Biometrical modeling of twin and family data using standard mixed model software.

Biometrical genetic modeling of twin or other family data can be used to decompose the variance of an observed response or 'phenotype' into genetic and environmental components. Convenient parameterizations requiring few random effects are proposed, which allow such models to be estimated using widely available software for linear mixed models (continuous phenotypes) or generalized linear mixed models (categorical phenotypes). We illustrate the proposed approach by modeling family data on the continuous phenotype birth weight and twin data on the dichotomous phenotype depression. The example data sets and commands for Stata and R/S-PLUS are available at the Biometrics website.

A simple correction for ties when censoring times depend on covariates.

We point out that the conventional methods for ties correction may be seriously biased when censoring times depend on covariates. A simple modification to the Efron correction method is suggested which works remarkably well in simulation studies. The method corresponds closely to breaking ties by random ordering. The modified correction method is easy to implement and computationally no more demanding than the Efron correction.

Day care center characteristics and children's respiratory health.

UNLABELLED: Day care centers provide an important exposure arena with potential harmful health effects for children. This study has linked health effect data from a survey among 942 3-5-year-old Oslo children with information on day care center characteristics collected during inspection of the 175 day care centers these children attended. The aim of the study was to estimate associations between dampness problems and other building characteristics and several respiratory health outcomes. Dampness problems (sign of molds, water leakage, damage to floor/wall) were observed in 51% of the day care centers. In multiple logistic regression analyses none of the studied symptoms and diseases (nightly cough, blocked or runny nose without common cold, wheeze, heavy breathing or chest tightness, the common cold, tonsillitis/pharyngitis, otitis media, bronchitis, pneumonia, asthma, and allergic rhinitis) were systematically associated with dampness problems or type of ventilation in day care centers. None of the studied indicators of day care center exposures were found to have a clear effect on day care children's respiratory health. Even so this study does not rule out negative health effects of day care center exposures. The study demonstrates that population-based studies of these relations are demanding with regard to assessment of exposure and health outcomes. PRACTICAL IMPLICATIONS: Simple and easy-to-register indicators of exposures like dampness problems and type of ventilation assessed in 175 day care centers were not related to respiratory health among 3-5-year-old Norwegian children attending the day care centers. The study does not rule out negative health effects of day care center exposures, but demonstrates methodological challenges needed to be addressed in studies of health effects of the day care environment.

Depression during pregnancy and after delivery: a repeated measurement study.

The aim of this study was to examine the risk of depression in the postpartum period (first four months after delivery) as compared to the remaining postnatal year and the pregnancy period. All postpartum women from two municipalities in Norway were included in a questionnaire study of mental health (n = 416). Over 50% of the women (n = 259) answered an identical questionnaire at an additional time either before or after the postpartum period. The level of depression was measured by the Edinburgh Postnatal Depression Scale (EPDS) and the Hopkins Symptom Check List-25 items (SCL-25). The point prevalence of depression (EPDS> or =10) in the first four months postpartum did not differ significantly as compared to other time periods during pregnancy and the postnatal year. This finding remained also after controlling for other risk factors of depression; high score on the life event scale, prior depression and poor partner relationship. There was a non-significant trend of lower prevalence of depression during early pregnancy and after the first eight postnatal months. In conclusion, our findings suggest that the first four months postpartum were not distinguished by higher depression prevalence as compared to other time periods during pregnancy and the first postnatal year.

Paternal contribution to birth weight.

STUDY OBJECTIVE: Understanding causes of variation in birth weight has been limited by lack of sufficient sets of data that include paternal birth weight. The objective was to estimate risks of low birth weight dependent on parental birth weights and to estimate father-mother-offspring correlations for birth weight to explain the variability in birth weight in terms of effects of genes and environmental factors. DESIGN: A family design, using trios of father-mother-firstborn child. SETTING: The complete birth population in Norway 1967-98. PARTICIPANTS: 67 795 families. MAIN RESULTS: The birth weight correlations were 0.226 for mother-child and 0.126 for father-child. The spousal correlation was low, 0.020. The relative risk of low birth weight in the first born child was 8.2 if both parents were low birth weight themselves, with both parents being above 4 kg as the reference. The estimate of heritability is about 0.25 for birth weight, under the assumption that cultural transmission on the paternal side has no effect on offspring prenatal growth. CONCLUSIONS: Paternal birth weight is a significant and independent predictor of low birth weight in offspring. The estimate of the heritability of birth weight in this study is lower than previously estimated from data within one generation in the Norwegian population.

