An Antiestrogenic Activity Score for tamoxifen and its metabolites is associated with breast cancer outcome.

PURPOSE: Endoxifen concentrations have been associated with breast cancer recurrence in tamoxifen-treated patients. However, tamoxifen itself and other metabolites also show antiestrogenic anti-tumor activity. Therefore, the aim of this study was to develop a comprehensive Antiestrogenic Activity Score (AAS), which accounts for concentration and antiestrogenic activity of tamoxifen and three metabolites. An association between the AAS and recurrence-free survival was investigated and compared to a previously published threshold for endoxifen concentrations of 5.97 ng/mL. PATIENTS AND METHODS: The antiestrogenic activities of tamoxifen, (Z)-endoxifen, (Z)-4-hydroxytamoxifen, and N-desmethyltamoxifen were determined in a cell proliferation assay. The AAS was determined by calculating the sum of each metabolite concentration multiplied by an IC50 ratio, relative to tamoxifen. The AAS was calculated for 1370 patients with estrogen receptor alpha (ERα)-positive breast cancer. An association between AAS and recurrence was investigated using Cox regression and compared with the 5.97 ng/mL endoxifen threshold using concordance indices. RESULTS: An AAS threshold of 1798 was associated with recurrence-free survival, hazard ratio (HR) 0.67 (95% confidence interval (CI) 0.47-0.96), bias corrected after bootstrap HR 0.69 (95% CI 0.48-0.99). The concordance indices for AAS and endoxifen did not significantly differ; however, using the AAS threshold instead of endoxifen led to different dose recommendations for 5.2% of the patients. CONCLUSIONS: Endoxifen concentrations can serve as a proxy for the antiestrogenic effect of tamoxifen and metabolites. However, for the aggregate effect of tamoxifen and three metabolites, defined by an integrative algorithm, a trend towards improving treatment is seen and moreover, is significantly associated with breast cancer recurrence.

Distribution of angiotensin type 1a receptor-containing cells in the brains of bacterial artificial chromosome transgenic mice.

In the central nervous system, angiotensin II (AngII) binds to angiotensin type 1 receptors (AT(1)Rs) to affect autonomic and endocrine functions as well as learning and memory. However, understanding the function of cells containing AT(1)Rs has been restricted by limited availability of specific antisera, difficulties discriminating AT(1)R-immunoreactive cells in many brain regions and, the identification of AT(1)R-containing neurons for physiological and molecular studies. Here, we demonstrate that an Agtr1a bacterial artificial chromosome (BAC) transgenic mouse line that expresses type A AT(1)Rs (AT1aRs) identified by enhanced green fluorescent protein (EGFP) overcomes these shortcomings. Throughout the brain, AT1aR-EGFP was detected in the nuclei and cytoplasm of cells, most of which were neurons. EGFP often extended into dendritic processes and could be identified either natively or with immunolabeling of GFP. The distribution of AT1aR-EGFP cells in brain closely corresponded to that reported for AngII binding and AT1aR protein and mRNA. In particular, AT1aR-EGFP cells were in autonomic regions (e.g., hypothalamic paraventricular nucleus, central nucleus of the amygdala, parabrachial nucleus, nuclei of the solitary tract and rostral ventrolateral medulla) and in regions involved in electrolyte and fluid balance (i.e., subfornical organ) and learning and memory (i.e., cerebral cortex and hippocampus). Additionally, dual label electron microscopic studies in select brain areas demonstrate that cells containing AT1aR-EGFP colocalize with AT(1)R-immunoreactivity. Assessment of AngII-induced free radical production in isolated EGFP cells demonstrated feasibility of studies investigating AT1aR signaling ex vivo. These findings support the utility of Agtr1a BAC transgenic reporter mice for future studies understanding the role of AT(1)R-containing cells in brain function.

Tamoxifen metabolite concentrations, CYP2D6 genotype, and breast cancer outcomes.

