Survival Analysis Techniques Research Articles

Sexual Debut Among Heterosexual Men of African and Caribbean Descent: Are the Youth Initiating Sex Earlier than the Older Generation?

Although age at first sex is considered a measure of sexual risk and vulnerability for HIV infection, there is a dearth of literature on age at sexual debut in the Canadian context. This study examined time variations to first sex among heterosexual African, Caribbean, and Black (ACB) men in four Ontarian cities. A population-based retrospective survey (n = 879) on timing to first sexual intercourse was conducted between 2018 and 2019 among self-identified heterosexual ACB men 16years or older and residing in London, Ottawa, Toronto or Windsor. We used the lognormal survival analysis technique to examine variations in time to first sexual intercourse among age cohorts and between cities. The findings showed a generational shift in the pattern of sexual initiation, with younger heterosexual ACB men initiating sexual intercourse earlier compared with those currently older than 50years. We observed those between 16 and 19years, 20 and 29years, and 30 and 39years of age to have significantly higher risk ratios of TR = 0.852, TR = 0.869, and TR = 0.855, respectively. At city level, the results show marked spatial variations, with youth in cities of Toronto, Ottawa, and London at the highest risk of early sexual debut relative to those in Windsor. Early initiation of first sexual intercourse among heterosexual ACB youth was observed with those in the larger cities being at a relatively higher risk. There is the need for programs aimed at delaying sexual debut among youth in general. It is, however, important to recognize the relative risk of those in the larger cities.

Consanguineous marriage and its relevance to divorce, polygyny and survival of marriage: evidence from a population-based analysis in Jordan

Background Consanguinity has been extensively studied in the context of its negative health outcomes for offspring and socio-demographic factors, but little evidence-based research has been done on its potential social benefits. Aim To examine the association between consanguineous marriage and the risk of divorce or separation, polygyny, and survival of marriage in Jordan. Subjects and methods Data were obtained from the 2018 Jordan Population and Family Health Survey, covering a nationally representative sample of ever-married women of reproductive age. Descriptive statistics, multiple logistic regression, and survival analysis techniques were used for data analysis. Results The prevalence of consanguineous marriage was found to be 27.5% in Jordan in 2018, while the prevalences of polygyny and divorced/separated rates were 4.4% and 4.7%, respectively. Both consanguineous and polygyny rates showed declining trends, but divorce/separation rate showed an increasing trend. Compared to non-consanguineous marriages, the risk of divorce/separation and polygyny were found to be lower among women with consanguineous marriage, while the survival of marriages was found to be higher for consanguineous marriages than for non-consanguineous marriages. Conclusions Findings revealed that consanguinity has some social benefits as it plays protective roles against divorce or separation, polygyny and enhances the survival of marriages.

Basics of survival statistics for oncologists

In clinical practice, survival curves show the fraction of patients who experienced the outcome of interest. As survival data contain “censored” observations where a patient is lost to scrutiny before experiencing the outcome, a sensible survival curve cannot be computed through simple division. This article aimed to provide various important aspects of survival analysis, censoring and various survival estimation techniques that are simple to calculate and understand, and a better visualization of statistical significance.

Identification of Bridge Key Performance Indicators Using Survival Analysis for Future Network-Wide Structural Health Monitoring.

Machine learning and statistical approaches have transformed the management of infrastructure systems such as water, energy and modern transport networks. Artificial Intelligence-based solutions allow asset owners to predict future performance and optimize maintenance routines through the use of historic performance and real-time sensor data. The industrial adoption of such methods has been limited in the management of bridges within aging transport networks. Predictive maintenance at bridge network level is particularly complex due to the considerable level of heterogeneity encompassed across various bridge types and functions. This paper reviews some of the main approaches in bridge predictive maintenance modeling and outlines the challenges in their adaptation to the future network-wide management of bridges. Survival analysis techniques have been successfully applied to predict outcomes from a homogenous data set, such as bridge deck condition. This paper considers the complexities of European road networks in terms of bridge type, function and age to present a novel application of survival analysis based on sparse data obtained from visual inspections. This research is focused on analyzing existing inspection information to establish data foundations, which will pave the way for big data utilization, and inform on key performance indicators for future network-wide structural health monitoring.