Foetal growth and duration of gestation relative to water chlorination.

OBJECTIVE: To assess the effect of exposure to chlorination byproducts during pregnancy on foetal growth and duration of pregnancy. METHODS: A population based study was conducted of 137,145 Norwegian children born alive in 1993--5. Information was obtained from the Norwegian medical birth registry, waterwork registry, and social science data service. The outcomes of interest were birth weight, low birth weight (<2500 g), small for gestational age, and preterm delivery (gestational age <37 weeks). The exposure assessment was based on quality of drinking water in the municipality where the mother lived during pregnancy. Municipal exposure was calculated with information on chlorination and the amount of natural organic matter in raw water measured as colour in mg precipitate/l. The main exposure category was high colour and chlorination, which was contrasted with the reference category of low colour and no chlorination. RESULTS: In logistic regression analysis adjusting for confounding, the risks of low birth weight (odds ratio (OR) 0.97, 95% confidence interval (95% CI) 0.89 to 1.06) and small for gestational age (OR 1.00, 95% CI 0.91 to 1.10) were not related to exposure. Contrary to the hypothesis, the risk of preterm delivery was slightly lower among the exposed than the reference category (OR 0.91, 95% CI 0.84 to 0.99). The risks of the studied outcomes were similar in newborn infants exposed to high colour drinking water without chlorination and chlorinated drinking water with low colour compared with the reference category. CONCLUSIONS: The present study did not provide evidence that prenatal exposure to chlorination byproducts at the relatively low concentrations encountered in Norwegian drinking water increases the risk of the studied outcomes.

[Day care centers, infections and asthma]. / Barnehage, infeksjoner og astma.

BACKGROUND: Our objective was to use a causal model for childhood asthma to determine whether the effect of day care attendance on asthma was mediated by recurrent respiratory tract infections. MATERIAL AND METHODS: The study is based on a cross-sectional survey among 1,447 children aged 6-16 years in Oslo. Their parents completed written questionnaires. A recursive logit model was used to estimate direct effects in terms of adjusted odds ratios (aOR). RESULTS: Year of birth, number of siblings and length of maternal education were significantly associated with after-school care attendance. Attendance increased the risk of early infections, aOR = 1.8 (1.3-2.5), and infections were associated with asthma, aOR = 4.9 (3.4-7.3). The crude association between after-school care and asthma was cOR = 1.5 (1.0-2.2), whereas the estimated direct effect was small and non-significant, aOR = 1.2 (0.8-1.9). The results may be influenced by over-reporting of infections among parents with children with asthma. INTERPRETATION: Our results suggest that children who attend day care have an increased risk of asthma, with early infections as a mediator of risk.

Mortality and temperature in Oslo, Norway, 1990-1995.

The associations between temperature and daily mortality was studied among the citizens of Oslo, Norway, 1990-1995. Data on daily mortality were linked with daily temperatures, relative humidity, wind velocity and air pollution. At temperatures below 10 degrees C, a 1 degrees C fall in the last 7 days average temperature increased the daily mortality from all diseases by 1.4%, respiratory diseases 2.1%, and cardiovascular diseases 1.7%. Above 10 degrees C, there was no statistically significant increase in daily mortality, except for respiratory mortality, which increased by 4.7% per 1 degrees C increase in the last 7 days average temperature. Daily mortality in Oslo increases with temperatures falling below 10 degrees C. The increase starts at lower temperatures than shown in warmer regions of the world, but at higher temperatures than in regions with even colder climates. As well insulated and heated dwellings are standard in Norway today, more adequate clothing during outdoor visits is probably the most important preventive measure for temperature related mortality.

Parameterization of multivariate random effects models for categorical data.

Alternative parameterizations and problems of identification and estimation of multivariate random effects models for categorical responses are investigated. The issues are illustrated in the context of the multivariate binomial logit-normal (BLN) model introduced by Coull and Agresti (2000, Biometrics 56, 73-80). We demonstrate that the BLN model is poorly identified unless proper restrictions are imposed on the parameters. Moreover, estimation of BLN models is unduly computationally complex. In the first application considered by Coull and Agresti, an identification problem results in highly unstable, highly correlated parameter estimates and large standard errors. A probit-normal version of the specified BLN model is demonstrated to be underidentified, whereas the BLN model is empirically underidentified. Identification can be achieved by constraining one of the parameters. We show that a one-factor probit model is equivalent to the probit version of the specified BLN model and that a one-factor logit model is empirically equivalent to the BLN model. Estimation is greatly simplified by using a factor model.

