Laura Ichikawa

Analysis of Case-Cohort Designs

The case-cohort design is most useful in analyzing time to failure in a large cohort in which failure is rare. Covariate information is collected from all failures and a representative sample of censored observations. Sampling is done without respect to time or disease status, and, therefore, the design is more flexible than a nested case-control design. Despite the efficiency of the methods, case-cohort designs are not often used because of perceived analytic complexity. In this article, we illustrate computation of a simple variance estimator and discuss model fitting techniques in SAS. Three different weighting methods are considered. Model fitting is demonstrated in an occupational exposure study of nickel refinery workers. The design is compared to a nested case-control design with respect to analysis and efficiency in a small simulation. In this example, case-cohort sampling from the full cohort was more efficient than using a comparable nested case-control design.

A Randomized Trial Comparing Acupuncture, Simulated Acupuncture, and Usual Care for Chronic Low Back Pain

BACKGROUND Acupuncture is a popular complementary and alternative treatment for chronic back pain. Recent European trials suggest similar short-term benefits from real and sham acupuncture needling. This trial addresses the importance of needle placement and skin penetration in eliciting acupuncture effects for patients with chronic low back pain. METHODS A total of 638 adults with chronic mechanical low back pain were randomized to individualized acupuncture, standardized acupuncture, simulated acupuncture, or usual care. Ten treatments were provided over 7 weeks by experienced acupuncturists. The primary outcomes were back-related dysfunction (Roland-Morris Disability Questionnaire score; range, 0-23) and symptom bothersomeness (0-10 scale). Outcomes were assessed at baseline and after 8, 26, and 52 weeks. RESULTS At 8 weeks, mean dysfunction scores for the individualized, standardized, and simulated acupuncture groups improved by 4.4, 4.5, and 4.4 points, respectively, compared with 2.1 points for those receiving usual care (P < .001). Participants receiving real or simulated acupuncture were more likely than those receiving usual care to experience clinically meaningful improvements on the dysfunction scale (60% vs 39%; P < .001). Symptoms improved by 1.6 to 1.9 points in the treatment groups compared with 0.7 points in the usual care group (P < .001). After 1 year, participants in the treatment groups were more likely than those receiving usual care to experience clinically meaningful improvements in dysfunction (59% to 65% vs 50%, respectively; P = .02) but not in symptoms (P > .05). CONCLUSIONS Although acupuncture was found effective for chronic low back pain, tailoring needling sites to each patient and penetration of the skin appear to be unimportant in eliciting therapeutic benefits. These findings raise questions about acupunctures purported mechanisms of action. It remains unclear whether acupuncture or our simulated method of acupuncture provide physiologically important stimulation or represent placebo or nonspecific effects.

Using Clinical Factors and Mammographic Breast Density to Estimate Breast Cancer Risk: Development and Validation of a New Predictive Model