Patterns and patient factors associated with loss to follow-up in the Muhimbili sickle cell cohort, Tanzania

BackgroundMonitoring patient’s clinical attendance is a crucial means of improving retention in care and treatment programmes. Sickle cell patients’ outcomes are improved by participation in comprehensive care programmes, but these benefits cannot be achieved when patients are lost from clinical care. In this study, patients are defined as loss to follow-up when they did not attend clinic for more than 9 months. Precise information on the retention rate and characteristics of those who are not following their clinic appointments is needed to enable the implementation of interventions that will be effective in increasing the retention to care.MethodThis was a retrospective study involving sickle cell patients registered in the Muhimbili Sickle Cohort in Tanzania. Descriptive and survival analysis techniques both non-parametric methods (Kaplan-Meier estimator and Log-rank test) and semi-parametric method (Cox’s proportional hazard model), were used. A p-value of 0.05 was considered significant to make an inference from the analysis.ResultsA total of 5476 patients were registered in the cohort from 2004 to 2016. Of these, 3350 (58.13%) were actively participating in clinics, while 2126 (41.87%) were inactive, of which 1927 (35.19%) were loss to follow-up. We used data from 2004 to 2014 because between 2015 and 2016, patients were referred to other government hospitals. From the survival analysis results, pediatric (HR: 14.29,95% CI: 11.0071–18.5768, p < 0.001) and children between 5 and 17 years [HR:2.61,95% CI:2.2324–3.0705, p < 0.001] patients were more likely to be loss to follow-up than the adult (18 and above years) patients. It was found that patients with above averages for hematocrit (HR: 2.38, 95% CI: 1.0076–1.0404, p = 0.0039) or mean cell volume (HR: 4.28, (95% CI: 1.0260–1.0598, p < 0.001) were more likely to be loss to follow-up than their counterparts.ConclusionLoss to follow-up is evident in the cohort of patients in long term comprehensive care. It is, therefore, necessary to design interventions that improve patients’ retention. Suggested solutions include refresher training for health care workers and those responsible for patient follow-up on techniques for retaining patients and comprehensive transition programs to prepare patients who are moving from pediatric to adult clinics.

Age at the onset of tobacco smoking in South Africa: a discrete-time survival analysis of the prognostic factors

BackgroundWhile knowledge of onset of smoking tobacco, and associated risk factors can aid the formulation of evidence-based policy and interventions, such information is scarce in South Africa. We assessed age at onset of tobacco smoking in South Africa and identified its risk factors.MethodsWe analysed data of 15,316 respondents aged 15–98 years from the 2012 South African National Health and Nutrition Examination Survey. Descriptive statistics and survival analysis techniques were used alongside weighted percentages.ResultsOverall lifetime prevalence of smoking was 20.5%. Among the 3360 ever-smoked respondents, the overall median age at smoking onset was 18 years (Inter-quartile range (IQR) =5) with 2% starting before age 10 while 60% had smoked before age 20. Likelihood of tobacco smoking was higher among adolescents (<=20 years) and those aged 20–29 years than those aged > = 60 years, thrice higher among males, 29% higher among urban dwellers and thrice higher in Western Cape and Free State than in North West Province. The onset of tobacco smoking was earlier among males, wealthier and “coloured” people from Northern and Eastern Capes.ConclusionThe onset of tobacco smoking peaked at 15–22 years and varied by province, sex, location, race and other characteristics. The age restrictions on smoking in South Africa has changed over time, coupled with the recent open and electronic advertisement of tobacco, and social media could have influenced the earlier onset of tobacco smoking in South Africa. Stricter regulations on tobacco-related advertisement and sales should be implemented.

