Risk of breast cancer in women with benign breast disease.

A group of 1,489 white women were treated in a private surgery practice from 1940 through 1975 for biopsy-proved benign breast disease, and 1,441 were followed through 1976 for the development of breast cancer. Average duration of follow-up was 12.9 years for a total of 18,617 person-years of observation. Information was collected from a set of questions devised in 1941 and asked of all subjects at the time of their initial office visit, follow-up interview conducted in 1976, and a standardized histology review of the slides from the initial benign lesions and the subsequent cancers. The current pathology review indicated that 66 of the women developed breast cancer. The incidence rate was 3.55 per 1,000 person-years, which is 2.10 times that of the general population. When multiple disease types and other variables were controlled for, excess risk of breast cancer was related to the presence of fibrocystic disease. In women with fibrocystic disease, excess risk was particularly related to the presence of epithelial hyperplasia and/or papias not related to the presence of fibroadenoma alone, but it was related to the presence of fibroadenoma in women with concomitant fibrocystic disease. The excess risk was also directly related to the estimated size of the initial benign mass and was greater for women with bilateral than with unilateral benign lesions.

Hutchinson Smoking Prevention Project: long-term randomized trial in school-based tobacco use prevention--results on smoking.

BACKGROUND: No long-term impact has yet been observed with the use of the social-influences approach to school-based smoking prevention for youth. However, whether this lack of impact is due to methodologic problems with the studies or to the failure of the interventions is unclear. The Hutchinson Smoking Prevention Project (HSPP), conducted from September 1984 through August 1999, aimed to attain the most rigorous randomized trial possible to determine the long-term impact of a theory-based, social-influences, grade 3-12 intervention on smoking prevalence among youth. METHODS: Forty Washington school districts were randomly assigned to the intervention or to the control condition. Study participants were children enrolled in two consecutive 3rd grades in the 40 districts (n = 8388); they were followed to 2 years after high school. The trial achieved high implementation fidelity and 94% follow-up. Data were analyzed with the use of group-permutation methods, and all statistical tests were two-sided. RESULTS: No significant difference in prevalence of daily smoking was found between students in the control and experimental districts, either at grade 12 (difference [Delta] = 0.2%, 95% confidence interval [CI] = -4.6% to 4.4%, and P =.91 for girls; Delta = 0.3%, 95% CI = -5.0% to 5.5%, and P =.89 for boys) or at 2 years after high school (Delta = -1.4%, 95% CI = -5.0% to 1.6%, and P =.38 for girls; Delta = 2.6%, 95% CI = -2.5% to 7.7%, and P =.30 for boys). Moreover, no intervention impact was observed for other smoking outcomes, such as extent of current smoking or cumulative amount smoked, or in subgroups that differ in a priori specified variables, such as family risk for smoking. CONCLUSION: The rigor of the HSPP trial suggests high credence for the intervention impact results. Consistent with previous trials, there is no evidence from this trial that a school-based social-influences approach is effective in the long-term deterrence of smoking among youth.

Monoclonal antibody therapy of spontaneous AKR T-cell leukemia.

We have previously shown that monoclonal antibodies against the Thy 1.1 differentiation antigen can inhibit the outgrowth of a lethal inoculum of transplanted AKR T-leukemic cells. In the present report we have extended these studies to examine antibody therapy of aged AKR/J mice with spontaneous leukemia. Infusion of anti-Thy 1.1 antibody in frankly leukemic mice led to uniform early mortality from cell lysis and agglutination. In contrast, anti-Thy 1.1 antibody therapy of mice in remission following treatment with cyclophosphamide prolonged remission duration (P less than 0.001) and modestly prolonged survival (P less than 0.01) compared to treatment with irrelevant antibody or chemotherapy alone. The major cause of failure was relapse of leukemia. In 85% (47 of 55) of cases relapse was due to cells that continued to express Thy 1.1, but in 15% of these relapsing animals all leukemic cells failed to express the target antigen. Our results suggest that monoclonal antibody against a normal T-cell antigen can add to the antileukemic effects obtained with chemotherapy alone. Nevertheless, the clinical benefit of unmodified antibody was modest, and antibodies conjugated to cytotoxic agents may be needed to overcome the limitations of unmodified antibodies.