We explored whether breast cancer outcomes are associated with endoxifen and other metabolites of tamoxifen and examined potential correlates of endoxifen concentration levels in serum including cytochrome P450 2D6 (CYP2D6) metabolizer phenotype and body mass index (BMI). Concentration levels of tamoxifen, endoxifen, 4-hydroxytamoxifen (4OH-tamoxifen), and N-desmethyltamoxifen (ND-tamoxifen) were measured from samples taken from 1,370 patients with estrogen receptor (ER)-positive breast cancer who were participating in the Women's Healthy Eating and Living (WHEL) Study. We tested these concentration levels for possible associations with breast cancer outcomes and found that breast cancer outcomes were not associated with the concentration levels of tamoxifen, 4-hydroxytamoxifen, and ND-tamoxifen. For endoxifen, a threshold was identified, with women in the upper four quintiles of endoxifen concentration appearing to have a 26% lower recurrence rate than women in the bottom quintile (hazard ratio (HR) = 0.74; 95% confidence interval (CI), (0.55-1.00)). The predictors of this higher-risk bottom quintile were poor/intermediate metabolizer genotype, higher BMI, and lower tamoxifen concentrations as compared with the mean for the cohort as a whole. This study suggests that there is a minimal concentration threshold above which endoxifen is effective against the recurrence of breast cancer and that ~80% of tamoxifen takers attain this threshold.

Sex differences in the subcellular distribution of angiotensin type 1 receptors and NADPH oxidase subunits in the dendrites of C1 neurons in the rat rostral ventrolateral medulla.

The rostral ventrolateral medulla (RVLM), a region critical for the tonic and reflex control of arterial pressure, contains a group of adrenergic (C1) neurons that project to the spinal cord and directly modulate pre-ganglionic sympathetic neurons. Epidemiological data suggest that there are gender differences in the regulation of blood pressure. One factor that could be involved is angiotensin II signaling and the associated production of reactive oxygen species (ROS) by NADPH oxidase, which is emerging as an important molecular substrate for central autonomic regulation and dysregulation. In this study dual electron microscopic immunolabeling was used to examine the subcellular distribution of the angiotensin type 1 (AT(1)) receptor and two NADPH oxidase subunits (p47 and p22) in C1 dendritic processes, in tissue from male, proestrus (high estrogen) and diestrus (low estrogen) female rats. Female dendrites displayed significantly more AT(1) labeling and significantly less p47 labeling than males. While elevations in AT(1) labeling primarily resulted from higher levels of receptor on the plasma membrane, p47 labeling was reduced both on the plasma membrane and in the cytoplasm. Across the estrous cycle, proestrus females displayed significantly higher levels of AT(1) labeling than diestrus females, which resulted exclusively from plasma membrane density differences. In contrast, p47 labeling did not change across the estrous cycle, indicating that ROS production might reflect AT(1) receptor membrane density. No significant differences in p22 labeling were observed. These findings demonstrate that both sex and hormonal levels can selectively affect the expression and subcellular distribution of components of the angiotensin II signaling pathway within C1 RVLM neurons. Such effects could reflect differences in the capacity for ROS production, potentially influencing short term excitability and long term gene expression in a cell group which is critically involved in blood pressure regulation, potentially contributing to gender differences in the risk of cardiovascular disease.

Evolution of a reactive surface via subsurface defect dynamics.

We find that the topography and composition of a reactive surface can evolve during epitaxy via motion of point and line defects within the material. We observe the response of a NiAl surface to an Al atom flux with low-energy electron microscopy. Initially, new NiAl layers grow as Al atoms exchange with bulk Ni atoms. When the surface is critically enriched in Al, condensation occurs at dislocations. They dissociate, move linearly, and leave tracks of altered composition and new atomic steps. We show how these dynamics depend on the identity and quantity of point defects near the surface.

How effective are tobacco industry bar and club marketing efforts in reaching young adults?

OBJECTIVE: Recently, the tobacco industry has focused marketing efforts on young adults through bar and club promotions, such as advertising and distribution of free cigarettes in these settings. This study estimates the fraction of the California young adult population that might be exposed and potentially influenced by these efforts. DESIGN AND PARTICIPANTS: Data were from 9364 young adult (18-29 years) respondents to the cross sectional population based 2002 California Tobacco Survey. As background, we analysed social smoking (only smoke with other smokers), and enjoyment of smoking while drinking. Our main focus was on bar and club attendance, what was observed in bars and clubs, and how this might differ according to respondents' risk for future smoking. RESULTS: Social smokers comprised 30.0 (2.2)% of all current smokers, including experimenters. Nearly three quarters (74.5 (2.3)%) of current smokers/experimenters said they enjoyed smoking while drinking. About one third (33.8 (1.2)%) of all young adults said they attended bars and clubs at least sometimes; attendance was significantly higher among smokers and those at risk for future smoking. Close to 60% (57.9 (2.2)%) of bar and club attenders reported seeing cigarette advertising and promotions in these settings. Again, smokers and those at risk were more likely to report seeing such advertising and promotions in these settings. CONCLUSIONS: About 20% of all young adults and about 30% of those at risk for future smoking (including current smokers) were exposed to tobacco advertising and promotions in bars and clubs. These California results may be conservative, but nonetheless indicate that the group potentially influenced is sizable.