Intergenerational associations in onset of cannabis use during adolescence: A data synthesis approach.

The present study examined associations between parents' and their children's ages of onset of cannabis use using a data synthesis methodology to pool data from 3 similarly designed intergenerational studies. Regarding age of first use of cannabis, prospective data were collected at one or more assessments from early to late adolescence in each generation. The extent to which parent and offspring gender separately or jointly moderated intergenerational effects was examined. Data were harmonized from studies originating in the states of Washington (Bailey, Hill, Epstein, Steeger, & Hawkins, 2018), New York (Thornberry, Henry, Krohn, Lizotte, & Nadel, 2018), and Oregon (Capaldi, Kerr, & Tiberio, 2018) when the parents were in late childhood to early adolescence; analyses concerned 1,081 parents and their children from 971 unique families. Parents' and their children's age of cannabis use onset during adolescence were modeled using discrete-time survival analysis techniques. Although data were successfully synthesized across the studies, the primary hypothesis was not supported: parents' earlier age of first cannabis use during adolescence was not significantly associated with earlier onset of cannabis use in the offspring generation. Rather, parents' histories of any cannabis use in adolescence-regardless of timing-were linked with increased risk for early onset cannabis use by their children compared to parents with no history of use during adolescence. There were no significant parent, child, or parent-by-child gender moderation effects. Thus, prevention of adolescent onset of cannabis in one generation may have prevention benefits for the next. (PsycInfo Database Record (c) 2020 APA, all rights reserved).

The life and death of residential dissonants in transit-oriented development: A discrete time survival analysis

Residential dissonants, residents who are not satisfied with land use patterns in their neighbourhood, are a threat to transit-oriented development (TOD) policy because of their unsustainable transport choices. However, it is not known if their level of dissatisfaction is reduced in TODs, and if so, the time duration it takes. This study tracks dissonance status of 98 TOD residents using five waves of panel data spanning over nine years from Brisbane, Australia. The residents were classified into TOD dissonant and TOD consonant (opposite of dissonants) groups and a discrete time survival analysis technique was applied to identify time-to-event for these groups. An event was recorded if a dissonant became a consonant, or vice versa. Two discrete time hazard models were estimated using binary logistic regression analysis (one for each transition) to identify socio-demographic and built environment characteristics associated with the occurrence of an event. Results showed that about 46% of the TOD residents were dissonants at baseline. The survival functions were significantly different between dissonant and consonant classes. About half of the dissonants took-on the characteristics of consonants in just four years. In contrast, TOD consonants remained consonants relatively longer (median survival duration is 9 years). Groups that were likely to become dissonants were those with low educational status, and people born overseas. The findings suggest that TODs have an autonomous effect on changing attitudes over time, which verifies the ‘reverse causality’ hypothesis, and therefore, TODs are likely to be dissonant free naturally presumably as residents experience the benefits of TOD living. The process could be sped up with targeted policy interventions (e.g., concessionary travel card, rent relief to bear the high cost of living in TODs) for those being as, or likely to be susceptible to become dissonant.

Risk factors, incidence, and prevalence of diabetes among rural farm and non-farm residents of Saskatchewan, Canada; a population-based longitudinal cohort study.

