Division of Risk Management, Department of Environmental and Life Sciences, Karlstad University, Karlstad, Sweden; Centre for Public Safety, Karlstad University, Sweden. Electronic address: firstname.lastname@example.org.
Residential fires represent the largest category of fatal fires in Sweden. The purpose of this study was to describe the epidemiology of fatal residential fires in Sweden and to identify clusters of events.
Data was collected from a database that combines information on fatal fires with data from forensic examinations and the Swedish Cause of Death-register. Mortality rates were calculated for different strata using population statistics and rescue service turnout reports. Cluster analysis was performed using multiple correspondence analysis with agglomerative hierarchical clustering.
Male sex, old age, smoking, and alcohol were identified as risk factors, and the most common primary injury diagnosis was exposure to toxic gases. Compared to non-fatal fires, fatal residential fires more often originated in the bedroom, were more often caused by smoking, and were more likely to occur at night. Six clusters were identified. The first two clusters were both smoking-related, but were separated into (1) fatalities that often involved elderly people, usually female, whose clothes were ignited (17% of the sample), (2) middle-aged (45-64years old), (often) intoxicated men, where the fire usually originated in furniture (30%). Other clusters that were identified in the analysis were related to (3) fires caused by technical fault, started in electrical installations in single houses (13%), (4) cooking appliances left on (8%), (5) events with unknown cause, room and object of origin (25%), and (6) deliberately set fires (7%).
Fatal residential fires were unevenly distributed in the Swedish population. To further reduce the incidence of fire mortality, specialized prevention efforts that focus on the different needs of each cluster are required.
Cooperation between various societal functions, e.g. rescue services, elderly care, psychiatric clinics and other social services, with an application of both human and technological interventions, should reduce residential fire mortality in Sweden.
We tested the association between alcohol use disorder (AUD) and divorce; estimated the genetic and environmental influences on divorce; estimated how much genetic and environmental influences accounted for covariance between AUD and divorce; and estimated latent genetic and environmental correlations between AUD and divorce. We tested sex differences in these effects.
We identified twin and sibling pairs with AUD and divorce information in Swedish national registers. We described the association between AUD and divorce using tetrachorics and used twin and sibling models to estimate genetic and environmental influences on divorce, on the covariance between AUD and divorce and the latent genetic and environmental correlations between AUD and divorce.
A total of 670?836 individuals (53% male) born 1940-1965.
Life-time measures of AUD and divorce.
AUD and divorce were related strongly (males: rtet = +0.44, 95% CI = 0.43, 0.45; females rtet = +0.37, 95% CI = 0.36, 0.38). Genetic factors accounted for a modest proportion of the variance in divorce (males: 21.3%, 95% CI = 7.6, 28.5; females: 31.0%, 95% CI = 18.8, 37.1). Genetic factors accounted for most of the covariance between AUD and divorce (males: 52.0%, 95% CI = 48.8, 67.9; females: 53.74%, 95% CI = 17.6, 54.5), followed by non-shared environmental factors (males: 45.0%, 95% CI = 37.5, 54.9; females: 41.6%, 95% CI = 40.3, 60.2). Shared environmental factors accounted for a negligible proportion of the covariance (males: 3.0%, 95% CI = -3.0, 13.5; females: 4.75%, 95% CI = 0.0, 6.6). The AUD-divorce genetic correlations were high (males: rA = +0.76, 95% CI = 0.53, 0.90; females +0.52, 95% CI = 0.24, 0.67). The non-shared environmental correlations were modest (males: rE = +0.32, 95% CI = 0.31, 0.40; females: +0.27, 95% CI = 0.27, 0.36).
Divorce and alcohol use disorder are correlated strongly in the Swedish population, and the heritability of divorce is consistent with previous studies. Covariation between AUD and divorce results from overlapping genetic and non-shared environmental factors. Latent genetic and non-shared environmental correlations for alcohol use disorder and divorce are high and moderate.
