New drug candidates for depression - a nationwide population-based study.

OBJECTIVE: To investigate whether continued use of non-aspirin NSAID, low-dose aspirin, high-dose aspirin, statins, allopurinol and angiotensin agents decreases the rate of incident depression using Danish nationwide population-based registers. METHODS: All persons in Denmark who purchased the exposure medications of interest between 1995 and 2015 and a random sample of 30% of the Danish population was included in the study. Two different outcome measures were included, (i) a diagnosis of depressive disorder at a psychiatric hospital as in-patient or out-patient and (ii) a combined measure of a diagnosis of depression or use of antidepressants. RESULTS: A total of 1 576 253 subjects were exposed to one of the six drugs of interest during the exposure period from 2005 to 2015. Continued use of low-dose aspirin, statins, allopurinol and angiotensin agents was associated with a decreased rate of incident depression according to both outcome measures. Continued uses of non-aspirin NSAIDs as well as high-dose aspirin were associated with an increased rate of incident depression. CONCLUSION: The findings support the potential of agents acting on inflammation and the stress response system in depression as well as the potential of population-based registers to systematically identify drugs with repurposing potential.

Patient-controlled hospital admission for patients with severe mental disorders: a nationwide prospective multicentre study.

OBJECTIVE: To assess whether implementing patient-controlled admission (PCA) can reduce coercion and improve other clinical outcomes for psychiatric in-patients. METHODS: During 2013-2016, 422 patients in the PCA group were propensity score matched 1:5 with a control group (n = 2110) that received treatment as usual (TAU). Patients were followed up for at least one year using the intention to treat principle utilising nationwide registers. In a paired design, the outcomes of PCA patients during the year after signing a contract were compared with the year before. RESULTS: No reduction in coercion (risk difference = 0.001; 95% CI: -0.038; 0.040) or self-harming behaviour (risk difference = 0.005; 95% CI: -0.008; 0.018) was observed in the PCA group compared with the TAU group. The PCA group had more in-patient bed days (mean difference = 28.4; 95% CI: 21.3; 35.5) and more medication use (P < 0.0001) than the TAU group. Before and after analyses showed reduction in coercion (P = 0.0001) and in-patient bed days (P = 0.0003). CONCLUSION: Implementing PCA did not reduce coercion, service use or self-harm behaviour when compared with TAU. Beneficial effects of PCA were observed only in the before and after PCA comparisons. Further research should investigate whether PCA affects other outcomes to better establish its clinical value.

Evidence for clinical progression of unipolar and bipolar disorders.

OBJECTIVE: It is a widely held belief that affective disorders are progressive of nature; however, some recent reviews have questioned this belief. The objective of the present systematic literature review was to present evidence for associations between number of affective episodes and (i) the risk of recurrence of episodes, (ii) probability of recovery from episodes, (iii) severity of episodes, (iv) the threshold for developing episodes, and (v) progression of cognitive deficits in unipolar and bipolar disorders. METHOD: A systematic review comprising an extensive literature search conducted in Medline, Embase, and PsychInfo up to September 2016 and including cross-references from identified papers and reviews. RESULTS: Most of the five areas are superficially investigated and hampered by methodological challenges. Nevertheless, studies with the longest follow-up periods, using survival analysis methods, taking account of the individual heterogeneity all support a clinical progressive course of illness. Overall, increasing number of affective episodes seems to be associated with (i) increasing risk of recurrence, (ii) increasing duration of episodes, (iii) increasing symptomatic severity of episodes, (iv) decreasing threshold for developing episodes, and (v) increasing risk of developing dementia. CONCLUSION: Although the course of illness is heterogeneous, there is evidence for clinical progression of unipolar and bipolar disorders.

Continuation of lithium after a diagnosis of chronic kidney disease.

OBJECTIVE: To investigate whether continued lithium or anticonvulsant treatment after a first diagnosis of chronic kidney disease (CKD) was associated with progression to irreversible end-stage kidney disease. METHODS: Nationwide cohort study including all individuals in Denmark in a period from 1995 to 2012 with a diagnosis of CKD and (i) a history of lithium treatment (N = 754, among whom 238 patients had a diagnosis of bipolar disorder) or (ii) a history of anticonvulsant treatment (N = 5.004, among whom 199 patients had a diagnosis of bipolar disorder). End-stage CKD was defined as chronic dialysis or renal transplantation. RESULTS: Continuing lithium (HR = 0.58 (95% CI: 0.37-0.90) and continuing anticonvulsants (HR = 0.53 (95% CI: 0.44-0.64) were associated with decreased rates of end-stage CKD. In the subcohorts of patients with a diagnosis of bipolar disorder, continuing lithium was associated with decreased end-stage CKD (HR = 0.40 (95% CI: 0.17-0.98), whereas continuing anticonvulsants was not (HR = 0.70 (95% CI: 0.21-2.37). There were no interactions of continuing lithium and anticonvulsants. CONCLUSION: After an initial diagnosis of CKD, patients who are selected by their physicians to continue lithium treatment may not necessarily have an increased risk of developing end-stage CKD. Shifting to an anticonvulsant per se may not be associated with an advantage; however, this requires further investigation.