Associations between outdoor air pollutants and hospitalization for respiratory diseases.

The concentration of particulate matter in outdoor air, as indicated by daily measures of particulate matter 10 microm in diameter (PM10) in many cities, has been found to be associated with the daily number of deaths and hospital admissions in these cities. To understand this association better, we studied the daily number of hospital admissions for respiratory diseases and the concentrations of eight pollutants in ambient air, during a period of 38 months, in an area with cold winters and air pollution that comes mainly from motor vehicles. We estimated the changes in risk of hospitalization by interquartile increases in pollutant concentrations by Poisson regression analyses. Controlling for periodic trends and weather, the relative risk of hospitalization associated with an interquartile increase of PM10 was 1.038 [95% confidence interval (CI) = 0.991-1.087]. In contrast, the relative risk associated with benzene was 1.105 (95% CI = 1.047-1.166). In a two-pollutant model, the relative risk estimates were 1.014 (95% CI = 0.966-1.063) for PM10 and 1.090 (95% CI = 1.031-1.153) for benzene. We evaluated other two- and three-pollutant models and concluded that pollutants other than PM10 are more strongly associated with hospital admissions for respiratory diseases.

Cesarean section among immigrants in Norway.

OBJECTIVE: We studied prevalences and risk factors for cesarean section among different groups of immigrants from countries outside Western Europe and North America in comparison to ethnic Norwegians. METHODS: The study is population based using data from the Medical Birth Registry of Norway. A total of 553,491 live births during the period 1986-1995 were studied, including 17,891 births to immigrant mothers. RESULTS: The prevalences of cesarean section ranged from 10.1% among women from Vietnam to 25.8% in the group of Filipino origin. The use of abdominal delivery was also high in the groups from Sri Lanka/India (21.3%), Somalia/Eritrea/Ethiopia (20.5%) and Chile/Brazil (24.3%), while the frequency among women from Turkey/Morocco (12.6%) and Pakistan (13.2%) was approximately the same as among ethnic Norwegians (12.4%). Feto-pelvic disproportion, fetal distress and prolonged labor were the most important diagnoses associated with the high prevalences, but the significance of these diagnoses differed among the groups. Other unknown factors come into play, particularly among women from Somalia/Eritrea/Ethiopia and Chile/Brazil. CONCLUSION: There was substantial variation in the use of cesarean section among ethnic groups in Norway. The diagnoses feto-pelvic disproportion, fetal distress and prolonged labor may be confounded by a number of factors including maternal request for cesarean section and difficulties in handling the delivery. Further research is needed to explain the observed differences.

Changes in condom use after HIV diagnosis.

We investigated changes in condom use after HIV diagnosis. The study population comprised 78 asymptomatic HIV-infected subjects recruited from a clinic in Oslo, Norway, during 1988-92. In total, 240 follow-up visits were carried out. The response variable was repeated measurements of whether or not condoms were always used during anal/vaginal intercourse at follow-up. The explanatory variables were: time from HIV diagnosis to follow-up, exposure group (heterosexually infected, homosexual men, or infected through injecting drug use), time of HIV diagnosis (before 1987 or later), and history of condom use before HIV diagnosis. Random effects logistic regression analysis was used to study incremental changes in condom use, accommodating an unbalanced repeated measurement design. The use of condoms increased monotonously over time after HIV diagnosis in all exposure groups. Condom use was least likely among injecting drug users, subjects diagnosed before 1987 and subjects without a history of frequent condom use before HIV diagnosis.

Herpes simplex virus type-2 antibodies in pregnant women: the impact of the stage of pregnancy.

In this study the impact of pregnancy duration on the measured level of HSV-2 antibodies was assessed. The study population comprised 35,940 pregnant women in Norway, in 1992-4, followed during pregnancy. A random sample of 960 women was selected. A mean of 2.6 serum samples from each woman were analysed for HSV-2 specific IgG antibodies at different times in pregnancy. Crude and adjusted odds ratios were estimated in logistic regression models taking all observations per women into account. Twenty-seven percent of the pregnant women had antibodies against HSV-2 in the first trimester. The adjusted odds ratio of being HSV-2 antibody positive decreased during the pregnancy and was 0.5 (0.2-0.9, 95% confidence interval) in the 40th as compared to the 10th week of pregnancy. About 50% of initially HSV-2 positive women did not have detecable antibodies by the end of the pregnancy. This may be explained by haemodilution during pregnancy. Our findings have diagnostic implications and should encourage further studies.