Senior citizens get around, to a large extent, as pedestrians, and safe walking is desirable for senior citizens allowing them to stay mobile, independent and healthy in old age. Senior citizens are over-represented in injury statistics, and fall-related injuries are common. The aim of this study was to investigate fall-related injuries including healthcare costs among senior citizen pedestrians injured when walking in public outdoor environments and to describe their self-reported causes and suggested preventive strategies. The data were based on a combination of information from injury data and a questionnaire. Three hundred senior citizens attended one emergency department after sustaining injuries from pedestrian falls; 60% suffered nonminor injuries, mostly fractures. One-fifth of the pedestrians were hospitalised for an average of 8 days with an indirect hospital cost of 6.2 million EUR (55 million SEK). Environmental factors such as ice were the most commonly described cause of the injury incident. Forty per cent of the respondents indicated that the municipality was responsible for the cause of the injury incident. Fewer respondents mentioned their own responsibility as a preventive strategy. Thirty per cent described a combination of improvements such as better road maintenance, changes in human behaviour and use of safety products as preventive strategies. It is of great importance to highlight general safety, products and preventive strategies to minimise injury risks, so that pedestrians can safely realise the known health benefits of walking and thereby limit healthcare costs.
Department of Clinical and Experimental Medicine, Occupational and Environmental Medicine, Faculty of Medicine and Health Sciences, Linköping University, Linköping, Sweden (Nyqvist, Helmfrid, Dr Wingren); and Department of Biology and Environmental Science, Faculty of Health and Life Sciences, Linneaus University, Kalmar, Sweden (Dr Augustsson).
The aim of this study was to examine mortality causes and cancer incidence in a population cohort that have resided in close proximity to highly metal-contaminated sources, characterized by contamination of, in particular, arsenic (As), cadmium (Cd), and lead (Pb).
Data from Swedish registers were used to calculate standardized mortality and cancer incidence ratios. An attempt to relate cancer incidence to metal contamination levels was made.
Significantly elevated cancer incidences were observed for overall malignant cancers in both genders, cancer in the digestive system, including colon, rectum, and pancreas, and cancers in prostate among men. Dose-response relationships between Cd and Pb levels in soil and cancer risks were found.
Cancer observations made, together with previous studies of metal uptake in local vegetables, may imply that exposure to local residents have occurred primarily via oral intake of locally produced foodstuffs.
Sweden has a long tradition of recording cause of death data. The Swedish cause of death register is a high quality virtually complete register of all deaths in Sweden since 1952. Although originally created for official statistics, it is a highly important data source for medical research since it can be linked to many other national registers, which contain data on social and health factors in the Swedish population. For the appropriate use of this register, it is fundamental to understand its origins and composition. In this paper we describe the origins and composition of the Swedish cause of death register, set out the key strengths and weaknesses of the register, and present the main causes of death across age groups and over time in Sweden. This paper provides a guide and reference to individuals and organisations interested in data from the Swedish cause of death register.
We investigated occupational exposure to diesel motor exhaust (DME) and the risk of lung cancer by histological subtype among men, using elemental carbon (EC) as a marker of DME exposure. 993 cases and 2359 controls frequency-matched on age and year of study inclusion were analyzed by unconditional logistic regression in this Swedish case-control study. Work and smoking histories were collected by a questionnaire and telephone interviews. DME was assessed by a job-exposure matrix. We adjusted for age, year of study inclusion, smoking, occupational exposure to asbestos and combustion products (other than motor exhaust), residential exposure to radon and exposure to air pollution from road traffic. The OR for lung cancer for ever vs. never exposure to DME was 1.15 (95% CI 0.94-1.41). The risk was higher for squamous and large cell, anaplastic or mixed cell carcinoma than for alveolar cell cancer, adenocarcinoma and small cell carcinoma. The OR in the highest quartile of exposure duration (=34 years) vs. never exposed was 1.66 (95% CI 1.08-2.56; p for trend over all quartiles: 0.027) for lung cancer overall, 1.73 (95% CI 1.00-3.00; p: 0.040) for squamous cell carcinoma and 2.89 (95% CI 1.37-6.11; p: 0.005) for the group of undifferentiated, large cell, anaplastic and mixed cell carcinomas. We found no convincing association between exposure intensity and lung cancer risk. Long-term DME exposure was associated with an increased risk of lung cancer, particularly to squamous cell carcinoma and the group of undifferentiated, large cell, anaplastic or mixed carcinomas.
aDepartment of Medicine Solna bDepartment of Biostatistics, Institute of Environmental Medicine cDepartment of Medicine, Division of Gastroenterology, Huddinge Hospital, Karolinska Institutet, Stockholm, Sweden.
Post-colonoscopy colorectal cancer (PCCRC), a cancer occurring within a short interval of a colonoscopy, might be partly explained as missed or incompletely resected lesions. Associated risk factors are age, sex, comorbidity, cancer location, and colonoscopy volume. There is a gap in the knowledge of prevalence of PCCRC and the impact of different risk factors in Sweden.
