Ultraviolet radiation and skin cancer risk in offshore workers.

BACKGROUND: Excess skin cancer mortality and incidence have been reported among both land-based and offshore petroleum workers. The association between skin cancer and ultraviolet radiation (UVR) exposure has not been examined in these workers, although they have long off-duty periods and high average income that may allow travelling to sunny destinations. In addition, they have access to solariums free of charge on many accommodation platforms. AIMS: To prospectively examine risk of incident cutaneous melanoma (CM) and non-melanoma skin cancer (NMSC), according to sun-tanning habits with adjustment for aromatic hydrocarbon exposure. METHODS: A cohort of men employed offshore from 1965 to 1999 was linked through the Cancer Registry of Norway 1999-2012. Cox regression adapted to a stratified case-cohort design was used to estimate hazard ratios with 95% confidence intervals. RESULTS: The cohort included 24917 men. During 13.5 years of follow-up, 112 CMs and 70 NMSCs occurred. A positive dose-response relationship was seen between sunburn frequency and risk of CM (Ptrend < 0.05) and NMSC (Ptrend < 0.01). Solarium use both before and after age 20 was related to increased risk of NMSC. Sunscreen use was associated with increased risk of NMSC (Ptrend < 0.001). CONCLUSIONS: UVR exposure seems to be a significant contributor to the elevated risk of skin cancer observed in North Sea offshore workers. The positive association between solarium use and NMSC risk adds to the growing body of literature on artificial UV devices as carcinogenic.

A combination of low serum concentrations of vitamins K1 and D is associated with increased risk of hip fractures in elderly Norwegians: a NOREPOS study.

UNLABELLED: The present study investigated the risk of incident hip fractures according to serum concentrations of vitamin K1 and 25-hydroxyvitamin D in elderly Norwegians during long-term follow-up. The results showed that the combination of low concentrations of both vitamin D and K1 provides a significant risk factor for hip fractures. INTRODUCTION: This case-cohort study aims to investigate the associations between serum vitamin K1 and hip fracture and the possible effect of 25-hydroxyvitamin D (25(OH)D) on this association. METHODS: The source cohort was 21,774 men and women aged 65 to 79 years who attended Norwegian community-based health studies during 1994-2001. Hip fractures were identified through hospital registers during median follow-up of 8.2 years. Vitamins were determined in serum obtained at baseline in all hip fracture cases (n = 1090) and in a randomly selected subcohort (n = 1318). Cox proportional hazards regression with quartiles of serum vitamin K1 as explanatory variable was performed. Analyses were further performed with the following four groups as explanatory variable: I: vitamin K1 ≥ 0.76 and 25(OH)D ≥ 50 nmol/l, II: vitamin K1 ≥ 0.76 and 25(OH)D < 50 nmol/l, III: vitamin K1 < 0.76 and 25(OH)D ≥ 50 nmol/l, and IV: vitamin K1 < 0.76 and 25(OH)D < 50 nmol/l. RESULTS: Age- and sex-adjusted analyses revealed an inverse association between quartiles of vitamin K1 and the risk of hip fracture. Further, a 50 % higher risk of hip fracture was observed in subjects with both low vitamin K1 and 25(OH)D compared with subjects with high vitamin K1 and 25(OH)D (HR 1.50, 95 % CI 1.18-1.90). The association remained statistically significant after adjusting for body mass index, smoking, triglycerides, and serum α-tocopherol. No increased risk was observed in the groups low in one vitamin only. CONCLUSION: Combination of low concentrations of vitamin K1 and 25(OH)D is associated with increased risk of hip fractures.

Benzene exposure and risk of lymphohaematopoietic cancers in 25 000 offshore oil industry workers.

BACKGROUND: The aim of this work was to examine the risk of lymphohaematopoietic (LH) cancer according to benzene exposure among offshore workers. METHODS: Cancer registry data were used to identify 112 cancer cases diagnosed during 1999-2011 in a cohort of 24 917 Norwegian men reporting offshore work between 1965 and 1999. Analyses were conducted according to a stratified case-cohort design with a reference subcohort of 1661 workers. Cox regression was used to estimate hazard ratios with 95% confidence intervals, adjusted for other benzene exposure and smoking. RESULTS: Most workers were exposed to benzene for <15 years. The upper range values of average intensity and cumulative exposure were estimated to 0.040 p.p.m. and 0.948 p.p.m.-years, respectively. Risks were consistently elevated among exposed workers for all LH cancers combined and for most subgroups, although case numbers were small and yielded imprecise risk estimates. There was evidence of dose-related risk patterns according to cumulative exposure for acute myeloid leukaemia (AML), multiple myeloma (MM) (P trends 0.052 and 0.024, respectively), and suggestively so for chronic lymphocytic leukaemia (CLL) according to average intensity (P trend 0.094). CONCLUSIONS: Our results support an association between cumulative and intensity metrics of low-level benzene exposure and risk for AML, MM, and suggestively for CLL.