Context Existing breast cancer prediction tools do not account for breast density, a strong risk factor for breast cancer and have been studied in white women only. Contribution The authors developed a breast cancer risk prediction model that incorporates a measure of breast density routinely reported with mammography. Its predictions were accurate, but it had only modest ability to distinguish women who did not develop cancer from those who did, and it misclassified risk in some subgroups. Implication The model requires validation in additional populations. A breast cancer prediction model that incorporates breast density does well in some but not all domains of predicting risk. Its accuracy should be better characterized before it is used clinically. The Editors In 2007, breast cancer will have been diagnosed in more than 178000 women in the United States, and more than 40000 women will have died of breast cancer (1). Most of these women never had their risk for breast cancer assessed, and even fewer considered chemoprevention (25). Providing women with an estimate of their risk for breast cancer would provide an opportunity for them to consider options to decrease their risk. Women at low short-term risk for breast cancer may experience less anxiety about their health and would be less likely to benefit from prevention efforts. Women at very high risk may warrant additional screening tests, such as breast magnetic resonance imaging (6), and might benefit from chemoprevention of breast cancer with tamoxifen or raloxifene. The standard risk assessment model available to practitioners (the Gail model) (7) identifies only a minority of women who eventually develop breast cancer being at high risk (8). Better breast cancer risk prediction tools are needed (9). The radiographic appearance of the breast has been consistently shown to be a major risk factor for breast cancer, whether it is defined by a qualitative assessment of the parenchymal pattern or a quantitative measure of percentage of density (1012). Women in whom more than 50% of total breast area is mammographically dense have high breast density and are at 3- to 5-fold greater risk for breast cancer than women in whom breast density is less than 25% (10, 1316). The increased risk for breast cancer associated with breast density is due in part to the lower sensitivity of mammography in dense breasts (1719), but the association remains strong after accounting for masking (20, 21). Mammographically dense breast tissue is rich in epithelium and stroma (10), and the association could represent activation of epithelial cells or fibroblasts (2225). Recently, several models have been published that incorporate breast density: One uses a continuous measure of breast density that is not available to clinicians and has not been validated (26), and the other predicts 1-year risk for breast cancer (27). We previously demonstrated that a simple model based on age, ethnicity, and a categorical measure of breast density had predictive accuracy similar to that of the Gail model in a multiethnic cohort of women receiving screening mammograms in northern California (28). We expand on that work by using data from more than 1 million ethnically diverse women throughout the United States to develop and validate a risk assessment tool that incorporates breast density and therefore might improve breast cancer screening and prevention efforts. Methods Study Population We included 1095484 women age 35 years or older who had had at least 1 mammogram with breast density measured by using the Breast Imaging Reporting and Data System (BI-RADS) classification system in any of the 7 mammography registries participating in the National Cancer Institutefunded Breast Cancer Surveillance Consortium (BCSC) (available at breastscreening.cancer.gov) (29). The BCSC is a community-based, ethnically and geographically diverse sample that broadly represents the United States (30). We excluded women who had a diagnosis of breast cancer before their first eligible mammography examination. Because our goal was to develop a model of long-term risk for invasive breast cancer, we excluded women with cancer diagnosed in the first 6 months of follow-up to minimize the number of cases of cancer included in the model that were diagnosed on the basis of the mammogram used for risk assessment. Women were also excluded if they had breast implants. Women in whom ductal carcinoma in situ was diagnosed were censored at the time of diagnosis in the primary analysis. When women had several mammograms, we based our analysis on findings from the first mammogram. Each registry obtains annual approval from its institutional review board for consenting processes or a waiver of consent, enrollment of participants, and ongoing data linkage for research purposes. All registries have received a Certificate of Confidentiality from the federal government that protects the identities of research participants. Measurement of Risk Factors Patient information was obtained primarily from self-report at the time of mammography. We selected 2 risk factors in addition to breast density for inclusion in the model on the basis of simplicity (yes or no) and a high attributable risk: history of breast cancer in a first-degree relative and history of a breast biopsy. Body mass index was later considered for addition to the model, but it was excluded to maintain parsimony and because it had minimal effect on model discrimination (the increase in the concordance statistic [c-statistic] was only 0.003). For modeling and validation, missing data for relatives with breast cancer and number of breast biopsies were set to 0. The 5-year Gail risk was computed for each woman by using the algorithms provided by the National Cancer Institute to calculate the Gail model risk for individual women (31). For Gail model calculations, missing data were coded as specified by that model (age at menarche as14 years, age at first live birth as<20 years, number of breast biopsies as 0, and number of first-degree relatives as 0). Ethnicity was coded by using the expanded race and ethnicity definition currently used in the Surveillance, Epidemiology, and End Results (SEER) database and U.S. Vital Statistics (non-Hispanic White, non-Hispanic Black, Asian or Pacific Islander, Native American/Alaskan Native, Hispanic, or other). We classified women who self-identified as mixed or other race with participants who did not report race and ethnicity. Breast Density Community radiologists at each site classified breast density on screening mammograms as part of routine clinical practice by using the American College of Radiology BI-RADS density categories (32): almost entirely fat (category 1), scattered fibroglandular densities (category 2), heterogeneously dense (category 3), and extremely dense (category 4). The BI-RADS category 2 was used as the reference group for breast density because it formed the largest group. Ascertainment of Breast Cancer Cases Breast cancer outcomes (invasive cancer and ductal carcinoma in situ) were obtained at each site through linkage with the regional population-based SEER program, state tumor registries, and pathology databases. Vital Status Vital status was obtained through linkage to SEER registries, state tumor registries, and the individual state vital statistics or the National Death Index. Model Development We used a proportional hazards model of invasive breast cancer to estimate the hazard ratios for each BI-RADS breast density category. Women entered the model 6 months after the index mammogram and were censored at the time of death, diagnosis of ductal carcinoma in situ, or the end of follow-up. All models were adjusted for age (in 5-year intervals) and race and ethnicity. The strength of the breast density association with breast cancer was greater for women younger than age 65 years (P for interaction< 0.001). Thus, separate models were fitted for women younger than age 65 years and for women age 65 years or older. No other interaction terms were included in the final model. We calculated similar estimates for first-degree relatives with breast cancer (yes or no) and a personal history of breast biopsy (yes or no) from the BCSC. All predictors met the proportional hazards assumption that was assessed by loglog plots and by including interaction terms with time for each predictor variable. We then developed an absolute risk model by using methods described in the Appendix Figure. The model primarily estimates predicted incidence of invasive breast cancer by using age, race or ethnicity, and breast density. These estimates are then adjusted for family history and biopsy history if available. We based our estimates of breast cancer incidence on the SEER age- and ethnicity-specific risk for invasive breast cancer (1992 to 2002) (33). Age-specific incidence for each ethnic group was estimated by fitting a third-order polynomial model to the SEER data. Age-specific incidence rates for the Native American and Alaskan Native group were inconsistent in SEER, so we excluded this group from further analyses. We calculated the baseline risk for the model by adjusting SEER incidence for the populations attributable risk for each breast density subgroup. We estimated the age- and ethnicity-specific distribution of mammographic breast density needed for these calculations by using data from a larger set of 3343047 mammograms from the BCSC. The distribution of breast density varied statistically significantly by age and by race or ethnicity (P< 0.001 for each comparison). The model used these variations by age and race to distribute the 5-year risk for invasive breast cancer across the 4 breast density subgroups. We used the methods described by Gail and colleagues (7) to translate the hazard ratios and risk factor distributions into absolute risks. The age-, sex-, and ethnicity-specific competing risks for death for women were calcula