Rates and predictors of remission, recurrence and conversion to bipolar disorder after the first lifetime episode of depression--a prospective 5-year follow-up study.

BACKGROUND: In depression, non-remission, recurrence of depressive episodes after remission and conversion to bipolar disorder are crucial determinants of poor outcome. The present study aimed to determine the cumulative incidences and clinical predictors of these long-term outcomes after the first lifetime episode of depression. METHOD: A total of 301 in- or out-patients aged 18-70 years with a validated diagnosis of a single depressive episode were assessed from 2005 to 2007. At 5 years of follow-up, 262 patients were reassessed by means of the life chart method and diagnostic interviews from 2011 to 2013. Cumulative incidences and the influence of clinical variables on the rates of remission, recurrence and conversion to bipolar disorder, respectively, were estimated by survival analysis techniques. RESULTS: Within 5 years, 83.3% obtained remission, 31.5% experienced recurrence of depression and 8.6% converted to bipolar disorder (6.3% within the first 2 years). Non-remission increased with younger age, co-morbid anxiety and suicidal ideations. Recurrence increased with severity and treatment resistance of the first depression, and conversion to bipolar disorder with treatment resistance, a family history of affective disorder and co-morbid alcohol or drug abuse. CONCLUSIONS: The identified clinical characteristics of the first lifetime episode of depression should guide patients and clinicians for long-term individualized tailored treatment.

Risk factors for miscarriage from a prevention perspective: a nationwide follow-up study.

OBJECTIVE: To identify modifiable risk factors for miscarriage and to estimate the preventable proportion of miscarriages that could be attributed to these. DESIGN: Nationwide observational follow-up study. SETTING: Denmark. POPULATION: Ninety-one thousand four hundred and twenty seven pregnancies included in the Danish National Birth Cohort between 1996 and 2002. METHODS: Information on potentially modifiable risk factors before and during pregnancy was collected by means of computer-assisted telephone interviews and linkage with Danish registers, ensuring almost complete follow-up of pregnancy outcome. Modifiable risk factors for miscarriage were identified by multiple Cox regression analysis, which provided the background for our estimations of population attributable fractions. In all, 88,373 pregnancies had full information on all covariates and were included in this analysis. MAIN OUTCOME MEASURES: Miscarriage before 22 completed weeks of gestation. RESULTS: The potentially modifiable pre-pregnant risk factors associated with increased miscarriage risk were: age of 30 years or more at conception, underweight, and obesity. During pregnancy the modifiable risk factors were: alcohol consumption, lifting of >20 kg daily, and night work. We estimated that 25.2% of the miscarriages might be prevented by reduction of all these risk factors to low risk levels. Modification of risk factors acting before and during pregnancy could lead to prevention of 14.7 and 12.5%, respectively, of the miscarriages. Maternal age at conception and alcohol consumption were the most important risk factors. CONCLUSIONS: Miscarriage risk is increased by multiple potentially modifiable risk factors and a considerable proportion of miscarriages may be preventable.

Cohabitation patterns among patients with severe psychiatric disorders in the entire Danish population.