Saskatchewan has a high prevalence of diabetes. It is the largest, rurally populated, predominantly agricultural province in Canada. This research aims to determine the risk factors associated with the incidence and longitudinal changes in the prevalence of diabetes among Saskatchewan's adult rural farm and non-farm residents. The Saskatchewan Rural Health Study (SRHS) is a prospective cohort study conducted in two phases: a baseline survey (2010, 8261 participants) and a follow-up survey (2014, 4867 participants). Generalized estimation equations and survival analysis techniques were used to determine diabetes prevalence and incidence risk factors, respectively. Incidence of diabetes among rural residents was 2.75%. Positive family history, high BMI, sleep apnea and an abnormal Epworth Sleepiness Score (ESS) were significant predictors for diabetes incidence. A substantial increase (1.98%) of diabetes prevalence was observed after four years of follow-up. Risk factors of diabetes prevalence were increasing age, male, low income, positive family history, high BMI, hypertension and heart attack. A mix of individual and contextual factors interacting in complex pathways were responsible for the high incidence and prevalence of diabetes among rural residents. The most original finding of that study was a positive association of sleep apnea, and ESS with incident diabetes warrants further research to identify a causal linkage. Increased diabetes risk among rural male insecticide users indicates an adverse consequence of unprotected chemical exposures in the agricultural field. Urgent population-based preventive measures should initiate to slow the increasing trend of diabetes prevalence among rural residents.

Insights From Survival Analyses During 12 Years of Anti–Vascular Endothelial Growth Factor Therapy for Neovascular Age-Related Macular Degeneration

Although multiple imputation models for missing data and the use of mixed-effects models generally provide better outcome estimates than using only observed data or last observation carried forward in clinical trials, such approaches usually cannot be applied to visual outcomes from retrospective analyses of clinical practice settings, also called real-world outcomes. To explore the potential usefulness of survival analysis techniques for retrospective clinical practice visual outcomes. This retrospective cohort study covered a 12-year observation period at a tertiary eye center. Of 10 744 eyes with neovascular age-related macular degeneration receiving anti-vascular endothelial growth factor (VEGF) therapy between October 28, 2008, and February 1, 2020, 7802 eyes met study criteria (treatment-naive, first-treated eyes starting anti-VEGF therapy). Eyes were excluded from the analysis if they received photodynamic therapy or macular laser, any previous anti-VEGF therapy, treatment with anti-VEGF agents other than ranibizumab or aflibercept, or had an unknown date or visual acuity (VA) value at first injection. Kaplan-Meier estimates and Cox proportional hazards modeling were used to consider VA reaching an Early Treatment Diabetic Retinopathy Study (ETDRS) letter score of 70 (Snellen equivalent, 20/40) or better, duration of VA sustained at or better than 70 (20/40), and VA declining to 35 (20/200) or worse. A total of 7802 patients (mean [SD] age, 78.7 [8.8] years; 4776 women [61.2%]; and 4785 White [61.3%]) were included in the study. The median time to attaining a VA letter score greater than or equal to 70 (20/40) was 2.0 years (95% CI, 1.87-2.32) after the first anti-VEGF injection. Predictive features were baseline VA (hazard ratio [HR], 1.43 per 5 ETDRS letter score or 1 line; 95% CI, 1.40-1.46), baseline age (HR, 0.88 per 5 years; 95% CI, 0.86-0.90), and injection number (HR, 1.12; 95% CI, 1.10-1.15). Of the 4439 of 7802 patients (57%) attaining this outcome, median time sustained at an ETDRS letter score of 70 (20/40) or better was 1.1 years (95% CI, 1.1-1.2). In this cohort study, patients with neovascular age-related macular degeneration beginning anti-VEGF therapy were more likely to experience positive visual outcomes within the first 2.0 years after treatment, typically maintaining this outcome for 1.1 years but then deteriorating to poor vision within 8.7 years. These findings demonstrate the potential usefulness of the proposed analyses. This data set, combined with the statistical approach for retrospective analyses, may provide long-term prognostic information for patients newly diagnosed with this condition.

Machine Learning for Mortality Analysis in Patients with COVID-19.

This paper analyzes a sample of patients hospitalized with COVID-19 in the region of Madrid (Spain). Survival analysis, logistic regression, and machine learning techniques (both supervised and unsupervised) are applied to carry out the analysis where the endpoint variable is the reason for hospital discharge (home or deceased). The different methods applied show the importance of variables such as age, O2 saturation at Emergency Rooms (ER), and whether the patient comes from a nursing home. In addition, biclustering is used to globally analyze the patient-drug dataset, extracting segments of patients. We highlight the validity of the classifiers developed to predict the mortality, reaching an appreciable accuracy. Finally, interpretable decision rules for estimating the risk of mortality of patients can be obtained from the decision tree, which can be crucial in the prioritization of medical care and resources.