Association between obesity and depression in middle-aged women

OBJECTIVE Evaluate the association between obesity and depression among middle-aged women. METHODS A total of 4641 female health plan enrollees aged 40-65 years completed a structured telephone interview including self-reported height and weight, the Patient Health Questionnaire (PHQ) assessment of depression; a brief measure of rate was 62%. RESULTS Prevalence of moderate or severe depression increased from 6.5% among those with body mass index (BMI) under 25 to 25.9% among those with BMI over 35. Prevalence of obesity increased from 25.4% among those with no depressive symptoms to 57.8% among those with moderate to severe depression. Independent of obesity, depression was associated with significant reductions in frequency of moderate (4.6 vs. 5.4 times per week) or vigorous (2.8 vs. 3.7 times per week) physical activity. Depression was associated with significantly higher daily caloric intake (1831 vs. 1543) among those with BMI over 30. CONCLUSIONS Among middle-aged women, depression is strongly and consistently associated with obesity, lower physical activity and (among the obese) higher caloric intake. Public health approaches to reducing the burden of obesity or depression must consider the strong association between these two common conditions.

Injectable hormone contraception and Bone density: Results from a prospective study

Background Depot medroxyprogesterone acetate (DMPA) injectable contraception may decrease bone density and increase the risk for osteoporosis in later life. Prospective data are scarce, especially of the effects of DMPA discontinuation on bone. Methods Between 1994 and 1999, we conducted a population-based prospective cohort study among women enrollees of a Washington State health maintenance organization. We enrolled 457 nonpregnant women, ages 18–39 years (183 DMPA users and 274 non-users). Bone density was measured by dual-energy x-ray absorptiometry every 6 months for 3 years. Results Bone density decreased notably among DMPA-exposed women at the spine (adjusted mean bone density was −0.0053 gm/cm2 for DMPA users compared with +0.0023 gm/cm2 for non-users for each 6-month interval) and total hip (−0.0060 compared with −0.0002 gm/cm2). This represents an annualized mean rate of change at the spine of −0.87% compared with +0.40% and, at the hip, −1.12% compared with −0.05%. Discontinuers of this method (N = 110) showed sizable increases in bone density over comparison women (for each 6-month interval, adjusted mean spine bone density was +0.0067 gm/cm2 compared with +0.0023 gm/cm2, respectively; adjusted mean hip bone density was +0.0035 compared with −0.0002 gm/cm2). Estimated annualized mean rates of change were +1.41% compared with +1.03% at the spine and +0.40% compared with −0.05% at the hip. After 30 months, mean bone density for discontinuers was similar to that of non-users. Conclusions In this study, DMPA use was strongly associated with bone density loss. Substantial postdiscontinuation recovery of bone provides evidence that the effects may be largely reversible.