Low serum concentrations of alpha-tocopherol are associated with increased risk of hip fracture. A NOREPOS study.

UNLABELLED: We investigated the risk of hip fracture according to circulating alpha-tocopherol, a plant-derived substance with antioxidant properties, in community-dwelling older Norwegians. We found a linear increasing risk of hip fracture with lower serum alpha-tocopherol concentrations, with a 51% higher risk in the lowest compared to the highest quartile. INTRODUCTION: Oxidative stress is a suggested contributing cause of osteoporosis and fractures. Vitamin E (α-tocopherol) has potent antioxidant properties in humans. The relationship between circulating α-tocopherol and fracture risk is not established. The aim of this study was to investigate the association between serum α-tocopherol concentrations and risk of hip fracture during up to 11 years of follow-up. METHODS: We performed a case-cohort analysis among 21,774 men and women aged 65-79 years who participated in four community-based health studies in Norway 1994-2001. Serum α-tocopherol concentrations at baseline were determined in 1,168 men and women who subsequently suffered hip fractures (median follow-up 8.2 years) and in a random sample (n = 1,434) from the same cohort. Cox proportional hazard regression adapted for gender-stratified case-cohort data was performed. RESULTS: Median (25, 75 percentile) serum α-tocopherol was 30.0 (22.6, 38.3) µmol/L, and it showed a linear inverse association with hip fracture: hazard ratio (HR) 1.11 (95% confidence interval (CI) 1.04-1.20) per 10-µmol/L decrease in serum α-tocopherol, adjusted for gender and study center. The lowest compared to the highest quartile conferred an HR of 1.51 (95% CI 1.17-1.95), adjusted for gender and study center. Adjustment for smoking, month of blood sample, BMI, education, physical inactivity, self-rated health, and serum 25-hydroxyvitamin D (25(OH)D) yielded similar results. Taking serum total cholesterol concentration into account attenuated the association somewhat: HR of hip fracture was 1.37 (95% CI 1.05-1.77) in first versus fourth quartile of serum α-tocopherol/total cholesterol ratio. CONCLUSIONS: Low serum concentrations of α-tocopherol were associated with increased risk of hip fracture in older Norwegians.

Hypertensive disorders in pregnancy and fetal death at different gestational lengths: a population study of 2 121 371 pregnancies.

OBJECTIVE: To compare the proportion of offspring that was stillborn in pregnancies with pre-eclampsia, gestational hypertension or chronic hypertension with those in normotensive pregnancies. DESIGN: Register-based observational study. SETTING: The Medical Birth Registry of Norway. POPULATION: All singleton births after 20 completed weeks of gestation in Norway from 1967 to 2006 (n = 2 121 371). METHODS: The proportion of stillborn offspring was estimated in normotensive pregnancies, and in pregnancies with pre-eclampsia, gestational and chronic hypertension at different gestational lengths. In addition, changes in the proportions of stillborn offspring by maternal hypertensive disorder from 1967-1986 to 1987-2006 were estimated. MAIN OUTCOME MEASURES: Fetal death. RESULTS: The prevalence of hypertensive disorders in pregnancy was 4.7%. In total, 17 933 fetal deaths occurred and 9.2% of these were in hypertensive pregnancies. In normotensive pregnancies, 0.8% (16 290/2 022 400) experienced fetal death. This was true for 1.9% (1170/62 261) of the pregnancies with pre-eclampsia, 1.2% (390/32 068) with gestational hypertension and 1.8% (83/4642) with chronic hypertension. There was a 44% overall reduction in fetal death rate from 1967-1986 to 1987-2006. The largest decline was in women with pre-eclampsia (80% reduction). In women with gestational hypertension and chronic hypertension, the overall reductions in fetal death rates were 49% and 57%, respectively, comparable with the 41% decline in normotensive pregnancies. CONCLUSIONS: In our nationwide study during 1967-2006, the risk of fetal death among women with hypertensive disorders in pregnancy has been greatly reduced, especially among pre-eclamptic women at term. The risk of fetal death among women with gestational or chronic hypertension has also decreased, but in a different manner.

