the survivors of male cohort are obtained from male cohort; 2) variable age difference between the parents as in 1) ; ; 3) constant age difference between the parents when survivors of male cohort are obtained from survivors of female cohort; and 4) variable age difference between the parents as in 3). The analyzed 4 models consist of female birth models of a multiregional stable population reflecting the effect of constant/variable age difference of parents when survivors of male cohort are obtained from male/female cohort. Models obtained are neither ordinary one-sex models nor ordinary two-sex models. In all the models paternity linked maternity (PLM) function was obtained in place of simple maternity function of the one-sex model and net parenthood function of the two-sex model. With the help of the PLM function, forecasts on future births can be made. As data for PLM function can be obtained more easily than those for net parenthood function, it is expected that the above models will be more suitable for practical applications than the ordinary two-sex model.^ieng

A note on Brass's model for the distribution of births in human populations.

PIP: Brass's model for the distribution of births considered only births to fertile women, making it inapplicable for countries with low fertility or samples of women who differ markedly in terms of fecundability. The authors show that a simple extension of a modified Poisson distribution, considering females to comprise 2 groups with respect to fecundability (high versus low ability to conceive), describes satisfactorily the data provided by Brass. Brass had noted that the major discrepancies between the observed and expected distribution of births in low-fertility countries reflect the practice of contraception and variations in the rate of childbearing during the reproductive period. The analysis indicated that the proportion of subfecund females is 70-90% in developed countries, reflecting high contraceptive usage. In contrast, in India, the data show that only 14% of fertile females belong to the low-risk group and about 4% remain sterile throughout the reproductive period. The revised model was applied to data from a Rural Development and Population Growth survey conducted in Varanasi, India, in 1978. Only women who were married when under 20 years of age and did not practice contraception were included in the test of the model. The observed (expected) number of females in each number of births category was as follows: 0, 41 (41); 1, 25 (24.3); 2, 50 (51.8); 3, 96 (93.5); 4, 182 (183.8); 5, 261 (272.3); 2, 221 (221.1); 7, 86 (79.6); 8, 13 (9.4); and 9 and over, 2 (0.2).^ieng

Probability distributions of number of births and their applications.

This paper develops probability distributions to describe the variations in the number of live birth-conceptions to a female during a given time interval (0, To) of length To. In the derivation, the interval (0, To) is divided into 2 consecutive segments, and fertility parameters within each segment are assumed to be constant but may differ between the segments. A method of obtaining maximum likelihood estimates of the parameters is outlined. The models are applied to an observed set of data. An application of the models in the evaluation of family planning programs is illustrated.

Probability distributions of number of births and their implications.

PIP: To apply models of the variation in the number if births to a couple throughout the reproductive period, it is necessary to account for variations that are dependent on age and parity. This paper presents time-dependent models for the number of live births/conceptions to a woman during given time intervals and applies these models to data derived from the Rural Development and Population Growth survey conducted in India in 1978. The reproductive span was segmented into 20-30 years and 30-35 years. Within each segment, fecundability was assumed to be constant but to vary between segments. The model assuming variability in fecundability provided a better fit to the data than the model assuming homogeneity. The data analysis suggests that about 4% of women are sterile at ages 20-35 years. The female population appears to consist of 2 groups with respect to fecundability: 1) a high-risk group (82.5%) with an average risk of conception of 0.7555 and 2) a low-risk group (17.5%) with an average conception rate of 0.2611. The average rate of conception for the 2 groups in the 20-39-year age period is 0.6607. These models can be used to demonstrate expected changes in fertility when contraceptives of varying degrees of effectiveness are used. If contraceptive effectiveness is increased 4 times, the expected number of births is decreased by 2.8 times/year.^ieng

Determinants of breast-feeding patterns in an urban society of India.

Breast feeding is the focus of rapidly growing interest in many areas of demographic research. However, relatively few rigorous studies on breast-feeding patterns and correlates in contemporary India have been published. This study uses data from a retrospective survey conducted in 1991-1992 to investigate current breast-feeding patterns and to identify the key factors that influence the duration of exclusive breast feeding and infant's age at the time of weaning in an urban Hindu society of the northeast Indian state of Assam. Applying life table procedures and a hazards regression model, we found evidence that the median duration of exclusive breast feeding and infant's age at the time of weaning were negatively associated with mother's education, per capita income, and social status of the household. Those infants who were breast-fed longer at night than in the daytime were also at greater risk of earlier introduction of non-breast-milk foods and of earlier termination of breast feeding than infants who were breast-fed longer during the day. Gender bias toward males in rearing infants prevails in this urban society, and male infants were found to have a significantly lower risk of early weaning than female infants.