BACKGROUND: Assortative mating has been demonstrated in mental disorders but the extent of cohabitation between patients with clinically diagnosed psychiatric disease has been poorly explored. Method We conducted a register-based study of all Danes between 18 and 70 years of age in a 13-year observational period, linking data on individuals' contacts with psychiatric services with data on individuals' cohabitation status. Two different Poisson regression analyses were performed: the first comparing the rates of commencing cohabitation with a psychiatric patient between individuals, depending on whether the individuals themselves had, or did not have, a psychiatric diagnosis; the second comparing the incidence rates of psychiatric diagnoses for individuals cohabitating with psychiatric patients with the similar rates for individuals living with unaffected cohabitants. RESULTS: In total, 159 929 (5.0%) out of 3 204 633 individuals were given a psychiatric diagnosis during the study period. Diagnosed individuals had an overall rate ratio (RR) of commencing cohabitation with a psychiatric patient of 1.95 [95% confidence interval (CI) 1.90-2.00] for women and 1.65 (95% CI 1.61-1.69) for men, when compared with unaffected individuals. The overall RR of receiving a psychiatric diagnosis while cohabitating with a psychiatric patient was 2.40 (95% CI 2.31-2.49) for women and 2.91 (95% CI 2.81-3.01) for men, when compared with those cohabitating with unaffected individuals. Individuals with schizophrenia and men with bipolar disorder had the highest RR of commencing cohabitation with a cohabitant with a similar diagnosis. CONCLUSIONS: Cohabitation among individuals with severe psychiatric disorders is increased. This has implications for research and for the clinical management of patients.

Decomposition of number of life years lost according to causes of death.

We study the competing risks model and show that the cause j cumulative incidence function integrated from 0 to τ has a natural interpretation as the expected number of life years lost due to cause j before time τ. This is analogous to the τ-restricted mean lifetime, which is the survival function integrated from 0 to τ. It is discussed how the number of years lost may be related to subject-specific explanatory variables in a regression model based on pseudo-observations, and the method is exemplified using data from a bone marrow transplantation study. Finally, inclusion of standard mortality rates is discussed.

Distribution and predictors of exercise habits among pregnant women in the Danish National Birth Cohort.

Physical activity is recommended during pregnancy, although strong evidence on reproductive health is lacking. We present exercise habits and predictors of exercise during pregnancy. From the Danish National Birth Cohort (1996-2002), 88,200 singleton pregnancies were analyzed in logistic regression. About one-third of the women exercised in early/mid pregnancy and slightly less in late pregnancy. Bicycling, swimming, and low-impact activities were most common. Exercising more than three times per week was strongly correlated with older age, being a student or out of work, eating disorders, moderate alcohol consumption, and a healthy diet. Multiparity, a normal or less good self-rated health, smoking, and a less health conscious diet were the strongest predictors of not doing exercise. Women of 25 years or older, with metabolic or psychiatric disorders, or who had received subfecundity treatment were more likely to increase their activity level substantially from early to late pregnancy than comparison groups. In conclusion, exercising during pregnancy correlated with a number of maternal characteristics. The findings may be used to identify pregnant women not likely to exercise, to target activities that may fit their needs, and, for research purposes, to identify adjustment variables or guide sensitivity analyses when data on confounders are lacking.

Does the association of education with breast cancer replicate within twin pairs? A register-based study on Danish female twins.

BACKGROUND: A positive association between socio-economic position and breast cancer has been widely observed, but not hitherto within twin pairs, where shared familial factors were adjusted for. METHODS: We used data on education and other factors from the Danish Twin Registry, The Danish Cancer Registry, and official registers on a total of 16 310 twins. Unpaired and intrapair Cox regression analyses were compared. RESULTS: In the unpaired analysis, an educational gradient in breast cancer risk was found. Similar results were seen in the intrapair analyses of all twins, although no longer statistically significant. When intrapair analyses were stratified on zygosity, the effect of education was attenuated in the monozygotic twins. CONCLUSION: The main findings support an effect of education beyond shared familial factors.

Influence of female bodyweight on IVF outcome: a longitudinal multicentre cohort study of 487 infertile couples.

This study investigated the impact of women's body mass index (BMI) on the outcome after consecutive IVF/intracytoplasmic sperm injection cycles in 487 patients initiating treatment with 5-year follow-up. The total number of cycles was 1417. In total 103 (21.1%) were overweight (BMI 25-29.9 kg/m²) and 59 (12.1%) were obese (BMI ≥ 30 kg/m²). Number of initiated cycles/woman (P=0.01), number of cancelled cycles/woman (P < 0.01) and the total dose of gonadotrophin used/cycle (P < 0.01) rose with increasing BMI. A negative linear association between BMI and the number of retrieved oocytes (B=-0.243, P < 0.001) and an inverse U-shaped relationship between BMI and the number of developed embryos was seen, with less embryos available among underweight and obese women (P=0.03). The number with positive serum human chorionic gonadotrophin/cycle decreased significantly with increasing BMI (P < 0.01). The ongoing pregnancy rate/cycle among the obese women was lower (20.8% versus 28.3% in normal-weight women; P=0.04). Live-birth rate per cycle was 15.2% versus 21.5%. Multiple logistic regression analysis showed that the only independent predictors of live birth were women's age (P=0.037), women's BMI (P=0.034) and men's age (P=0.040).