Adolescent smoking decline during California's tobacco control programme.

OBJECTIVE: California's comprehensive tobacco control programme was 13 years old in 2002; by then, children entering adolescence at the start of the programme were young adults. This study examines whether adolescent smoking declined over this period, whether any decline carried through to young adulthood, and whether it was specific to California. SETTING AND PARTICIPANTS: Most data were from the 1990-2002 California Tobacco Surveys (CTS) (adolescents 12-17 years, > 5000/survey, young adults 18-24 years, > 1000/survey). Additional data were from the national 1992/93-2001/02 Current Population Survey (CPS) (young adults 18-24 years, > 15,000/survey). RESULTS: Over the 13 year period in California, ever puffing declined by 70% in 12-13 year olds, by 53% in 14-15 year olds from 1992-2002, and by 34% in 16-17 year olds from 1996-2002 (CTS). As noted, the decline commenced progressively later in each older group. Smoking experimentation (1+ cigarettes) and established smoking (> 100 cigarettes in lifetime) showed similar patterns. Compared to 1990, the percentage of California young adults (CTS data) who ever experimented declined by 14%, with half of the decline from 1999-2002. CPS young adult smoking prevalence (established and now smoke everyday or some days) was constant in the rest of the USA over the entire period, but California showed a recent 18% decline from 1998/99 to 2001/02. CONCLUSIONS: California's comprehensive programme may have kept new adolescent cohorts from experimenting with cigarettes. Low young adolescent experimentation rates at programme start appeared to carry through to young adulthood, resulting in a recent drop in young adult smoking prevalence in California not observed in the rest of the USA.

Compliance and support for smoke-free school policies.

Our objective was to examine factors associated with compliance and support for a smoke-free campus before and after a 1995 campus-wide smoking ban for everyone, including teachers and visitors, in California. Adolescent (12-17 years) data from the 1993, 1996, 1999 and 2002 (N approximately 6000 each year) California Tobacco Surveys (population-based telephone surveys) were analyzed. Trends in compliance with smoke-free school policies and support for smoke-free campuses were examined among students in public and private schools. Perceived compliance with the no-smoking rule by most or all student smokers increased from 43.7 +/- 1.6% in 1993 to 71.5 +/- 1.4% in 2002. While non-smokers have overwhelmingly favored smoke-free school grounds since 1993 (more than 85% each survey year), support among current smokers increased from 55.8 +/- 4.7% in 1996 to 69.1 +/- 6.8% in 2002. Student smokers who saw teachers smoking in school were less likely to favor school smoking bans (odds ratio = 0.25, 95% confidence interval 0.12-0.49). The percentage of private school students seeing teachers smoke on school grounds has been at least twice that of public school students since 1996. Compliance with and support for smoke-free schools increased since smoking was banned on campus for everyone. Perceived compliance by teachers, much lower in private schools, appears to undermine student smokers' support of this policy. Increased efforts are necessary to communicate to teachers the importance of their modeling of policy compliance to students.

Ferromagnetic stability in Fe nanodot assemblies on Cu(111) induced by indirect coupling through the substrate.

We report collective ferromagnetic behavior with high Curie temperatures (T(c)) in Fe dot assemblies supported by the Cu(111) surface. Our ability to tune the average size and spacing of the individual dots allows us to conclude that enhanced magnetic anisotropy cannot account for this high-T(c) ferromagnetic order. Because our Monte Carlo simulations have ruled out the dipolar interaction as the dominant factor in this system, we attribute the origin of the ferromagnetic order to indirect exchange coupling via the Cu(111) substrate.

Changes in population attitudes about where smoking should not be allowed: California versus the rest of the USA.