PIP: This study examines breast feeding patterns among Hindu mothers from Guwahati city, Assam state, India. Data were obtained from a 1991-92 household survey and a subsample of 1650 women aged under 50 years who lived in the capital city. 38.0% had less than a primary education (8.9% of wage earners and 29.1% of non-wage earners). 40.6% had some primary education (3.9% of wage earners and 36.7% of non-wage earners). 21.4% graduated from high school (11.7% of wage earners and 9.7% of non-wage earners). 75.5% of the sample were non-wage earners. Findings indicate that among the last two births the median duration of breast feeding was 16 weeks. 75% of infants were introduced to non-breast-milk products by the age of 26 weeks. The median age of weaning was 19 months. 64% of infants were not breast fed at 36 months. As level of mother's education increased, the duration of breast feeding decreased. The longest duration was among mothers with low educational attainment. At 6 months, almost double the infants of highly educated mothers were receiving non-breast-milk foods compared to less educated women. Wage earners did not provide non-breast-milk food earlier than non-wage earners. However, the median age at weaning was 22 months for non-wage earners and 14 months for wage earners. Maternal age was unrelated to median duration of breast feeding. Although older and younger mothers introduced non-breast-milk foods almost at the same time of 4 months, younger mothers breast fed longer. This longer breast feeding pattern suggests its use for birth spacing. Income and caste membership were unrelated to duration of exclusive breast feeding. The hazard analysis indicates that higher educated mothers were at greater risk of early introduction of non-breast-milk foods and an earlier stop to breast feeding. Female children had a higher risk of an earlier stop to breast feeding.

The socio-demographic correlates of post-partum amenorrhoea in an urban society of India.

PIP: Studies in different parts of the world have shown duration of breast feeding to be positively correlated with duration of postpartum amenorrhea. Breast feeding is, however, associated with a variable period of amenorrhea among nursing mothers. This study of breast feeding practices demonstrated that socioeconomic and educational status largely determine the pattern of postpartum amenorrhea and, consequently, fertility during lactation. Undernourished nursing mothers were found to have a longer duration of postpartum amenorrhea relative to better nourished peers. These findings suggest that any improvement in maternal education and nutritional status may lead to a faster return of fertility in lactating women. Efforts to improve the educational, health, and nutritional status of nursing mothers should be integrated with services which provide adequate modern contraceptive care. The authors used survey data collected from the Assamese-speaking Hindus in urban Guwahati during 1991-92.^ieng

A probability model for number of births and its application in estimation of fecundability for a heterogeneous population.

"Under some simple assumptions relating to the human reproduction process, an extension of a probability model describing the variations in the number of births to a female during a specified period of time (O, T) of length T, is proposed. The model is extended to a situation when the female population consists of two distinct groups with respect to fecundability. A procedure of estimation of some of the underlying parameters is given. The applicability of the model is illustrated with an observed distribution." Data for the illustration are from a population survey undertaken in Varanasi, India, in 1978.

Sex preference and third birth intervals in a traditional Indian society.

The traditional preference for sons may be the main hindrance to India's current population policy of two children per family. In this study, the effects of various sociodemographic covariates, particularly sex preference, on the length of the third birth interval are examined for the scheduled caste population in Assam, India. Life table and hazards regression techniques are applied to retrospective sample data. The analysis shows that couples having two surviving sons are less likely to have a third child than those without a surviving son and those with only one surviving son. Age at first marriage, length of preceding birth intervals, age of mother, and household income have strong effects on the length of the third birth interval.