MTHFR polymorphisms and 5-FU-based adjuvant chemotherapy in colorectal cancer.

BACKGROUND: Methylenetetrahydrofolate reductase is a pivotal enzyme in folate metabolism and 5-fluorouracil (5-FU) cytotoxicity. Two common single-nucleotide polymorphisms (SNPs), MTHFR 677C>T (rs1801133) and 1298A>C (rs1801131), reduce enzyme activity. Initially, these SNPs were claimed to predict clinical efficacy, but further studies have yielded contradictory results. We tested whether these two polymorphisms are determinants of clinical outcome in a large patient group with a long follow-up time. PATIENTS AND METHODS: We included 331 patients who had been treated with adjuvant 5-FU/leucovorin chemotherapy after intended curative resection between 1997 and 2003. Clinical data, including relapse rates, overall survival, and tumor stage, were collected. DNA was extracted from formalin-fixed tumor tissue and analyzed for the MTHFR 677C>T and 1298A>C SNPs with real-time PCR. RESULTS: The MTHFR 677C>T and 1298A>C polymorphisms were not associated with survival or relapse-free survival (P > 0.2). The 677 CC genotype was associated to toxicity (odds ratio = 1.83, P = 0.01). CONCLUSIONS: The MTHFR 677C>T and 1298A>C polymorphisms probably do not predict efficacy of adjuvant 5-FU treatment in colorectal cancer after complete resection; however, the 677C>T polymorphism may be associated with lower toxicity in 5-FU treatment. Implementation of SNP analysis for these polymorphisms for individualized treatment is premature.

Epidemiological studies of the non-specific effects of vaccines: II--methodological issues in the design and analysis of cohort studies.

We review sources of bias which can affect non-randomized cohort studies of non-specific effects of vaccines on child mortality. Using examples from the literature on non-specific effects, we describe different sources of selection and information bias, and, where possible, outline analysis strategies to mitigate or eliminate such biases.

Leisure time physical exercise during pregnancy and the risk of miscarriage: a study within the Danish National Birth Cohort.

OBJECTIVE: To examine the association between leisure time physical exercise during pregnancy and the risk of miscarriage. DESIGN: Prospective study with elements of retrospective data collection. SETTING: Denmark 1996-2002. POPULATION: A total of 92,671 pregnant women enrolled in the Danish National Birth Cohort and interviewed subsequently. METHODS: Data on exercise during pregnancy and potential confounders were obtained through computer-assisted telephone interviews either during pregnancy or after an early miscarriage. Outcome of pregnancy was identified by register linkage. Using Cox regression analysis, we estimated the hazard ratio (HR) of miscarriage according to weekly amount of exercise and the type of exercise. The HR was estimated for <11, 11-14, 15-18, and 19-22 weeks of gestation, respectively. MAIN OUTCOME MEASURES: Miscarriage, defined as fetal loss before 22 completed weeks of gestation. RESULTS: A stepwise increasing relation was found between amount of exercise and risk of miscarriage, where risk of miscarriage increased by amount of exercise up to HR = 3.7 (95% CI 2.9-4.7) for women who exercised more than 7 hours per week compared with nonexercisers. Particularly 'high-impact exercise' was associated with an increased risk of miscarriage. No association was seen between exercise and risk of miscarriage after 18 weeks of gestation. CONCLUSIONS: This study suggests that exercise early in pregnancy is associated with an increased risk of miscarriage. The results should, however, be interpreted cautiously as potential bias arising from retrospective data collection may explain part of the association.

Treatment with antidepressants and lithium is associated with increased risk of treatment with antiparkinson drugs: a pharmacoepidemiological study.

OBJECTIVE: To estimate the risk for persons treated with antidepressants or lithium of subsequent treatment with antiparkinson drugs (APD). METHODS: The Danish national prescription database supplied data on all persons who received antidepressants, lithium, or antidiabetics (first control group). A second control group was included comprising persons from the general population. Outcome was purchase of APD and the study period was 1995 to 1999. RESULTS: In total, 1 293 789 persons were included. The rate ratio of treatment with APD after treatment with antidepressants was 2.27 (95% CI 2.14 to 2.42) for men and 1.50 (95% CI 1.43 to 1.58) for women. Figures for lithium were almost identical. CONCLUSION: Persons treated with antidepressants or lithium are at increased risk of subsequently treatment with APD, showing an association between anxiety/affective disorder and Parkinson's disease.