The VLA-COSMOS 3 GHz Large Project: Average radio spectral energy distribution of active galactic nuclei

Context.As the Square Kilometer Array (SKA) is expected to be operational in the next decade, investigations of the radio sky in the range of 100 MHz–10 GHz have become important for simulating SKA observations. In determining physical properties of galaxies from radio data, the radio spectral energy distribution (SED) is often assumed to be described by a simple power law, usually with a spectral index of 0.7 for all sources. Even though radio SEDs have been shown to exhibit deviations from this assumption, both in differing spectral indices and complex spectral shapes, it is often presumed that their individual differences can be canceled out in large samples.Aims.Since the average spectral index around 1 GHz (observed-frame) is important for determining physical properties of large samples of galaxies, we aim to test whether individual differences in the spectra of radio-identified active galactic nuclei align with the simple assumption ofα = 0.7 and test the evolution of the parameters of the synchrotron aging model with redshift and radio luminosity.Methods.We use a sample of 744 radio-excess active galactic nuclei (RxAGN), defined as those that exhibit more than a 3σradio luminosity excess with respect to the value expected only from the contribution from star formation, out toz ∼ 4. We constructed their average radio SED by combining Very Large Array (VLA) observations of the COSMOS field at 1.4 GHz and 3 GHz with Giant Meterwave Radio Telescope (GMRT) observations at 325 MHz and 610 MHz. To account for nondetections in the GMRT maps, we employed the survival analysis technique. We binned the RxAGN sample into luminosity- and redshift-complete subsamples. In each bin, we constrained the shape of the average radio SED by fitting a broken power-law model.Results.We find that the RxAGN sample can be described by a spectral index ofα1 = 0.28 ± 0.03 below the break frequencyνb = (4.1 ± 0.2) GHz andα2 = 1.16 ± 0.04 above it, while a simple power-law model, capturing fewer spectral features, yields a single spectral index of 0.64 ± 0.07. By binning in 1.4 GHz of radio luminosity and redshift, we find that the power-law spectral index is positively correlated with redshift and that the broken power-law spectral index above 4 GHz is positively correlated with both the redshift and source size. By selecting sources with sizes less than 1 kpc, we find a subsample of flat-spectrum sources, which can be described by a spectral index ofα = 0.41 ± 0.07 and a broken power-law spectral index ofα1 = 0.1 ± 0.1 (α2 = 0.55 ± 0.09) below (above) a break frequency ofνb = (2.7 ± 0.5) GHz.Conclusions.We have constrained the radio SED for a sample of RxAGN in the COSMOS field using available VLA and GMRT data, corresponding to the rest-frame frequency range from ∼0.3 GHz to ∼10 GHz. We describe our derived average radio SED of RxAGN using power-law and broken power-law models, yielding a radio SED that steepens above ∼4 GHz.

Warm, dry conditions inhibit aspen growth, but tree growth and size predict mortality risk in the southwestern United States

Widespread, rapid aspen (Populus tremuloides Michx.) mortality since the beginning of the 21st century, sometimes called sudden aspen decline (SAD), has been documented in many locations across North America, but it has been particularly pronounced in the southwestern United States. We investigated the relationship among aspen growth, mortality, and climate across three forest types in northern Arizona using cross-dated tree-ring samples from 126 live and 132 dead aspens. Aspen growth was negatively correlated with warm temperatures and positively associated with higher precipitation. Using survival analysis techniques to investigate the links between aspen mortality, tree traits, and climatic conditions, we found that tree traits played a larger role in mortality risk than climate factors. Trees with larger diameters, older trees, and trees with faster growth rates over the past 50 years had a reduced risk of mortality. Management actions aimed at maintaining the most vigorous, fastest growing aspen in the region could help mitigate the impacts of a warmer, drier future.