BACKGROUND: The decade long California Tobacco Control Program is unique to the nation in its duration, emphasis, and level of funding. Programme emphasis is on changing social norms about smoking as a means to discourage smoking and thus reduce the harmful health effects of tobacco to the population. METHODS: Data from the 1992-93, 1995-96, and 1998-99 Tobacco Use Supplements to the national Current Population Survey (n > 175 000 each period) were used to examine changes in norms regarding where smoking should "not be allowed at all" in both California and in the rest of the USA. Venues queried were restaurants, hospitals, work areas, bars, indoor sports venues, and indoor shopping malls. RESULTS: There were substantial increases in the percentages of the adult population (18+ years) stating that smoking should not be allowed in the venues queried in California by 1998-99 compared to 1992-93; only modest increases were observed in the rest of the USA. In fact, for most venues, the percentages for the rest of the USA were lower in 1998-99 than in California in 1992-93. Further, the percentage increase over this period in respondents stating that smoking should not be allowed in four or more of the six venues was 30% in California and 23% in the rest of the USA. The most dramatic percentage increase in California occurred among current smokers (93%). CONCLUSIONS: A strong, comprehensive tobacco control programme such as California's can influence population norms, including those of smokers, with respect to where smoking should not be allowed.

Correlates of obesity in postmenopausal women with breast cancer: comparison of genetic, demographic, disease-related, life history and dietary factors.

BACKGROUND: Obesity in women has been associated with a variety of factors, including genetic predisposition, social class, early age at menarche, exercise, alcohol consumption and diet. Obesity is a risk factor for the occurrence and the recurrence of breast cancer in postmenopausal women, perhaps because of increased exposure to estrogen, insulin and insulin-like growth factors (IGFs). The progesterone receptor (PR) and the steroid hormone receptor coactivator pCIP/ACTR/AIB1/TRAM1/RAC3 (AIB1) are hypothesized to mediate signaling crosstalk between these hormonal pathways. Polymorphisms in both genes have been described and their association with breast cancer risk reported. If genetic factors contribute to obesity, and the PR and AIB1 genes influence estrogenic, insulin and IGF pathways, then genetic patterns resulting from PR and AIB1 polymorphisms may be associated with obesity in postmenopausal women. OBJECTIVE: We compared the PR and AIB1 genotypes of postmenopausal women with breast cancer with demographic, disease-related, reproductive, lifestyle and dietary variables in terms of the strength of their relationship with obesity (BMI> or =30 kg/m2). SUBJECTS: A total of 301 postmenopausal women previously diagnosed with Stage I, II or IIIA breast cancer, who are enrolled in the Women's Healthy Eating and Living (WHEL) study (age: 34.5-70.8 y, BMI: 17.8-54.6 kg/m2). MEASUREMENTS: The PR polymorphism PROGINS was identified by PCR. The length of the AIB1 polyglutamine repeat was determined by PCR and nondenaturing gel electrophoresis or DNA sequencing. BMI was obtained at the baseline clinic visit upon entry into the WHEL study. Information about date of diagnosis, stage of disease, tumor hormone receptor status and adjuvant treatment received were obtained from medical records. Reproductive, menstrual history, demographic, family history of cancer, smoking history and exercise frequency and intensity information were obtained from questionnaires. Dietary and alcohol intake data came from four 24-h telephone recalls of food intake obtained at the study entry. RESULTS: The combined inheritance of PROGINS A1/A1 and AIB1 28/29, 28/30, 28/31, 29/29 or 29/30 (AIB1 LG) genotypes (adjusted odds ratio (OR)=2.22 (95% confidence interval 1.25-3.93)) and early age at menarche (<12 y) (adjusted OR=2.34 (1.12-4.86)) were each associated with the risk for obesity. Current use of tamoxifen (adjusted OR=0.49 (0.28-0.87)) and an alcohol intake > or =10 g/day (adjusted OR=0.28 (0.11-0.77)) were inversely associated with BMI > or =30 kg/m2. CONCLUSION: Early age at menarche and a PROGINS A1/A1+AIB1 LG genetic pattern had comparable levels of association with obesity in this cross-sectional sample of postmenopausal women with breast cancer. Since this was a cross-sectional rather than a case-control design, the association between PROGINS and AIB1 genotype and obesity found in this sample should be considered preliminary, and must be re-evaluated with a new and larger sample.

Perforant path activation of ectopic granule cells that are born after pilocarpine-induced seizures.