Experimental radiotherapy of murine lymphoma with 131I-labeled anti-Thy 1.1 monoclonal antibody.

Monoclonal antibodies against the Thy 1.1 differentiation antigen are ineffective in the treatment of transplanted AKR T-cell lymphoma once a palpable tumor nodule is present, due to the inability of the host to eliminate antibody-coated tumor cells. To overcome this limitation, we have evaluated the use of 131I-labeled anti-Thy 1.1 antibodies for the therapy of established AKR/J SL2 lymphoma (Thy 1.1+) nodules growing in congeneic AKR/Cu mice (Thy 1.2+). In these experiments, 131I-anti-Thy 1.1 antibody specifically localized to a s.c. tumor with a mean of 6.5% of the infused dose per g of tumor at 24 h after infusion. The proportion of infused anti-Thy 1.1 antibody localizing to tumor was constant following antibody doses of up to 400 micrograms/animal. Antibody iodinated with up to 2 atoms of iodine per antibody of molecule maintained binding activity and localization to tumor equivalent to antibody labeled with less iodine. The concentrations of 131I-anti-Thy 1.1 in tumor would result in delivery of a mean of 1600 cGy to tumor following infusion of 500 muCi of 131I-labeled anti-Thy 1.1 antibody. In comparison, 500 muCi 131I-labeled irrelevant antibody would deliver a mean of 380 cGy to tumor. Treatment of animals with palpable tumor nodules with 500 muCi 131I-anti-Thy 1.1 led to regression of the tumor nodule in 44% of animals, significantly prolonged survival, and cured two of five of the animals treated prior to the development of metastatic disease. In contrast, unlabeled anti-Thy 1.1 led to tumor response in 6% of animals, and up to 1000 muCi 131I-labeled irrelevant antibody had no effect on tumor growth. Therapy was limited by the emergence of variant tumor cells lacking the target antigen and by bone marrow toxicity following 131I-labeled antibody doses of greater than or equal to 1000 muCi/animal. These studies demonstrate that 131I-labeled monoclonal antibodies can have a significant antitumor effect in a situation where unmodified antibody is ineffective.

Investigation of circular asymmetry in cancer mortality of Hiroshima and Nagasaki A-bomb survivors.

Data on Hiroshima and Nagasaki A-bomb survivors are used to investigate, for each city, possible circular asymmetry in cancer mortality around the hypocenter. Using the Cox regression method and controlling for age ATB, sex, followup year, distance from the hypocenter, and type of shielding, it is found that in Hiroshima cancer mortality was significantly higher in the westerly direction from the hypocenter. Mortality from stomach cancer, leukemia, and colon cancer were higher in the westerly direction. In Nagasaki, only lung cancer exhibited circular asymmetry, and was significantly higher in the westerly direction. For various reasons, the results tend to support the possibility of an asymmetry in radiation dose in Hiroshima, but not in Nagasaki. Also, possible asymmetry in nondose variables associated with cancer is suggested in both cities, particularly in variables associated with lung cancer. Indications for future work and implications for future dose-mortality investigations are discussed.

Children's perceptions of smokers and nonsmokers: a longitudinal study.

In a longitudinal study we investigated 5th- and 7th-grade children's perceptions of smokers and nonsmokers, changes in perceptions from 5th to 7th grades, and the degree to which these perceptions predict smoking in 9th grade. The participants were 1,663 students from 14 school districts in Washington state. The results showed large developmental shifts from 5th to 7th grade in children's perceptions of both smokers and nonsmokers: Students at 7th grade saw smokers in a much more positive light and nonsmokers in a much more negative light than they did at 5th grade. Children's positive perceptions at 5th grade of smokers predicted smoking 4 years later at 9th grade and were stronger predictors than positive perceptions at 7th grade. The results suggest that smoking prevention interventions must begin before 5th grade to counter perceptions predictive of subsequent smoking.