Dynamic Survival Prediction Combining Landmarking with a Machine Learning Ensemble: Methodology and Empirical Comparison

AbstractDynamic prediction models provide predicted survival probabilities that can be updated over time for an individual as new measurements become available. Two techniques for dynamic survival prediction with longitudinal data dominate the statistical literature: joint modelling and landmarking. There is substantial interest in the use of machine learning methods for prediction; however, their use in the context of dynamic survival prediction has been limited. We show how landmarking can be combined with a machine learning ensemble—the Super Learner. The ensemble combines predictions from different machine learning and statistical algorithms with the goal of achieving improved performance. The proposed approach exploits discrete time survival analysis techniques to enable the use of machine learning algorithms for binary outcomes. We discuss practical and statistical considerations involved in implementing the ensemble. The methods are illustrated and compared using longitudinal data from the UK Cystic Fibrosis Registry. Standard landmarking and the landmark Super Learner approach resulted in similar cross-validated predictive performance, in this case, outperforming joint modelling.

Unlucky at work, unlucky in love: job loss and marital stability

This paper analyses the relationship between a husband’s job loss and marital stability, focusing on involuntary employment terminations due to plant closures or dismissals. Using discrete survival analysis techniques on data from the German Socio-Economic Panel, we find plant closures and dismissals to be associated with a 54 and 74% higher risk of marital dissolution respectively, though the strength of association varies significantly by how long ago the change in employment status occurred. We extend the previous literature by considering heterogeneity in the relationship depending on whether new employment was found. Our analysis shows that the dissolution risk remains elevated even in couples where the husband has taken up a new position. Surprisingly, the relative risk of dissolution following the first period in a new job after a job loss is about the same as the relative risk of dissolution following the first period without employment. The relationship between finding a new job and marital dissolution appears to be mediated by changes in working hours as well as wages. In two extensions, we also consider the role of the wife’s employment status in moderating the relationship and show that a wife’s job loss is not associated with a similar increase in the probability of divorce as a husband’s.

Selecting appropriate endpoints for assessing treatment effects in comparative clinical studies for COVID-19

To evaluate the efficacy and safety of a new treatment for COVID-19 vs. standard care, certain key endpoints are related to the duration of a specific event, such as hospitalization, ICU stay, or receipt of supplemental oxygen. However, since patients may die in the hospital during study follow-up, using, for example, the duration of hospitalization to assess treatment efficacy can be misleading. If the treatment tends to prolong patients' survival compared with standard care, patients in the new treatment group may spend more time in hospital. This can lead to a “survival bias” issue, where a treatment that is effective for preventing death appears to prolong an undesirable outcome. On the other hand, by using hospital-free survival time as the endpoint, we can circumvent the survival bias issue. In this article, we use reconstructed data from a recent, large clinical trial for COVID-19 to illustrate the advantages of this approach. For the analysis of ICU stay or oxygen usage, where the initiating event is potentially an outcome of treatment, standard survival analysis techniques may not be appropriate. We also discuss issues with analyzing the durations of such events.

Development of nonlinear probabilistic S-N curves using survival analysis techniques with application to steel bridges

This paper presents the application of survival analysis techniques, widely used in medical research, to the development of nonlinear probabilistic S-N curves for fatigue including the effects of other covariates. The number of cycles to failure is treated as a fictitious “time-to-event” parameter. Parametric survival analyses were performed using fatigue test data for steel bridges. The loglogistic accelerated failure time survival model was determined to be best suited for steel bridges. Results indicate that the current linear S-N curves for bridges do not provide a consistent probability of failure. Survival analysis can be a powerful tool for probabilistic fatigue analyses.

Deep learning-based survival analysis for brain metastasis patients with the national cancer database.