Granule cells in the dentate gyrus are born throughout life, and various stimuli can affect their development in the adult brain. Following seizures, for instance, neurogenesis increases greatly, and some new cells migrate to abnormal (ectopic) locations, such as the hilus. Previous electrophysiological studies of this population have shown that they have intrinsic properties that are similar to normal granule cells, but differ in other characteristics, consistent with abnormal integration into host circuitry. To characterize the response of ectopic hilar granule cells to perforant path stimulation, intracellular recordings were made in hippocampal slices from rats that had pilocarpine-induced status epilepticus and subsequent spontaneous recurrent seizures. Comparisons were made with granule cells located in the granule cell layer of both pilocarpine- and saline-treated animals. In addition, a few ectopic hilar granule cells were sampled from saline-treated rats. Remarkably, hilar granule cells displayed robust responses, even when their dendrites were not present within the molecular layer, where perforant path axons normally terminate. The evoked responses of hilar granule cells were similar in several ways to those of normally positioned granule cells, but there were some differences. For example, there was an unusually long latency to onset of responses evoked in many hilar granule cells, especially those without molecular layer dendrites. Presumably this is due to polysynaptic activation by the perforant path. These results indicate that synaptic reorganization after seizures can lead to robust activation of newly born hilar granule cells by the perforant path, even when their dendrites are not in the terminal field of the perforant path. Additionally, the fact that these cells can be found in normal tissue and develop similar synaptic responses, suggests that seizures, while not necessary for their formation, strongly promote their generation and the development of associated circuits, potentially contributing to a lowered seizure threshold.

Hilar neuropeptide Y interneuron loss in the aged rat hippocampal formation.

Neuropeptide Y-immunoreactive (NPY-I) interneurons in the dentate gyrus are vulnerable to various insults, including septohippocampal cholinergic deafferentation. The present study examined whether a loss of NPY-I neurons occurs during aging, when the functional integrity of the septohippocampal pathway is thought to be compromised. Sets of male Long Evans rats (consisting of young and aged rats, with and without spatial learning impairments assessed by the Morris water maze) were examined. Light microscopic analysis revealed that hilar NPY-I neuronal number in matched dorsal sections was significantly decreased in aged compared to young rats. Ultrastructural analysis disclosed that the microenvironment (the types of processes apposed to the plasmalemmal surface) of NPY-I neurons also differed significantly between young and aged rats. In particular, a subgroup of NPY-I neurons, distinguished by a higher percentage of unmyelinated axon coverage of the plasmalemmal surface, was present in young, but not aged, rats. Neither the number nor the microenvironment of NPY-I neurons significantly differed between aged animals that were impaired versus unimpaired in spatial learning performance. To our knowledge these findings represent the first report of an age-associated decline in the number of a specific, neurochemically identified neuronal subpopulation within the hippocampal formation. Additionally, they closely parallel observations in 192 IgG-saporin-lesioned animals, suggesting that a distinct subgroup of NPY-I interneurons is particularly dependent on the viability of septohippocampal cholinergic innervation for its survival. Since neuronal loss was not correlated with performance, this alteration by itself does not appear to be sufficient to produce learning impairment.

Increased mu-opioid receptor labeling is found on inner molecular layer terminals of the dentate gyrus following seizures.

The hippocampal formation is a brain region sensitive to seizure development, a phenomenon thought to be mediated in part by mu-opioid receptor (MOR) activation. Previous studies have found a delayed increase in MOR immunoreactivity (IR) in the inner molecular layer (IML) of the dentate gyrus after experimentally induced seizures. However, whether these increases in MOR-IR are restricted to certain cell types or cellular compartments (i.e., presynaptic, postsynaptic, or glial profiles) has not been determined. Thus, the present study examined which subcellular profiles demonstrate changes in MOR-IR after kainic acid (KA)-induced seizures. Light microscopic (LM) analysis demonstrated seizure-induced increases in MOR-IR at three points of the IML (dorsal blade, ventral blade, and crest) at three levels of section (septal, mid-septotemporal, and temporal). Electron microscopic analysis of the IML revealed that MOR-IR was present in the same types of cellular profiles in both control and KA-treated rats. However, a significant increase in the number of MOR-labeled terminal profiles was revealed in KA-treated rats compared to controls. Additionally, some MOR-labeled terminals in KA-treated rats possessed excitatory-type morphology and contained enkephalin or dynorphin, peptides found in mossy fiber terminals. These data suggest that most of the seizure-induced increases in MOR expression in the IML are associated with terminals originating from several different neuronal populations, including granule cells, and possibly, surviving GABAergic interneurons, septal cholinergic, and/or supramamillary projection neurons.

Self-assembly of nanometer-scale magnetic dots with narrow size distributions on an insulating substrate.