Teacher training as a behavior change process: principles and results from a longitudinal study.

For students to realize the benefits of behavior change curricula for disease prevention, programs must be implemented effectively. However, implementation failure is a common problem documented in the literature. In this article, teacher training is conceptualized as a behavior change process with explicit teacher motivation components included to help effect the intended behavior (i.e., implementation). Using this method, the Hutchinson Smoking Prevention Project, a randomized controlled trial in school-based smoking prevention, conducted 65 in-service programs, training nearly 500 teachers (Grades 3-10) from 72 schools. Implementation was monitored by teacher self-report and classroom observations by project staff. The results were favorable. All eligible teachers received training, virtually all trained teachers implemented the research curriculum, and 89% of observed lessons worked as intended. It is concluded that teacher training conceptualized as a behavior change process and including explicit teacher motivation components can promote effective implementation of behavior change curricula in public school classrooms.

Descriptions of early triers.

The onset of smoking behavior in adolescents has been described as a process, beginning when children are young. Little empirical evidence is available, however, on the nature and specifics of the onset process in young children. More information is needed about the early stages of smoking onset in order to design interventions for young children and for early triers. The purpose of the present study was to describe several onset-related variables in young girls and boys and to discuss implications for designing prevention interventions that target young children. A total of 1,663 5th-grade students completed a questionnaire assessing smoking behavior, psychosocial characteristics, and perceptions of a "smoker" image. Saliva samples for cotinine analysis were also collected. Students were classified as either never-triers (never tried a cigarette) or early triers (tried one or more cigarettes) on the basis of self-reported smoking. Most students who had tried a cigarette were in the early stages of smoking onset, because approximately 30% had tried one cigarette and less than 10% had tried a second. Triers versus never-triers differed on their reported images of smokers, and several psychosocial characteristics predicted trying a cigarette and intentions to smoke for boys.

Adolescents' reasons for smoking.

Part of the Hutchinson Smoking Prevention Project, this study provides information on reasons reported by adolescents for their smoking behavior. A total of 1,615 10th grade students in 20 rural and suburban school districts in Washington state were asked why they currently smoke and why they first smoked. For beginning smoking, curiosity, social norms, and social pressure were the most frequently given reasons for smoking. For current smoking, pleasure and addiction were mentioned most often. A number of gender differences occurred with 10th grade students beginning and current smoking. More females than males reported social norms and social pressure as reasons for beginning to smoke, whereas for currently smoking, pleasure was given as a reason more frequently by females than males. When developing interventions, smoking prevention programs need to consider students' current and past smoking behavior, and the reasons they give for smoking.

Bounds for a joint distribution function with fixed sub-distribution functions: Application to competing risks.

This paper gives sharp bounds for the joint survival function G(t(1), t(2),...,t(r)) identical with P(X(1) > t(1), X(2) > t(2),...,X(r) > t(r)), and for the marginal survival functions S(j)(t) identical with P(X(j) > t), j = 1,2,...,r, when the sub-survival functions S(j) (*)(t) identical with P(X(j) > t, X(j) = min(k=1,2),...,(r)X(k)) are fixed. Theorem 1 gives the bounds for r = 2, and Theorem 2 gives the bounds for general r. Theorem 3 applies the result to the competing risks problem, and presents empirical bounds based on the observations. Finally, an example illustrates the bounds.

Teaching the principles of clinical trials design and management.

This paper presents the aims, conduct and references that we feel would be appropriate for a formal interdisciplinary graduate course in the design and management of clinical trails. Recent papers in the medical and statistical literature, accompanied by in-class discussion, constitute a good "text". Opportunity to gain "experience" can be provided by assignments to review and critique clinical trials reported in the literature, and especially by an assignment to design "from scratch" a clinical trial. Practice in interdisciplinary discussions and decision-making is afforded when biostatistics students and other research-oriented students take the course together.