Genetic polymorphisms in the genes encoding human interleukin-7 receptor-alpha: prognostic significance in allogeneic stem cell transplantation.

Interleukin-7 (IL-7) is essential for T-cell development in the thymus and for the maintenance of peripheral T cells. IL-7 signals through IL-7R, that consists of the gammac-chain and an alpha-chain. Sequencing of IL-7Ralpha has revealed the existence of four single nucleotide polymorphisms (SNPs) (+510C/T, +1237 A/G, 2087T/C and +3110A/G), which all give rise to amino-acid substitutions. The aim of the present investigation was to evaluate the significance of IL-7Ralpha SNPs for the outcome in allogeneic stem cell transplantation (SCT). IL-7Ralpha polymorphisms were determined in 195 recipient and donor pairs from either matched sibling donors or matched unrelated donors (MUD). Genotyping of 173 normal controls was performed in parallel. In MUD transplants, the +1237 genotype of the donor was associated with survival after SCT, the mortality being highest and intermediate for the GG and AG genotypes, respectively (P = 0.023). This pattern was more pronounced with respect to treatment-related mortality (P = 0.003), while IL-7Ralpha genotypes were unrelated to the risk of relapse of leukaemia. The IL-7Ralpha +1237 genotype of the recipient and the genotypes of the other three polymorphisms, were not significantly associated with the outcome of SCT. These findings suggest that the IL-7Ralpha polymorphisms may be of importance for treatment-related mortality after SCT.

Birth characteristics, sibling patterns, and acute leukemia risk in childhood: a population-based cohort study.

BACKGROUND: The occurrence of acute lymphoblastic leukemia (ALL) and acute myeloid leukemia (AML) during childhood may be influenced by factors operating in fetal life. Furthermore, childhood ALL has been suggested to be linked to patterns of infection during infancy. PURPOSE: To explore these hypotheses and other associations, we studied the impact of sibling patterns (e.g., birth order) and birth characteristics (e.g., birth weight) on the risk of childhood ALL and AML. METHODS: By linkage of records of population-based registries, a cohort of all children whose mothers were born in Denmark from April 1935 through March 1978 was established. Children who developed ALL or AML during the period from April 1968 through December 1992 were identified by linkage with the Danish Cancer Registry. Birth weights were obtained for children born during the period from January 1973 through December 1992 by linkage with the Medical Birth Registry. RESULTS: The cohort of approximately 2.0 million children was followed for the diagnosis of ALL or AML for 20.9 million person-years. A total of 704 cases of childhood ALL were identified. Among 0-4 year olds, the relative risks (RRs) of ALL for birth order positions 1, 2, 3, and 4+ were 1.00 (reference), 0.85 (95% confidence interval [CI] = 0.68-1.07), 0.91 (95% CI = 0.66-1.25), and 0.57 (95% CI = 0.30-1.06), respectively (P for trend = .09). A decreasing trend was not observed among 5-14 year olds. A significant log-linear association between birth weight and the risk of ALL was observed for both age groups. Overall, the RR of ALL increased by a factor of 1.46 (95% CI = 1.18-1.81) (P = .0005) for each kilogram of increase in birth weight. A total of 114 cases of childhood AML were identified. Children born second or later in the birth order had an increased risk of AML (RR = 1.53; 95% CI = 1.01-2.32) compared with firstborns. A particularly high risk of AML at ages 2 (RR = 2.53; 95% CI = 1.46-4.40) and 3 years was associated with having siblings compared with being an only child at those ages. Similar to the findings for ALL risk, there was a significant association between birth weight and AML risk. The relative increase in AML risk per 1-kg increase in birth weight was 2.14 (95% CI = 1.19-3.85; P = .009). CONCLUSION AND IMPLICATIONS: The association between birth weight and childhood leukemia suggests the importance of intrauterine factors. A plausible explanation may be that increasing birth weight is associated with a higher rate of cell proliferation and/or a larger number of precursor cells being at risk of malignant transformation. The inverse association between birth order and ALL risk among 0-4 year olds was weak, but it was compatible with the hypothesis that delayed exposure to infection may increase the risk of ALL in this age group. The association of childhood AML with birth order and sibship size at young ages deserves further attention in the search for environmental factors that affect childhood AML risk.

Prognostic factors in small cell lung cancer: multivariate model based on 778 patients treated with chemotherapy with or without irradiation.