Changes in fetal death during 40 years-different trends for different gestational ages: a population-based study in Norway.

OBJECTIVE: To study changes in gestational-age-specific fetal death risks during a 40-year period. DESIGN: Register-based observational study. SETTING: The Medical Birth Registry of Norway. POPULATION: All pregnancies after 16 weeks of gestation in Norway from 1967 to 2006 (n = 2, 182, 756). METHOD: Changes in fetal death risk since 1967-1971 (reference) were estimated as absolute risks (rates) and relative risks (RR) in ongoing pregnancies at the following gestational weeks; 16-22, 23-29, 30-36 and 37-43. MAIN OUTCOME MEASURES: Fetal death. RESULTS: In all pregnancies lasting longer than 22 weeks, the fetal death rate decreased during 1967-2006. The greatest decline was in term pregnancies (37-43 weeks) from 10.8 to 3.3 fetal deaths per 1000 at risk (crude RR 0.35; 95% CI 0.31-0.38) comparing 1967-1971 with 2002-2006. In pregnancies at 30-36 weeks the fetal death rate declined from 4.5 to 1.1 per 1000 (crude RR 0.23; 95% CI 0.21-0.26). At 23-29 weeks, the rate declined from 2.8 to 1.3 per 1000 (crude RR 0.46; 95% CI 0.40-0.52). An opposite trend was observed at early gestation (16-22 weeks) with an increase from 1.7 to 3.4 fetal deaths per 1000 ongoing pregnancies (crude RR 2.05; 95% CI 1.84-2.27). Adjustments for maternal age, parity, multiple pregnancies, paternal age and pre-eclampsia did not significantly alter the estimated associations. CONCLUSION: Since 1967 the risk of fetal death has been reduced by almost 70% in pregnancies lasting longer than 22 weeks; however, at 16-22 weeks of gestation there was an increase in risk. The causes of this increase should be further explored because it may be attributed to an increase in early delivery caused by the increased proportion of women being treated with cervical cone excision before pregnancy.

The association of maternal age with placental weight: a population-based study of 536,954 pregnancies.

OBJECTIVE: To study the association of maternal age with placental weight, birthweight and placental weight/birthweight ratio. DESIGN: Population-based study. SETTING: Medical Birth Registry of Norway. POPULATION: All singleton births in Norway in the period 1999-2008 (n = 536,954). METHODS: Z-scores of placental weight and birthweight were calculated and divided into deciles. The proportions with a small or a large placenta were calculated within each maternal age group. Also, the odds ratios of having a small (lowest decile) and a large (highest decile) placenta, according to maternal age, were estimated, with and without adjustment for birthweight in grams, parity, smoking, pre-eclampsia and diabetes. MAIN OUTCOME MEASURES: Placental weight, birthweight and placental weight/birthweight ratio. RESULTS: The mean placental weight increased with maternal age: 647.1 g in women below the age of 20 years and 691.3 g in women aged 45 years or older. Among the oldest group of women (≥45 years) 15.8% of placentas were in the highest decile of placental weight z-score, whereas this was true for just 7.0% of women below the age of 20 years (Wald test, P < 0.001). Using women younger than 20 years of age as a reference, the odds ratio for having a placenta in the highest decile of placental weight z-score was 2.50 (95% CI 1.92-3.26) for women aged 45 years or older, after adjustment for offspring birthweight, parity, maternal smoking, pre-eclampsia and diabetes. CONCLUSION: We found an association between increased placental weight and maternal age, and this finding may be important in understanding the causes of adverse events associated with high maternal age.

Maternal human parvovirus B19 infection and the risk of fetal death and low birthweight: a case-control study within 35 940 pregnant women.