Bone mineral density in women using depot medroxyprogesterone acetate for contraception

OBJECTIVE To evaluate the possible effects of depot medroxyprogesterone acetate injectable contraception on bone mineral density in reproductive-age women. METHODS We conducted a population-based cross-sectional comparison of bone mineral density levels in women using depot medroxyprogesterone acetate contraception and in women of similar age not using this method. The study recruited 457 nonpregnant women aged 18-39 years who were enrollees of a Washington state health maintenance organization. One hundred eighty-three women were receiving injections and 274 were not. Bone mineral density at several anatomic sites (spine, femoral neck, greater trochanter, and whole body) was measured using dual-energy x-ray absorptiometry. Data on other factors potentially related to bone density were collected through questionnaire and examination. RESULTS Overall, age-adjusted mean bone density levels were lower for users of this method than for nonusers at all anatomic sites: The mean difference was 2.5% for the spine (P = .03) and 2.2% for the femoral neck (P = .12). Exposure to depot medroxyprogesterone acetate continued to be significantly (P < .01) associated with decreased bone density at the femoral neck, spine, and trochanter after multivariate adjustment for other risk factors related to bone density. Age-specific comparisons indicated that the major differences in bone density between users and nonusers occurred in the youngest age group (women 18-21 years); the mean femoral neck bone density was 10.5% lower (P < .01) for the exposed women, and differences were consistent (P < .01) across all anatomic sites. We also noted a significant dose-response relation between longer use of depot medroxyprogesterone acetate and decreased bone density levels in this age group (P < .01 for all sites). CONCLUSION These results provide evidence that contraception with depot medroxyprogesterone acetate, particularly long-term use, may adversely affect bone mineral density levels in young women aged 18-21 years. The implications for future bone health need further study.

Associations of child sexual and physical abuse with obesity and depression in middle-aged women ☆

OBJECTIVE Examine whether (1) childhood maltreatment is associated with subsequent obesity and depression in middle-age; (2) maltreatment explains the associations between obesity and depression; and (3) binge eating or body dissatisfaction mediate associations between childhood maltreatment and subsequent obesity. METHODS Data were obtained through a population-based survey of 4641 women (mean age=52 years) enrolled in a large health plan in the Pacific Northwest. A telephone survey assessed child sexual and physical abuse, obesity (BMI>or=30), depressive symptoms, binge eating, and body dissatisfaction. Data were analyzed using logistic regression models incorporating sampling weights. RESULTS Both child sexual and physical abuse were associated with a doubling of the odds of both obesity and depression, although child physical abuse was not associated with depression for the African American/Hispanic/American Indian subgroup. The association between obesity and depression (unadjusted OR=2.82; 95% CI=2.20-3.62) was reduced somewhat after controlling for sexual abuse (adjusted OR=2.54; 1.96-3.29) and for physical abuse (adjusted OR=2.63; 2.03-3.42). Controlling for potential mediators failed to substantially attenuate associations between childhood maltreatment and obesity. CONCLUSIONS This study is the first to our knowledge that compares associations of child abuse with both depression and obesity in adults. Although the study is limited by its cross-sectional design and brief assessments, the fact that child abuse predicted two debilitating conditions in middle-aged women indicates the potential long-term consequences of these experiences.

Relationship Between Mammographic Density and Breast Cancer Death in the Breast Cancer Surveillance Consortium