Statistical methods for the analysis of longitudinal data from school-based smoking prevention studies.

The features which make longitudinal data obtained from school-based smoking prevention studies well-suited for efficient analysis by survival analysis methods are discussed. Survival analysis methods, in particular relative risk regression models, are described and illustrated through an example involving data from the Waterloo Smoking Prevention Project--Study 1. Indications of some of the possible applications for these techniques in the evaluation of interventions to prevent smoking and the study of the smoking onset process are provided.

Smoking cessation among high school seniors.

BACKGROUND: Effecting smoking cessation among adolescents provides an opportunity to reduce smoking levels among teens and among the adults those teens are soon to become. Better understanding of high school students' cessation behavior is needed to inform the development of effective strategies to support teen quitting. METHODS: Data from 1,210 established smokers among 6,427 12th graders from rural and suburban communities in Washington State were used to determine the frequency of a declared intention to quit, one or more quit attempts in the past year, ongoing quitting, and the attainment of long-term cessation. Logistic regression models investigated the association of these outcomes with the extent of peer smoking, age of becoming a regular smoker, sex, and grade point average (GPA). RESULTS: A serious intention to quit was declared by 67% of established smokers, and at least one attempt to quit was made in the past 12 months by 60%. However, only 21% of those who had attempted to quit in the past year were still abstaining at the time of the survey, and overall, merely 3% had achieved cessation beyond 12 months. The likelihood of success of quit attempts was strongly dependent on the extent of smoking among peers. Subjects who had become regular smokers at the youngest ages were more likely to undertake a quit attempt than those who started older. Females were less successful than males in attaining ongoing abstinence. Cessation behavior showed no clear association with students' GPA. CONCLUSIONS: A high fraction of high school smokers want to quit, but are unsuccessful in doing so. Thus, there is a need for programs to provide cessation skills to young smokers already motivated to quit.

A comparison of generalized linear mixed model procedures with estimating equations for variance and covariance parameter estimation in longitudinal studies and group randomized trials.

Response data in longitudinal studies and group randomized trials are gathered on units that belong to clusters, within which data are usually positively correlated. Therefore, estimates and confidence intervals for intraclass correlation or variance components are helpful when designing a longitudinal study or group randomized trial. Data simulated from both study designs are used to investigate the estimation of variance and covariance parameters from the following procedures: for continuous outcomes, restricted maximum likelihood (REML) and estimating equations (EE); for binary outcomes, restricted pseudo-likelihood (REPL) and estimating equations (EE). We evaluate these procedures to see which provide valid and precise estimates as well as correct standard errors for the intraclass correlation coefficient or variance components. REML seems the better choice for estimating terms related to correlation for models with normal outcomes, especially in group randomized trial situations. Results for REML and EE are mixed when outcomes are continuous and non-normal. With binary outcomes neither REPL nor EE provides satisfactory estimation or inference in longitudinal study situations, while REPL is preferable for group randomized trials.

Initiation and use of smokeless tobacco in relation to smoking.

Questionnaire data obtained from 1,631 tenth grade students in 14 school districts in the State of Washington are used in this investigation of the relationship between the onset processes for smokeless tobacco use and smoking. Emphasized is the use of time-to-event data on the ages of occurrence of six events in these onset processes. Concepts and methods for the statistical analysis of time-to-event data are demonstrated. The occurrence of events in the smoking onset process are strongly related to increases in the subsequent onset rate for smokeless tobacco use. Compared with before initial smoking has occurred, the onset rates for weekly smokeless tobacco use after initial smoking has occurred are 2.03 (P less than .001) and 6.72 (P less than .001) times as large for males and females, respectively. Furthermore, both initial and weekly use of cigarettes contributes to the risk of subsequent weekly smokeless tobacco use. Conversely, the steps in the onset process of smokeless tobacco use are strongly related to increases in the subsequent smoking onset rate. Possible implications for intervention in prevention of smokeless tobacco use and for further research are discussed.