PurposePrognostic indices such as the Brain Metastasis Graded Prognostic Assessment have been used in clinical settings to aid physicians and patients in determining an appropriate treatment regimen. These indices are derivative of traditional survival analysis techniques such as Cox proportional hazards (CPH) and recursive partitioning analysis (RPA). Previous studies have shown that by evaluating CPH risk with a nonlinear deep neural network, DeepSurv, patient survival can be modeled more accurately. In this work, we apply DeepSurv to a test case: breast cancer patients with brain metastases who have received stereotactic radiosurgery.MethodsSurvival times, censorship status, and 27 covariates including age, staging information, and hormone receptor status were provided for 1673 patients by the NCDB. Monte Carlo cross‐validation with 50 samples of 1400 patients was used to train and validate the DeepSurv, CPH, and RPA models independently. DeepSurv was implemented with L2 regularization, batch normalization, dropout, Nesterov momentum, and learning rate decay. RPA was implemented as a random survival forest (RSF). Concordance indices of test sets of 140 patients were used for each sample to assess the generalizable predictive capacity of each model.ResultsFollowing hyperparameter tuning, DeepSurv was trained at 32 min per sample on a 1.33 GHz quad‐core CPU. Test set concordance indices of 0.7488 ± 0.0049, 0.6251 ± 0.0047, and 0.7368 ± 0.0047, were found for DeepSurv, CPH, and RSF, respectively. A Tukey HSD test demonstrates a statistically significant difference between the mean concordance indices of the three models.ConclusionOur results suggest that deep learning‐based survival prediction can outperform traditional models, specifically in a case where an accurate prognosis is highly clinically relevant. We recommend that where appropriate data are available, deep learning‐based prognostic indicators should be used to supplement classical statistics.

CEO Overconfidence and the Timeliness of Goodwill Impairments

ABSTRACT We use survival analysis techniques to examine whether overconfidence affects the likelihood and timeliness of goodwill impairments. We predict that overconfident CEOs have a lower likelihood of impairment in any firm-quarter, and take longer, on average, to impair goodwill. Using the Cox proportional-hazards model and the accelerated failure time model, we find evidence consistent with both predictions. In cross-sectional tests, we find having more financial experts on the board mitigates the effect of CEO overconfidence on the timeliness of goodwill impairments, while uncertainty in predicting a firm's future performance strengthens the effect. Additional results show that overconfident CEOs hold overly optimistic expectations of their firms' performance, and that they underweight negative market signals prior to the impairment decisions.

Religiosity differences in the transition to first intercourse before age 18 among Polish students

Poland is a Catholic country and is one of a few countries in Europe where the percentage of people declaring themselves as non-believing remains rather low. Thus, most young Poles are brought up in Catholic families and receive religious instruction at school.The purpose of this study is to estimate the effects of religiosity and gender on the risk of first intercourse before age 18 among university students in Poland. We analysed data from two comparative studies of affective life, sexual behaviour, and attitudes conducted among university students in 2001 and 2013. We used survival analysis techniques to test for relative differences in the effects of religiosity on the timing of first sexual intercourse. While the results confirm that religiosity was a significant differentiating factor in the sexual behaviours of students in Poland, they also show that religiosity was a significant predictor of sexual abstinence only if a young person regarded religion as very important. The unadjusted median age at first sex for the participants who attached no or little importance to religion was more than 2 years lower than the median age among those who considered religion very important and attended church services regularly. The findings also indicate that there were significant gender differences, with young males transitioning to sex earlier than young females, albeit only among the 2001 cohort. The multivariate analyses confirm the absence of significant gender differences in FSI before age 18 within religious groups (controlling for other factors) in the cohort of students born after 1990. The results for control variables are consistent with the literature indicating that growing up in a two-parent household has a protective effect, and that teenage lifestyles are significant. Our observation of a distinct polarity of students with respect to religiosity and the timing of sexual debut may be a starting point for further analyses of FSI determinants.