BACKGROUND Women with elevated mammographic density have an increased risk of developing breast cancer. However, among women diagnosed with breast cancer, it is unclear whether higher density portends reduced survival, independent of other factors. METHODS We evaluated relationships between mammographic density and risk of death from breast cancer and all causes within the US Breast Cancer Surveillance Consortium. We studied 9232 women diagnosed with primary invasive breast carcinoma during 1996-2005, with a mean follow-up of 6.6 years. Mammographic density was assessed using the Breast Imaging Reporting and Data System (BI-RADS) density classification. Hazard ratios (HRs) and 95% confidence intervals (CIs) were estimated by Cox proportional hazards regression; women with scattered fibroglandular densities (BI-RADS 2) were the referent group. All statistical tests were two-sided. RESULTS A total of 1795 women died, of whom 889 died of breast cancer. In multivariable analyses (adjusted for site, age at and year of diagnosis, American Joint Committee on Cancer stage, body mass index, mode of detection, treatment, and income), high density (BI-RADS 4) was not related to risk of death from breast cancer (HR = 0.92, 95% CI = 0.71 to 1.19) or death from all causes (HR = 0.83, 95% CI = 0.68 to 1.02). Analyses stratified by stage and other prognostic factors yielded similar results, except for an increased risk of breast cancer death among women with low density (BI-RADS 1) who were either obese (HR = 2.02, 95% CI = 1.37 to 2.97) or had tumors of at least 2.0 cm (HR = 1.55, 95% CI = 1.14 to 2.09). CONCLUSIONS High mammographic breast density was not associated with risk of death from breast cancer or death from any cause after accounting for other patient and tumor characteristics. Thus, risk factors for the development of breast cancer may not necessarily be the same as factors influencing the risk of death after breast cancer has developed.

Vaginal Douching as a Risk Factor for Cervical Chlamydia trachomatis Infection

Objective To evaluate the association between vaginal douching and cervical Chlamydia trachomatis infection. Methods We analyzed cross-sectional data from a study conducted at Group Health Cooperative of Puget Sound, a nonprofit health maintenance organization in western Washington state. Participants were nonpregnant women Group Health enrollees between the ages of 18 and 34 years who were attending two primary care clinics either for nonurgent visits, primarily routine preventive health visits, or in response to an invitation from the study. Before the clinical examination, all completed a self-administered survey assessing demographic and behavioral characteristics, including the timing, frequency, products used, and reasons for douching. Chlamydial infection was ascertained via cell culture isolation of C trachomatis from endocervical specimens obtained at the same visit. Results Chlamydia trachomatis was isolated from cervical cultures in 58 (3.4%) of 1692 study participants. Women who reported douching in the 12 months before their clinic visit had an increased likelihood of chlamydial infection compared with women who did not douche (prevalence odds ratio [OR] 2.29, 95% confidence interval [CI] 1.22, 4.30, after adjusting for confounding factors). The likelihood was higher for women who reported douching more often: OR 2.60 (95% CI 1.29, 5.24) for women who douched one to three times per month, and OR 3.84 (95% CI 1.26, 11.70) for those douching four times or more per month. These associations were slightly stronger when women who reported douching because of an infection were excluded from the analysis. Conclusion: These results support the hypothesis that vaginal douching predisposes to acquisition of cervical chlamydial infection and are compatible with previous studies that report associations between douching and sequelae of chlamydial infection, including pelvic inflammatory disease, ectopic pregnancy, and infertility.

Evaluating Organized Breast Cancer Screening Implementation The Prevention of Late-Stage Disease?

The objective of our study was to evaluate organized breast cancer screening implementation by measuring the association between screening program enrollment and late-stage disease. Our setting was a health plan using mailed mammography reminders to women ages ≥40. We conducted yearly cross-sectional summaries of mammography experience and late-stage (regional or distant Surveillance Epidemiology and End Results Reporting (SEER) stage) breast cancer occurrence for all of the health-plan women ages ≥40 (1986–1998). We estimated the odds of late-stage breast cancer among health-plan and surrounding community women because it was too early to compare changes in mortality. We also estimated the odds of late-stage disease (1995–1998) associated with program enrollment and mammography screening among health-plan women. We found that mammography-within-two-years increased within the health plan from 25.9% to 51.2% among women ages 40–49 and from 32.9% to 74.7% among women ages ≥50. Health-plan late-stage rates were lower than those in the surrounding community [ages 40–49: odds ratio (OR), 0.87; 95% confidence interval (CI), 0.77–0.99; ages 50–79: OR, 0.86; 95% CI, 0.80–0.92] and declined parallel to the community. Among health-plan cancer cases, women ages ≥43 who were enrolled in the screening program and who had at least one program mammogram were less likely to have late-stage disease compared with the women not enrolled in the program (OR, 0.31; 95% CI, 0.16–0.61) but the odds of late-stage was also reduced among program-enrolled women not receiving program mammograms (OR, 0.45; 95% CI, 0.21–0.95). We concluded that enrollment in organized screening is associated with increased likelihood of mammography and reduced odds of late-stage breast cancer. Addressing the concerns of un-enrolled women and those without mammograms offers an opportunity for further late-stage disease reduction.