PIP: Hazard regression techniques were used to examine the impact of socioeconomic and behavioral factors on third birth intervals among scheduled caste couples who did not practice family planning to space and limit births in villages in Karimganj district of Southern Assam, India. Life table techniques were used to estimate median birth intervals. Data were obtained from a retrospective study of the effects of socioeconomic factors on fertility among a scheduled caste population in the rural areas of Karimganj District, Assam, during 1988-89. Independent variables are marriage age (over or under 15 years), maternal age at the birth of the third child (under 25, 25-30, and 30-35 years), length of the first and second birth intervals (under 48, 48-68, and over 68 months), mother's education and income per capita (over or under Rs. 60), subcaste, gender and survival status of the first two births. Analysis of potential interaction of covariates revealed no statistically significant relationships. The median third birth interval was just under three years and ranged from 32.3 to 40.9 months. The third birth interval increased with age of marriage, length of the two preceding birth intervals, and family income. A shorter third birth interval followed two surviving female children or only one surviving child of either sex. The generalized Wilcoxon test confirmed statistically significant different survival functions among subgroups. The proportional hazards regression analysis revealed that couples with one male and one female surviving child were only slightly more likely to have a long third birth interval than couples with two surviving male children; differences do not reach statistical significance. A greater likelihood of a short third birth interval occurred among those with only one surviving child of either sex or two female surviving children. The hazard model with all social and biological factors showed statistically insignificant effects and a positive relationship between surviving male and female children. Risk after two surviving males was 26% lower than after two surviving girls. A 29% higher risk of a third birth occurred among females married under 15 years of age. Shorter third birth intervals occurred among those with shorter preceding birth intervals. A lower risk of a third birth occurred among high income groups.

A comparative study of the life tables of Assam, Kerala and India, 1980.

The authors construct "life tables for Assam...using the Sample Registration System data of 1980 [and applying] Greville's method of constructing abridged life tables. These tables are compared with those of Kerala and India. Brass' two parameter logit system [when] fitted to survivors shows that risk of mortality of Kerala females, is much lower than that of India and Assam."

On some bivariate distributions of number of births.

PIP: This paper presents probability models to describe the variation in number of live birth-conceptions to a group of females in two consecutive segments of the reproductive period. In derivation of the models, it is assumed that the parameters considered in the models remain constant in each segment but may differ between the segments. The models utilize the observed distribution of women according to the number of births given in two consecutive segments arranged in a bivariate table, providing more cells with significant number of observations for estimating the fertility parameters. A procedure to obtain the maximum likelihood estimates of some of the parameters is given. The models are applied to data for 977 Indian women surveyed in the 1978 Rural Developments and Population Growth Sample Survey.^ieng

Desired family size--a parity dependent model and its application.

PIP: This paper proposes a model for number of births that considers the following factors: contraceptive practice and its effectiveness, the chance of undergoing induced abortion, the chance of undergoing sterilization following spontaneous induced abortion or live birth, and the length of time involved in deciding to undergo sterilization following a live birth. These variables are assumed to be dependent on parity as well as family size. The model is used to measure the percentage reduction in total fertility when several vertical family planning programs are in operation. Under the various program combinations, the proportions of females according to number of births during the 1st 30 years of marriage, average family size, and percentage reduction in number of births are calculated. The fertility reduction is higher among women with smaller family size desires, regardless of family planning program effort. The data further indicate that a higher reduction in total fertility can be attained through sterilization than through the use of highly effective contraception.^ieng

A waiting time distribution for the first conception and its application to a non-contracepting traditional society.

PIP: This paper provides a description of a probability model for determining the time to first birth for a finite marital or cohabitation duration in a traditional society with early marriage and in an advanced society with delayed childbearing due to contraception. The model generates estimates for fertility parameters, such as the risk of conception, the risk of ovulation or withdrawal from using contraception, and the proportion of adolescent sterile females or couples using contraception. The model assumes homogeneity of risk of conception and a one-to-one correlation between conception and live birth. The model has the potential for further adjustment by incorporating a Pearsonian type III distribution and/or abortion as another possible outcome of a conception that may follow a geometric distribution pattern. The model is applied to data from 3931 rural households in Eastern Uttar Pradesh in India in 1987. None of the couples used contraception before the birth of the first child. The minimum chi square method is used to estimate the expected frequencies of important parameters in the model (the adolescent sterile group, the time spent by a woman in the adolescent sterility state in an exponential distribution, and the time spent in the ovulating state before conception in an exponential distribution). Findings indicate that the proposed model provides a good fit to the distribution of waiting time to first conception for marriage cohorts I and II. Findings indicate that it took 1.65 years and 1.54 years for women in the respective cohorts to reach the state of ovulation after marriage and 2.38 years and 1.72 years to reach first conception after ovulation. A comparison of this study's estimates with Talwar's estimates of nonfecund females shows, respectively, 74% and 78% of females aged 12-14 years being nonfecund in cohort I. 56% of study participants aged 15 in cohort II were estimated to be nonfecund compared to only 40% in Talwar's estimates.^ieng

The role of breast-feeding beyond postpartum amenorrhoea on the return of fertility in India: a life table and hazards model analysis.