OBJECTIVES: To assess the association between maternal parvovirus B19 infection and fetal death, birthweight and length of gestation. DESIGN: Case-control study. SETTING: Population based. POPULATION: Cases were all 281 women with fetal death within a cohort of 35 940 pregnant woxmen in Norway. The control group consisted of a random sample of 957 women with a live born child. METHOD: Information on pregnancy outcome was obtained from the Medical Birth Registry of Norway. First trimester serum samples were tested for antibodies against parvovirus B19 (IgM and IgG). In seronegative women, further serum was analysed to detect seroconversion during pregnancy. MAIN OUTCOME MEASURES: Fetal death, length of gestation and birthweight. RESULTS: Two of 281 (0.7%) of the women who experienced fetal death and nine of 957 (0.9%) of the controls had presence of IgM antibodies, crude odds ratio 0.8; 95% CI (0.2-3.5). In initially, seronegative women, 3.1% (2/65) with fetal death and 2.6% (8/307) with a live birth seroconverted, crude odds ratio 1.2; 95% CI (0.2-5.7). Presence of maternal parvovirus-specific IgG or IgM antibodies in the first trimester, or seroconversion during pregnancy were not associated with lower birthweight or reduced length of gestation in live born children, but was associated with low birthweight in stillborn offspring. CONCLUSION: Maternal parvovirus B19 infection was not associated with fetal death in our study. Very few cases of fetal death may be attributed to maternal parvovirus B19 infection.

Association between level of physical activity and lung function among Norwegian men and women: the HUNT study.

OBJECTIVE: To estimate the association between level of physical activity in 1984-1986 and 1995-1997 and lung function in 1995-1997 among Norwegian men and women aged 28-80 years. DESIGN: In 1984-1986 and 1995-1997, all residents of Nord-Trøndelag County, Norway, aged > or =20 years were invited to participate in the Nord-Trøndelag Health Studies. These analyses included a sample that took part in both studies and underwent spirometry (n = 8047). We used linear regression models adjusting for potential confounders stratified by sex and age groups (28-49 years, 50-69 years and > or =70 years) to estimate the association between forced expiratory volume in 1 second (FEV1) and physical activity. RESULTS: Men and women who were physically active in 1985 and 1995 had the highest lung function in both sexes and in all age groups. The reduction in FEV1 ranged from 20 ml to 170 ml, similar to 1-7% of predicted values dependent on physical activity level. Lung function was also associated with body mass index (BMI), height, smoking and subjective health. CONCLUSIONS: The findings show that a high level of physical activity corresponds to about 3-5 years of normal decline in FEV1 (30 ml/year), and may therefore overcome the disadvantages of a decline in FEV1 from increasing age.

The estimated impact of the CCR-5 delta32 gene deletion on HIV disease progression varies with study design. Oslo HIV Cohort Study Group.

OBJECTIVES: To study the impact of the genotype CCR-5 wild-type +/A32 on the progression rate to AIDS and death, and to discuss sources of bias according to study design. METHODS: A prospective study of 310 HIV-positive subjects with follow-up time from study entry (prevalent cohort), and a prospective study of 105 HIV-positive subjects with well-defined time of HIV seroconversion, with follow-up time from the retrospectively assessed date of HIV seroconversion (retrospective incident cohort). RESULTS: Slower progression to AIDS among subjects with CCR-5 +/delta32 than those with CCR-5 +/+ genotype was estimated in the prevalent cohort (P=0.07, log-rank test). Slower progression to death from any cause was also estimated for subjects with CCR-5 +/delta32 (P < 0.05, log-rank test). No differences in survival after AIDS diagnosis were seen (P=0.89, log-rank test). No differences in the progression rate to AIDS (P=0.82, log-rank test) or death (P=0.78, log-rank test) were estimated in the retrospective incident cohort. CONCLUSIONS: The varying estimates of the impact of CCR-5 genotype on progression to AIDS in this and other studies, may be real and reflect differences in the dependence of HIV on the CCR-5 receptor, or may be due to systematic errors caused by study design. Several methodological difficulties occur when the factor studied, such as CCR-5 genotype, is associated both with the risk of being HIV-infected and the progression of disease.

Differences in mortality rates and causes of death between HIV positive and HIV negative intravenous drug users.

In order to study differences in mortality and causes of death between HIV positive and HIV negative intravenous drug users (IVDU), 1009 (180 HIV positive and 829 HIV negative) IVDU in Oslo, Norway, were followed from their first HIV test for a mean period of 36 months (range 1-67 months). Eighty-seven (55 HIV negatives and 32 HIV positives) died during the follow-up period. The risk of death for IVDU was 31 times higher than for the general population. The estimated probability of survival after 3 years of follow-up was 0.92 for the whole cohort, 0.93 for the HIV negative group and 0.87 for the HIV positive group (P < 0.001, log rank test). In a Cox regression analysis, HIV positivity, > 30 years of age and > 5 years of IV drug use prior to study entry were all significantly associated with a fatal outcome. Eleven per cent (n = 20) of the HIV positives and 4% (n = 38) of the HIV negatives died from drug overdose, which accounted for 68% of all deaths; 2.2% of the HIV positives and 0.4% (n = 4) of the total cohort died from AIDS. Drug overdose represented the major threat to life among IVDU in this study. Because of the dynamics of the HIV epidemic, AIDS may have an increasing impact on mortality. However, in order to forecast the number of AIDS cases among IVDU the high non-AIDS mortality must be controlled for.