The Hutchinson Smoking Prevention Project trial: design and baseline characteristics.

INTRODUCTION: This paper describes the experimental design and baseline characteristics of the Hutchinson Smoking Prevention Project (HSPP), a 15-year trial to determine to what extent a grade 3-12 school-based tobacco use prevention intervention can deter tobacco use throughout and beyond high school. DESIGN: Trial design features include use of the school district as the unit of randomization, inclusion of the school district's entire enrollment of 3rd graders, long-term follow-up of the entire original cohort, and sample size and evaluation methods that account for the group-randomization and intraclass correlation of endpoints within school districts. The theory-based intervention is teacher-led and includes grade 3-10 curriculum units, teacher training, grade 9-12 tobacco use cessation materials, and high school staff newsletters. RESULTS: Baseline data were collected on the trial cohort of 8388 children and their parents and on the 40 collaborating school districts and communities. A comparison of the distribution of baseline variables between experimental conditions shows good balance. CONCLUSIONS. The HSPP trial's experimental design will provide a rigorous test of the intervention. The balance in baseline variables between the experimental and control conditions will help provide assurance that the trial's intervention effectiveness results, scheduled for publication in 2000, will be unbiased.

Predicting adolescent smoking: a prospective study of personality variables.

OBJECTIVE: We investigated how personality variables measured at 5th grade predict daily smoking in 12th grade. METHODS: A group of 3130 fifth graders was assessed for a propensity toward Rebelliousness, Risk Taking, Problem Helplessness, Affect Regulation, and Early Maturation and susceptibility to Peer Compliance and Peer Approval. Daily smoking status was determined for 94.7% of them 7 years later, in 12th grade. RESULTS: Rebelliousness and Risk Taking were the most significant predictors of smoking. There was no statistical evidence that the extent of prediction depended on gender or history of early smoking. No predictive evidence was seen for the other personality measures, including those describing susceptibility to peer influences. CONCLUSIONS: These results show that propensity toward rebelliousness and risk taking in childhood predicts adolescent smoking. They suggest that smoking prevention programs would do well to address the needs and expectations of rebellious and risk-taking youth.

Experimental design and methods for school-based randomized trials. Experience from the Hutchinson Smoking Prevention Project (HSPP).

Nonadherence to accepted design principles for randomized trials has been a limitation of school-based intervention research. Designed to overcome these limitations, the Hutchinson Smoking Prevention Project (HSPP) is a 15-year randomized trial to determine the extent to which a school-based (grades 3-12) tobacco use prevention intervention can deter youth tobacco use throughout and beyond high school. This paper presents the HSPP experimental design, together with methods for its implementation, and an evaluation of the degree to which HSPP has adhered to principles of randomized trials. Results from the experimental design and its conduct include (1) a recruitment rate of 97.6% (40 of 41 targeted school districts), (2) full and active participation for the trial's duration by 100% of the 40 school districts recruited, (3) implementation by virtually all teachers (99%+), with 86% implementation fidelity, and (4) outcome determination for 94.3% (7910) of 8388 original study participants identified 12 years previously at baseline. The high degree of rigor achieved by the HSPP experimental design ensures confidence in the trial's soon-to-be available intervention effectiveness results. Equally important, for future school-based trials, the HSPP design and its execution have illustrated that school-based research can adhere to the principles of rigorous randomized trials, with high rates of implementation, and very high rates of recruitment, maintenance, and follow-up of study participants, even for studies with decade-long follow-up periods. Rigor in school-based trials can be achieved through a combination of (1) commitment to the principles of randomized trials, (2) attention to the special challenges of trials specific to the school setting, (3) adoption and meticulous execution of proven methods for trial conduct, and (4) establishment at the outset of principles for maintaining positive collaborative relationships with participating school districts for the duration of the trial. These findings are important in light of the great potential for using the nation's schools to access youth for health promotion/risk-factor prevention.