This paper investigates the effects of continued breast-feeding after resumption of menses on fertility, using data from two retrospective surveys in India and single decrement life table and multivariate time-dependent hazards analyses. Breast-feeding even after the return of menses is found to be associated with longer birth intervals. The interaction of breast-feeding duration after resumption of menses and postpartum amenorrhoea has a significant effect on the risk of conception after return of menses.

PIP: Researchers used data from two retrospective studies of traditional societies in India to conduct a single decrement life table and multivariate time-dependent hazards analyses to determine the contraceptive role of continued breast feeding after the return of menses. One study was conducted during 1988-1989 in rural areas of Karimganj District in Assam. The other study was conducted during 1987-1988 in rural areas of Eastern Uttar Pradesh. The overall median duration of months of breast feeding after resumption of menses was 16.06 months in Assam compared to 23.74 months in Uttar Pradesh. The overall duration of postpartum amenorrhea was 7.83 months in Assam and 5.95 months in Uttar Pradesh. In Assam, the overall birth interval after the return of menses for nonlactating mothers was around six months shorter than that of lactating mothers. It was 10 months in Uttar Pradesh. Breast feeding reduced the rate of conception after the return of menses by 16% in Assam and by 49% in Uttar Pradesh. The risk of conception after menses was reduced by 7.4% per month in Assam and by 1.8% per month in Uttar Pradesh. Breast feeding duration after return of menses and postpartum amenorrhea interact to increase the interval between births. When women did not breast feed after menstruation began, the risk of conception increased by 4% among Assam women and by 5% among Uttar Pradesh women for every additional month of postpartum amenorrhea. When women still breast fed nine months after the return of menses, the relative risk of conception was essentially zero. Yet, when they still breast fed for 3-9 months after the return of menses, the relative risk increased by 14%. These findings show that breast feeding after return of menses continued to decrease the risk of conception and to increase the interval between births.

Birth spacing, breastfeeding, and early child mortality in a traditional Indian society: a hazards model analysis.

There are few studies of the interrelationships among breastfeeding, child spacing, and child mortality in traditional societies that incorporate extensive controls for social and demographic characteristics of the mother and child. In this paper, we investigate the impact of breastfeeding and the length of the preceding birth interval on early child mortality (defined as a death in the first two years of life) using data from a traditional society of India. Multivariate hazards models are used to analyze the data. Most prior analyses related the impact of breastfeeding duration to the duration of child survivability by taking breastfeeding as a fixed covariate. The present study has a methodological focus in the sense that breastfeeding information from retrospective survey data is treated as a time-dependent covariate both as a status variate as well as a duration--with empirical findings compared across the two specifications. The effects of postpartum amenorrhoea and various other demographic and socioeconomic characteristics of mother and child are also studied. The results suggest that breastfeeding duration has a strong impact in reducing the relative risk of early child mortality; but it does not explain the effect of the length of the preceding birth interval on early child mortality.

PIP: The impact of breastfeeding and the length of the preceding birth interval on early child mortality were studied in India. Multivariate hazard models used the data of a 1988-89 retrospective survey from 39 villages selected randomly in the Karimganj district of Southern Assam representing 1805 scheduled caste households. In eligible couples the females were under 50 years old. The impact of 8 independent variables on child survival up to the age of 24 months was studied: mother's age at the birth of child, length of preceding birth interval, parity of mother, sex of child, mother's education, subcaste, postpartum amenorrhea (PPA), and breastfeeding. A total of 4020 censored and 618 uncensored observations were analyzed. The results from Model I indicated that preceding birth intervals of length or= 27 months, parity of mother, sex of child, mother's education, and subcaste significantly affected early child death. In Model II, as a result of introduction of PPA as a control variable, only the age of mother attained significance. Models III and IV, which included breastfeeding, made the length of the preceding birth interval significant (15-27 months and or= 27 months). Models V and VI expanded on Models III and IV by including PPA and reducing the strength of the breastfeeding variable. Babies born between 15 and 27 months and or= 27 months from the preceding birth were 25% and 72%, respectively, less likely to die in the first 24 months. Children born to a mother younger than 25 years were more likely to die early compared to mothers 25-35 years or older. The risk of death increased by 1.085 times for every increase in parity of children. The relative risk of death of a male child compared to a female child was 0.4. The children of illiterate mothers faced a 1.327 higher risk of death than those born to educated mothers. The duration of breastfeeding reduced child mortality by 6.5% for every additional month.