The self-assembly of iron dots on the insulating surface of NaCl(001) is investigated experimentally and theoretically. Under proper growth conditions, nanometer-scale magnetic iron dots with remarkably narrow size distributions can be achieved in the absence of a wetting layer. Furthermore, both the vertical and lateral sizes of the dots can be tuned with the iron dosage without introducing apparent size broadening, even though the clustering is clearly in the strong coarsening regime. These observations are interpreted using a phenomenological mean-field theory, in which a coverage-dependent optimal dot size is selected by strain-mediated dot-dot interactions.

Was there significant tax evasion after the 1999 50 cent per pack cigarette tax increase in California?

OBJECTIVES: Several states, including California, have implemented large cigarette excise tax increases, which may encourage smokers to purchase their cigarettes in other lower taxed states, or from other lower or non-taxed sources. Such tax evasion thwarts tobacco control objectives and may cost the state substantial tax revenues. Thus, this study investigates the extent of tax evasion in the 6-12 months after the implementation of California's 0.50 dollars/pack excise tax increase. DESIGN AND SETTING: Retrospective data analysis from the 1999 California Tobacco Surveys (CTS), a random digit dialled telephone survey of California households. MAIN OUTCOME MEASURES: Sources of cigarettes, average daily cigarette consumption, and reported price paid. RESULTS: Very few (5.1 (0.7)% (+/-95% confidence limits)) of California smokers avoided the excise tax by usually purchasing cigarettes from non- or lower taxed sources, such as out-of-state outlets, military commissaries, or the internet. The vast majority of smokers purchased their cigarettes from the most convenient and expensive sources: convenience stores/gas (petrol) stations (45.0 (1.9)%), liquor/drug stores (16.4 (1.6)%), and supermarkets (8.8 (1.2)%). CONCLUSIONS: Despite the potential savings, tax evasion by individual smokers does not appear to pose a serious threat to California's excise tax revenues or its tobacco control objectives.

Patterns of cigar use in California in 1999.

BACKGROUND: Adult cigar use in California increased substantially between 1990 and 1996. METHODS: Cigar smoking prevalence is from the 1990, 1996, and 1999 California Tobacco Surveys (CTS), large cross-sectional random-digit-dialed surveys designed to identify trends in tobacco use in the California population. Questions added to the 1999 CTS allowed a more detailed assessment of cigar smoking patterns. RESULTS: Adult cigar use prevalence increased from 2.5% (95% confidence interval [CI], 2.2-2.8) in 1990 to 4.9% (95% CI=4.5-5.3) in 1996, and declined to 4.4% (95% CI=4.1-4.7) in 1999. Nearly the entire decrease was accounted for by less use in adults who had never been cigarette smokers. Among current cigar smokers in 1999, 43.3% (95% CI=37.8-48.8) had not smoked a cigar in the last month, just 16.2% (95% CI=7.9-24.5) of never cigarette smokers smoked three or more cigars in the past month, but 10.4% (95% CI=5.0-15.8) of former cigarette smokers-the group with the highest level of cigar consumption-reported daily use. CONCLUSIONS: Cigar use may have peaked in California around 1996; in 1999, the intensity of use was generally at modest levels. California's bans on smoking in bars and restaurants may limit cigar smoking while drinking, so that the observed patterns may or may not reflect those in the rest of the United States.

Persistence of depressive symptoms in adolescents.

To examine factors related to the persistence of depressive symptoms, we used a longitudinal follow-up survey of 1,176 adolescents aged 12 to 18 in the United States who reported notable depressive symptoms at baseline. Adolescents were interviewed by telephone at baseline in 1989 and at follow-up in 1993. The outcome of interest was a self-report measure of depressive symptoms experienced within the past 12 months at follow-up. Overall, 38.5% of adolescents reported persistent depressive symptoms. Marked gender differences were found, with 44.5% of girls reporting notable depressive symptoms at follow-up compared with 28.2% for boys. Moreover, current established smokers or experimenters were significantly more likely to report notable depressive symptoms compared with never smokers (42.2%, 41.4% and 33.6%, respectively). Significant multivariate predictors of notable depressive symptoms at follow-up were female gender, change in sleep problems from 1989 to 1993, change in cigarette smoking status from 1989 to 1993, engaging in physical fights, and lack of participation in sports. Many adolescents report continued depressive symptoms over a period spanning four years. Several predictors of persistent depression were identified that could be important components of interventions targeting depressed adolescents.