This is a retrospective population-based observational cohort study of the colonoscopies performed on adults during the years 2001-2010 that were identified from Swedish health registers. The rate of PCCRC (diagnosed 6-36 months after the first colonoscopy) was defined as the number of PCCRCs divided by the number of colorectal cancers (CRC) in the interval of 0-36 months. Univariate and multivariate Poisson regression analyses examined associations with PCCRC.
There were 289?729 colonoscopies performed on 249?079 individuals included in the study. There were 16?319 individuals with a colorectal cancer diagnosis 0-36 months after a colonoscopy. Of these, 1286 (7.9%) were PCCRCs. In the multivariate analysis, young age (18-30 years) and former polyp diagnosis had the highest risks [relative risk (RR)=3.3; 95% confidence interval: 2.1-5.2 and RR=3.1; 95% confidence interval: 2.7-3.6]. The impact of other risk factors, such as female sex, comorbidity, right sided colorectal cancer location, and time period, was consistent with the finding in other studies.
The prevalence of PCCRC in Sweden seems to be relatively high, indicating that there is room for improvement in colonoscopy quality. The high RR of PCCRC in the youngest age group, even though there were only a few cases, has not been described in other studies.
*Department of Medical and Health Sciences, Linköping University, Linköping ‡Occupational and Environmental Medicine, University of Gothenburg, Gothenburg †Division of Occupational and Environmental Medicine, Lund University, Lund, Sweden.
Individuals with chronic pain often report increased pain sensitivity compared with pain-free individuals; hence, it is crucial to determine whether and how different pain characteristics influence or interact with pain sensitivity. An alternative to experimental pain sensitivity testing is the self-reported pain sensitivity questionnaire (PSQ), which captures pain sensitivity in various body areas.This study compares PSQ in individuals with and without pain and clarifies how pain sensitivity relates to spreading of pain on the body, and to intensity, frequency, duration of pain and to age and sex.
A total of 5905 individuals with pain and 572 individuals without pain from the general population in southeastern Sweden completed and returned a postal questionnaire.
The mean PSQ score was 3.9 (95% confidence interval [CI], 3.88-3.98) in individuals with pain and 3.5 (95% CI, 3.38-3.64) in pain-free individuals. Hence, PSQ was the highest among individuals with pain, with a difference of 0.4 (95% CI, 0.30-0.56). There was a considerable variation in the PSQ values (mean=3.5; SD=1.54) among pain-free individuals. Pain sensitivity was positively related to spreading, intensity, and frequency of pain, with a correlation coefficient of 0.3. PSQ was higher in widespread pain, 4.5 (95% CI, 4.27-4.69) in women and 4.3 (95% CI, 3.94-4.71) in men, than in local pain, 3.7 (95% CI, 3.61-3.91) in women and 3.8 (95% CI, 3.66-3.95) in men. The score for women with regional pain was between local and widespread pain at 4.0 (95% CI, 3.95-4.11) and that for men with regional pain was 3.8 (95% CI, 3.69-3.87), which is equal to that of local pain.
The positive association between pain sensitivity and spreading of pain on the body provides some evidence that the extent of spreading may be related to the degree of pain sensitivity. Before clinical use of PSQ, psychometric development and further research are needed.
This study examined the health-related quality of life (HRQoL) of teenagers with and without asthma, including the impact of their sex, allergic conditions, smoking, living conditions and physical activity.
The Obstructive Lung Disease in Northern Sweden (OLIN) studies recruited a cohort of schoolchildren in 2006. The parents of all children aged seven to eight years in three municipalities were invited to complete a questionnaire and 2585 (96%) participated. The cohort was followed up at the ages of 11-12 years and 14-15 years with high participation rates. At 14-15 years, the HRQoL questionnaire KIDSCREEN-10 and Asthma Control Test were added.
Girls with current asthma at 14-15 years had a lower mean HRQoL score than girls without asthma (46.4 versus 49.3, p
Non-Hodgkin's lymphoma (NHL) has been subject to several epidemiological studies and various occupational and non-occupational exposures have been identified as determinants. The present study is a pooled analysis of two earlier methodologically similar case-referent studies encompassing 199 cases of NHL and 479 referents, all alive. Exposure information, mainly on occupational agents, was obtained by mailed questionnaires to the subjects. Exposure to white spirits, thinner, and aviation gasoline as well as work as a painter was connected with increased odds ratios, whereas no increased risk was noted for benzene. Farming was associated with a decreased odds ratio and exposure to phenoxy herbicides, wood preservatives, and work as a lumberjack showed increased odds ratios. Moreover, exposure to plastic and rubber chemicals and also contact with some kinds of pets appeared with increased odds ratios. Office employment and housework showed decreased odds ratios. This study indicates the importance of investigating exposures not occurring very frequently in the general population. Solvents were studied as a group of compounds but were also separated into various specific compounds. The present findings suggest that the carcinogenic property of solvents is not only related to the aromatic ones or to the occurrence of benzene contamination, but also to other types of compounds.