Changing paternity and time since last pregnancy; the impact on pre-eclampsia risk. A study of 547 238 women with and without previous pre-eclampsia.

BACKGROUND: Long time interval between pregnancies has been found to increase the risk of pre-eclampsia in second pregnancy. Our aim was to investigate whether this effect is influenced by a history of pre-eclampsia or a change in paternity. METHODS: We studied 547 238 women with a first and second pregnancy registered in the Medical Birth Registry of Norway, 1967-1998. The relative risk of pre-eclampsia in the second delivery according to time interval between deliveries was estimated as odds ratios (OR) in logistic regression models, controlling for changing paternity, maternal age and calendar time period in women with and without previous pre-eclampsia. RESULTS: A change of paternity for the second pregnancy was associated with a reduced risk of pre-eclampsia after controlling for the time since first delivery (adjusted OR = 0.80, 95% CI : 0.72-0.90), but the interaction between change in paternity and time between deliveries was significant only for women with no previous pre-eclampsia. The interaction between history of pre-eclampsia and time interval between the two deliveries was highly significant, and for women with no previous pre-eclampsia the risk of pre-eclampsia in second pregnancy increased with increasing time interval (for intervals longer than 15 years the adjusted OR was 2.11, 95% CI : 1.75-2.53). For women with previous pre-eclampsia the risk tended to decrease with increasing time interval between deliveries. CONCLUSIONS: The protective impact of a new father for the second pregnancy challenges the hypothesis of primipaternity, and implies that the increase in pre-eclampsia risk ascribed to new father by others is due to insufficient control for interpregnancy interval.

Exposure to nitrogen dioxide and the occurrence of bronchial obstruction in children below 2 years.

BACKGROUND: The objective of the investigation was to test the hypothesis that exposure to nitrogen dioxide (NO2) has a causal influence on the occurrence of bronchial obstruction in children below 2 years of age. METHODS: A nested case-control study with 153 one-to-one matched pairs was conducted within a cohort of 3754 children born in Oslo in 1992/93. Cases were children who developed > or = 2 episodes of bronchial obstruction or one episode lasting >4 weeks. Controls were matched for date of birth. Exposure measurements were performed in the same 14-day period within matched pairs. The NO2 exposure was measured with personal samplers carried close to each child and by stationary samplers outdoors and indoors. RESULTS: Few children (4.6%) were exposed to levels of NO2 > or = 30 microg/m3 (average concentration during a 14-day period). In the 153 matched pairs, the mean level of NO2 was 15.65 microg/m3 (+/-0.60, SE) among cases and 15.37 (+/-0.54) among controls (paired t = 0.38, P = 0.71). CONCLUSIONS: The results suggest that NO2 exposure at levels observed in this study has no detectable effect on the risk of developing bronchial obstruction in children below 2 years of age.

Symptoms of anxiety and depression among mothers of pre-school children: effect of chronic strain related to children and child care-taking.

BACKGROUND: Factors that affect maternal mental health were studied when the children were 30 and 50 months old, and changes in the importance of these factors over time were analyzed. A specific aim was to elucidate the role of chronic strain related to children and child care-taking. This study follows up previous work on the influence of social class, strain and social support on maternal mental distress when the children were 18 months old. METHODS: The sample is population based, and 1,081 parents were invited to fill out questionnaires. Maternal mental distress was measured by the Hopkins Symptom Checklist (SCL-25). Multiple regression analyses were conducted at each time point and chi-square tests were used to analyze the changes between the estimated regression coefficients over time. RESULTS: Chronic strain related to children and child care-taking consistently predicted maternal mental distress. Among the specific child related strains, problems with child care-taking were significantly associated with maternal symptom levels at all time points. The importance of two specific child problem behaviors (activity level and the child being a worrier) on maternal mental health changed over time. LIMITATIONS: Conclusions about causality can not be drawn based on cross-sectional analyses. The self-report measures used here may be biased by the current mood state. CONCLUSIONS: Problems with child care arrangements and combining work and child care-taking are predictive of maternal mental health when the children are 18, 30 and 50 months old. The risk and protective factors found here may have implications for prevention and intervention.