The relationships between prognostic factors and duration of survival in small cell lung cancer were investigated in a consecutive series of 874 patients treated with combination chemotherapy with or without irradiation. The series included 443 patients with limited and 431 patients with extensive stage disease based on staging including bone marrow examination and peritoneoscopy with liver biopsy but no routine scans. The median durations of survival for the two disease categories were 48 and 30 weeks, respectively. The influence on survival of various pretreatment factors was investigated by use of univariate methods and Cox's multivariate regression model. Patients in each stage were treated according to one of three controlled trials. Variations among the applied treatment regimens did not result in significant differences in duration of survival among patients with limited disease. An alternating regimen was superior to continuous therapy in patients with extensive disease and raised serum lactate dehydrogenase. Prognosis was correlated with disease extent. Surgical resection as well as limited stage disease thus both contributed to survival. Poor performance status, reduced hemoglobin concentration, and raised values for serum lactate dehydrogenase were significantly associated with a reduced duration of survival in both stages. Females with limited disease lived significantly longer than males while advanced age was a negative prognostic factor in extensive disease. Plasma sodium and serum urate were both predictive of survival in limited disease. Proved metastatic disease affecting specific sites or total number of metastatic sites did not carry significant prognostic information in a model including a general variable characterizing stage of disease. Fifty of the 778 patients, on whom the multiple regression model was based, were alive and disease free 2 years after the start of the treatment. Two-year survival rates were strongly correlated to groupings based on prognostic factors, and information about disease extent was not mandatory for predicting the probability of long term disease-free survival.

Determinants of complete remission induction and maintenance in chemotherapy with or without irradiation of small cell lung cancer.

The possible influence of pretreatment patient characteristics upon the probabilities of complete remission (CR) induction and maintenance was investigated in a series of 815 nonresected patients with small cell lung cancer. All patients underwent pretreatment staging which enabled allocation of 391 patients to trials for limited stage disease and 424 patients to trials for extensive disease. Three controlled trials for each disease stage were conducted between 1973 and 1981. All therapeutic regimens consisted of combinations of between three and six agents (lomustine, cyclophosphamide, methotrexate, vincristine, doxorubicin, etoposide) with or without irradiation. Thirty-five % of the limited stage patients and 18% of the extensive stage patients were alive and had achieved a complete remission 16 weeks after initiation of the treatment, i.e., after four cycles of chemotherapy. Relationships between pretreatment characteristics and the probability to pass this benchmark were examined by logistic regression analysis. The probability of CR was negatively related to increased serum lactate dehydrogenase and male sex in both disease stages. Pretreatment anemia (less than 12 g/liter) and poor performance status were associated with a reduced CR rate in limited and extensive stage disease, respectively. Factors related to the maintenance of complete remission were subsequently examined in the 211 complete responders by use of Cox's regression analysis. Complete responders with extensive disease prior to treatment had greater cumulative risk of relapse than those with limited disease (P less than 0.01). Hyponatremia had a significant negative influence on the remission duration in limited disease while age greater than 60 years and bone marrow metastases had significantly negative influence in extensive disease. Using the models it was possible to identify subgroups of patients with CR rates ranging from 5 to 55% and to stratify complete responders according to estimated risks of subsequent relapse.

A prospective study showing long-term infection with human papillomavirus 16 before the development of cervical carcinoma in situ.

Human papillomavirus 16 (HPV16) is a predominant cause of cervical neoplasia. However, no population-based study with long-term follow-up has clarified the temporal relationship between HPV16 infection and occurrence of carcinoma in situ, or the importance of recurrent or persistent infection. This nested case-control study was carried out in a population-based cohort of women participating in cytological screening whose initial smear, taken in 1969-1995, was normal. During up to 26 years of follow-up, carcinoma in situ was diagnosed in 484 eligible women. Archival smears from these women were compared with smears from 619 individually matched controls. After DNA extraction, a highly sensitive PCR system was used to detect HPV16. Among case women, the prevalence of HPV16 positivity was 56% at the time of diagnosis. The relative risk of cervical carcinoma in situ increased from 3.6 (95% confidence interval, 1.2-11.0) 13 years before diagnosis to 11.1 (95% confidence interval, 5.5-22.2) 1 year before diagnosis. Having a positive smear at entry to the cohort increased risk >5-fold, whereas having persistent infection with HPV in two subsequent smears increased risk 30-fold. We estimated that among HPV16-positive women, the median incubation period from infection to carcinoma in situ was 7-12 years. We conclude that evidence of persistent and/or recurrent infection is associated with a drastically higher risk of cervical carcinoma in situ than occasional infection with HPV16.