OBJECTIVES: Familial cancer risks were studied in offspring whose parents had a similar (concordant) or a different (discordant) cancer in order to assess the modification of cancer risks from one generation to another. METHODS: We used the nation-wide Swedish Family-Cancer Database to calculate familial rate ratios (FRRs) to the offspring when their parents had concordant and discordant cancers. Cancer sites were included if there was at least one pair of parents with the same cancer. In situ cancers were included in a separate analysis in order to increase the numbers of cases. RESULTS: The risk of colon, all bowel, lung and breast cancer and melanoma increased 1.1-1.2 times when one parent and 1.3-1.6 times when two parents had any discordant cancer, suggesting involvement of environmental and hereditary effects shared by many forms of cancer. When both parents had colon cancer or melanoma, the respective risks in the offspring were 3.0 and 9.3 but only based on single triplets. For all bowel cancer the risk was 3.4, approximately multiplicative from the familial one parent-offspring risk. For concordant lung and breast cancer triplets the risk in offspring was 11.8 and 29.4, respectively. CONCLUSIONS: Even discordant cancer in parents increased cancer risk in offspring. This may be due to environmental and hereditary causes, and deserves consideration in epidemiological studies. The high risks in families where both parents had the same cancer suggest interactions of hereditary and environmental factors.
Analysis of familial cancer risks between discordant sites provides etiologic understanding on genetic and environmental risks factors of site-specific cancers. We used the Swedish nation-wide Family-Cancer Database to analyze familial risks in discordant cancers of offspring and parents. Familial risk ratios (FRRs) were calculated for cancer in offspring aged 15 to 53 years at 22 sites, discordant from parental sites. We confirmed many reported associations. Consistent novel findings associated parental-offspring sites of pancreas-breast, breast-testis and uterus-nervous system. For these, the FRRs were modest, 1.2 to 1.5 in the whole Database, but the FRRs increased in those whose parents were diagnosed before age 50. Pancreas and liver cancers showed FRRs of 2.5 to 3.3 in offspring of women and of 1.3 in offspring of men. One or both of these cancers was/were associated with cancers of stomach, colon, breast, uterus, ovary and prostate. Melanoma was associated with pancreas, breast, skin and nervous-system cancers and with leukemias. Myeloma showed a concordant FRR of about 4.0 and was associated with prostate cancer and non-thyroid endocrine-gland cancers. Mutations in known cancer-related genes may explain some of these findings, but new susceptibility genes are yet to be found. For melanoma, pancreatic and liver cancer, environmental factors are important etiologic factors and may contribute to the familial effects observed.
OBJECTIVE: To follow up cancer incidence and mortality in a group of Swedish battery workers exposed to nickel hydroxide and cadmium oxide. METHODS: 869 workers, employed at least one year between the years 1940 and 1980 were followed up until 1992. Vital status and causes of death were obtained from the Swedish cause of death registry. Cancer morbidity was retrieved from the Swedish cancer registry. Regional reference rates were used to compute the expected numbers of deaths and cancers. RESULTS: Up to 31 December, 1992, a total of 315 deaths (292 in men and 23 in women) had occurred in the cohort. For men, the overall standardised mortality ratio (SMR) was 106 (95% confidence interval (95% CI) 93.7 to 118) and for women 83.8 (95% CI 53.1 to 126). The SMRs for total cancer mortality were 125 (95% CI 98.2 to 157) for men and 69.5 (95% CI 25.5 to 151) for women. The SMR for lung cancer in men was 176 (95% CI 101 to 287). No lung cancers were found among female workers. Up to 31 December, 1991, a total of 118 cancers had occurred in the cohort. A significantly increased standardised incidence ratio (SIR) was found for cancer of the nose and nasal sinuses in men, three cases v 0.36 expected, yielding an SIR of 832 (95% CI 172 to 2430). Applying a 10 year latency period in cohort members exposed to > or = 1000 micrograms cadmium/m3, the SIR was 1107 (95% CI 134 to 4000). Similarly, for cohort members exposed to 2000 micrograms nickel/m3, the SIR was 1080 (95% CI 131 to 3900). CONCLUSION: There was an increased overall risk for lung cancer, but no exposure-response relation between cumulative exposure to cadmium or nickel and risk of lung cancer. There was a highly significant increased risk of cancer of the nose and nasal sinuses, which may be caused by exposure to nickel or cadmium or a combination of both exposures.