No correlation in post-neonatal deaths for twins. A study of the early mortality of twins based on the Norwegian Medical Birth Registry.

The records of 10,204 pairs of twins born in Norway 1967-84 in the Norwegian Birth Registry were examined. The probabilities of stillbirth, neonatal death and post-neonatal death, conditional on the fate of the cotwin, were calculated. The neonatal death rate was four times increased when the cotwin was stillborn. When the cotwin died neonatally, the probability of postneonatal death was increased ninefold. The concordance rate for stillbirth and for neonatal death was above 0.4. In striking contrast, there was no correlation in postneonatal death for members of a twin pair. This finding suggests that genetic or common environmental factors play a minor role in the aetiology of post-neonatal death.

Cohort study of the relationship between air pollution and short term health effects as determined by peak expiratory flow measurements.

The aim of the study was to establish if air pollution has short term effects on health and well-being for individuals living in an industrialized area of Norway. A cohort study was designed so that two groups (one randomly selected from the general population and one with preexisting lung disease) were followed hour by hour during two months in the winter and in the summer of 1988. In order to minimize the problems of confounding factors, each individual served as her/his own control. Each participant described through the use of a diary the presence of symptoms from the upper and lower respiratory tract as well as general symptoms of ill health. Measurements of lung function by the use of peak expiratory flow meters were done four times a day. In addition, every second week the participants were subjected t a full spirometric test. Samples of urine and blood were examined, and bacteriological test from the throat was performed at the beginning and at the end of the study. A comprehensive measurement program of outdoor air contaminants (including nitrogen oxides, sulphur dioxide) is presented. Estimation of each participant's exposure was performed hour by hour based on detailed modelling of the measured levels, known emissions of pollutants and meteorological conditions, as well as diary information on the participant's movements through the various micro-environments. The estimated exposures were generally low. In this presentation, a linear regression model and their corresponding parameter estimates were applied on an individual basis to evaluate any effect of air contamination on lung function.(ABSTRACT TRUNCATED AT 250 WORDS)

Procollagen type 1 amino-terminal propeptide (P1NP) and risk of hip fractures in elderly Norwegian men and women. A NOREPOS study.

The current study aimed to assess a possible association between the bone turnover marker procollagen type 1 amino-terminal propeptide (P1NP) and future hip fractures in elderly Norwegian men and women and to elucidate the relation between P1NP, bone mineral density and 25-hydroxyvitamin D (25(OH)D). Men and women aged 71 to 77 from two population based health studies in Norway (1999-2001) were followed for a median period of 7.3 years with respect to hip fractures. The study was designed as a case-cohort study. P1NP and 25(OH)D were analysed in frozen serum samples obtained at baseline in hip fracture patients (n=340) and in randomly selected sex stratified sub-cohorts. Bone mineral density was measured by dual-energy X-ray absorptiometry (DXA) in a subset of participants. Cox proportional hazards regression with inverse probability weighting and robust variance was performed. No significant correlation between 25(OH)D and P1NP was found. A negative correlation between P1NP and BMD was observed in women (Rho=-0.36, p=0.001). A similar trend was observed in men. No association between quartiles of P1NP and rate of subsequent hip fractures was found. Spline analyses suggested a higher rate of hip fracture at P1NP levels above 60 µg/L in both men and women. A higher hip fracture rate, which was independent of BMD, was also indicated in women with very low levels of P1NP.

A short matrix-version of the Edinburgh Depression Scale.

OBJECTIVE: Given the importance of depression as a world health problem, depression assessment should be standard practice in large-scale health surveys. The aim of the study was to construct a short matrix-version of the Edinburgh Depression Scale (EDS) that can be used in questionnaire studies. METHOD: On the basis of the complete EDS scale of ten items, answered by 2730 women, stepwise multiple regression analysis was used to find the combination of items that explains the maximum proportion of the variance of the full scale sum score. The selected EDS items were thereafter correlated with the Hopkins Symptom Check List (SCL-25) for external validation. RESULTS: The sum of five selected items from the EDS correlated at r = 0.96 with the full version. The EDS-5 scores correlated strongly with the SCL-25 (r = 0.75). CONCLUSION: The EDS-5 version shows good psychometric properties and may, for some scientific purposes, substitute the full EDS scale.