This study aimed to investigate the association between daily air pollution levels and the occurrence of acute respiratory signs and symptoms among people with asthma or asthma-like problems. Thirty eight subjects in the southern Swedish city of Landskrona kept a daily diary for 10 weeks. The daily prevalence of symptoms, supplementary bronchodilator use and peak flow deviations were compared with measurements of environmental nitrogen dioxide (NO2), sulphur dioxide, temperature and humidity in the city. The occurrence of severe asthma, both during the day and during the evening, was significantly positively associated with the concurrent 24 h average concentration of NO2, which never exceeded 72 microg x m(-3). A correlation of borderline significance was found between the use of on-demand medication and the NO2 level. However, peak flow deviations were not associated with air pollution or weather conditions, which may be explained by the beneficial effect of bronchodilators used by 28 of the subjects. The results of this study confirm those of some earlier studies and suggest that aggravation of asthma is related to daily variations in air quality, as indicated by relatively low ambient concentrations of nitrogen dioxide. These results also indicate that it may be appropriate to examine severe asthma symptoms separately.
BACKGROUND: Pertussis vaccination in infancy has been suggested to increase the risk for development of asthma and allergy. OBJECTIVE: To assess sensitization rates and development of atopic diseases in a prospective randomized controlled trial of pertussis vaccine. PATIENTS AND METHODS: A total of 669 children were randomized to 1 of 4 vaccine groups (2-component acellular pertussis, 5-component acellular pertussis, whole-cell pertussis vaccines, and placebo [diphtheria and tetanus toxoids]). Diphtheria and tetanus toxoids were also given to the children in the pertussis vaccine groups. The children were evaluated by means of questionnaires at age 2 months, 7 months, and 2 1/2 years; skin prick tests at age 7 months and 2 1/2 years; and blinded clinical investigation at age 2 1/2 years. The families were contacted at regular intervals to assess possible adverse effects after the vaccinations and symptoms of whooping cough. RESULTS: The cumulative incidence of atopic diseases was 30% and incidence rates were similar in the 4 groups after adjusting for family history. Exposure to environmental tobacco smoke and home dampness did not confound these results. The frequency of adverse effects did not differ appreciably between atopic and nonatopic children, with the exception that a nodule at the vaccination site was more frequent after whole-cell pertussis vaccination in the nonatopic children. Among 47 children with proven pertussis, atopic disease appeared in 19 (40%). Of these 47 children, 9 (19%) developed asthma, as compared with 58 (9%) noninfected children (P=.03). CONCLUSIONS: We found no support for a drastic increase in allergic manifestations after pertussis vaccination. There was a positive association between whooping cough and asthma by 2 1/2 years of age. There seems to be little reason to withhold pertussis vaccination from infants, irrespective of family history of allergy.
BACKGROUND: Poor indoor air quality has been suggested to be related to the increase in the prevalence of asthma that has occurred in the western world, especially among children and young persons. Apart from the home, school is the most important indoor environment for children. OBJECTIVES: The aims were to study the prevalence of current asthma among secondary pupils and its relationship to the school environment, but also to personal factors and domestic exposures. METHODS: Data on asthmatic symptoms, other health aspects, and domestic exposures were gathered using a questionnaire which was sent to 762 pupils in the seventh form (13-14 years old) in 11 randomly chosen schools in the county of Uppsala in Sweden. Pupils answering 'yes' to having had asthma diagnosed by a physician, and having had recent asthma attacks, or who used asthma medication were defined as having current asthma. Data on exposures at school were gathered by measurements in 28 classrooms. The relationship between asthma and exposures was analysed by multiple logistic regression. RESULTS: The questionnaire was completed by 627 (82%). Current asthma was found among 40 pupils (6.4%). Current asthma was more common in those who had an atopic disposition, or food allergy, or who had attended a day care centre for several years. Controlling for these factors, current asthma was related to several factors in the school environment. There were more pupils with current asthma in schools that were larger, had more open shelves, lower room temperature, higher relative air humidity, higher concentrations of formaldehyde or other volatile organic compounds, viable moulds or bacteria or more cat allergen in the settled dust. CONCLUSIONS: Although the pupils attended school for a minor part of their time, our study indicates that the quality of the school environment is of importance and may affect asthmatic symptoms.
OBJECTIVE: To study the association of the daily frequency of registration of patients with acute asthma at the emergency department of a central hospital in the south-west of Sweden with levels of air pollution and meteorological observations. METHODS: A retrospective longitudinal study was made of asthma patients taken from a hospital registry. This information was correlated with measurements of ozone, nitrogen dioxide, sulphur dioxide, toluene, temperature and relative humidity. Patients were from the catchment area of the Central Hospital of Halmstad containing around 120,000 inhabitants. A total of 4127 visits of patients with acute asthma to the emergency department at the Central Hospital of Halmstad were registered during a period of 1247 days from January 1990 to May 1993. The differential optical absorption spectroscopy (DOAS) technique was used to monitor levels of air pollutants over a distance of 1000 m in the central part of the town of Halmstad. Data on temperature, relative humidity, precipitation, wind speed and wind direction for the time period were supplied by the Swedish Meteorological and Hydrological Institute (SMHI). RESULTS: There were many statistically significant correlations between the levels of air pollutants and the meteorological measurements and a strong negative correlation between ozone and nitrogen dioxide. There was a statistically significant effect on asthma visits in children of low temperature and high nitrogen dioxide levels, and on asthma visits in adults of high temperature and high levels of ozone. CONCLUSIONS: There was a different reaction pattern in children and adults with asthma regarding temperature and ozone/nitrogen dioxide. The strong correlations between temperature and air pollution and between the levels of ozone and nitrogen dioxide made the true relation between asthma, air pollution and temperature hard to evaluate statistically.
Does the prevalence of respiratory symptoms differ between regions? As a part of the European Community Respiratory Health Survey, we present data from an international questionnaire on asthma symptoms occurring during a 12 month period, smoking and symptoms of chronic bronchitis. The questionnaire was mailed to 10,800 persons aged 20-44 yrs living in three regions of Sweden (Västerbotten, Uppsala and Göteborg) with different environmental characteristics. The total response rate was 86%. Wheezing was reported by 20.5%, and the combination of wheezing without a cold and wheezing with breathlessness by 7.4%. The use of asthma medication was reported by 5.3%. Long-term cough and/or morning cough together with problems with phlegm was reported by 12.8%; the prevalence being highest in the most polluted area (Göteborg). When using multivariate analysis, no significant difference in asthma-related symptoms was found between the centres. Women reported cough more frequently, but otherwise gender did not influence symptom prevalence. Our results indicate that bronchitis symptoms occur more frequently in Göteborg, the most polluted of the Swedish centres, but that the prevalence rates of asthma-related symptoms do not differ between these three regions.
Allergic sensitization and symptoms from the airways in relation to air pollution were compared in 10-12-year-old school children (n = 1113) from urban Konin in central Poland and both urban and rural parts of Sundsvall in northern Sweden. The measurements included parental questionnaires, skin-prick tests and serial peak flow measurements during 2 weeks with simultaneous monitoring of outdoor air pollutants. The skin-prick test technique was validated by IgE antibody determinations. The levels of common industrial pollutants, SO2 and smoke particles were much higher in Konin than in urban Sundsvall and the levels of NO2 were similar. Various respiratory symptoms were more often reported among school children in Konin (except for wheezing and diagnosed asthma). Multiple logistic regression analyses yielded the following increased odds ratios for children in Konin as compared with the reference group (rural Sundsvall): chest tightness and breathlessness 3.48 (95% confidence interval 2.08-5.82), exercise-induced coughing attacks 3.69 (95% confidence interval 1.68-8.10), recurrent episodes of common cold 2.79 (95% confidence interval 1.53-5.09) and prolonged cough 4.89 (95% confidence interval 2.59-9.23). In contrast, as compared with rural Sundsvall, the adjusted odds ratio for a positive skin-prick test was decreased in Konin, but increased in urban Sundsvall, 0.58 (95% confidence interval 0.37-0.91) and 1.67 (95% confidence interval 1.15-2.42) respectively. The study confirms that living in urban, as compared with rural areas, is associated with an increased prevalence of respiratory symptoms and sensitization to allergens. These differences could be explained by air pollution. Respiratory symptoms were more common in a similar urban group of Polish children who were exposed to even higher levels of air pollution. These children, however, had a much lower prevalence of sensitization to allergens, as compared with the Swedish children. This indicates that differences in lifestyle and standard of living between western Europe and a former socialist country influences the prevalence of atopy.