Design and Analysis of Monte Carlo Experiments: Attacking the Conventional Wisdom.

The design and analysis of Monte Carlo experiments, with special reference to structural equation modelling, is discussed in this article. These topics merit consideration, since the validity of the conclusions drawn from a Monte Carlo study clearly hinges on these features. It is argued that comprehensive Monte Carlo experiments can be implemented on a PC if the experiments are adequately designed. This is especially important when investigating modern computer intensive methodologies like resampling and Markov Chain Monte Carlo methods. We are faced with three fundamental challenges in Monte Carlo experimentation. The first problem is statistical precision, which concerns the reliability of the obtained results. External validity, on the other hand, depends on the number of experimental conditions, and is crucial for the prospects of generalising the results beyond the specific experiment. Finally, we face the constraint on available computer resources. The conventional wisdom in designing and analysing Monte Carlo experiments embodies no explicit specification of meta-model for analysing the output of the experiment, the use of case studies or full factorial designs as experimental plans, no use of variance reduction techniques, a large number of replications, and "eyeballing" of the results. A critical examination of the conventional wisdom is presented in this article. We suggest that the following alternative procedures should be considered. First of all, we argue that it is profitable to specify explicit meta-models, relating the chosen performance statistics and experimental conditions. Regarding the experimental plan, we recommend the use of incomplete designs, which will often result in considerable savings. We also consider the use of common random numbers in the simulation phase, since this may enhance the precision in estimating meta-models. The use of fewer replications per trial, enabling us to investigate an increased number of experimental conditions, should also be considered in order to improve the external validity at the cost of the conventionally excessive precision.

Day care attendance, recurrent respiratory tract infections and asthma.

OBJECTIVE: Our objective was to use a causal model for childhood asthma to determine whether the effect of day care attendance on asthma was mediated by recurrent respiratory tract infections. DESIGN: A cross-sectional survey among 1447 children aged 6-16 years in Oslo. Their parents completed written questionnaires. A recursive logit model was used to estimate direct effects in terms of adjusted odds ratios (aOR). RESULTS: Year of birth, number of siblings and length of maternal education were significantly associated with day care attendance. Attendance at day care increased the risk of early infections, aOR = 1.8 (1.3-2.5), and infections were associated with asthma, aOR = 4.9 (3.4-7.3). The crude association between day care and asthma was cOR = 1.5 (1.0-2.2), whereas the estimated direct effect was small and nonsignificant, aOR = 1.2 (0.8-1.9). The results may be influenced by overreporting of infections among parents of children with asthma. CONCLUSIONS: Our results suggest that children who attend day care have an increased risk of asthma with early infections as a mediator of risk.

Adrenaline-mediated glycogen phosphorylase activation is enhanced in rat soleus muscle with increased glycogen content.

The effect of glycogen content on the activation of glycogen phosphorylase during adrenaline stimulation was investigated in soleus muscles from Wistar rats. Furthermore, adrenergic activation of glycogen phosphorylase in the slow-twitch oxidative soleus muscle was compared to the fast-twitch glycolytic epitrochlearis muscle. The glycogen content was 96.4 +/- 4.4 mmol (kg dw)(-1) in soleus muscles. Three hours of incubation with 10 mU/ml of insulin (and 5.5 mM glucose) increased the glycogen content to 182.2+/-5.9 mmol (kg dw)(-1) which is similar to that of epitrochlearis muscles (175.7+/-6.9 mmol (kg dw)(-1)). Total phosphorylase activity in soleus was independent of glycogen content. Adrenaline (10(-6) M) transformed about 20% and 35% (P < 0.01) of glycogen phosphorylase to the a form in soleus with normal and high glycogen content, respectively. In epitrochlearis, adrenaline stimulation transformed about 80% of glycogen phosphorylase to the a form. Glycogen synthase activation was reduced to low level in soleus muscles with both normal and high glycogen content. In conclusion, adrenaline-mediated glycogen phosphorylase activation is enhanced in rat soleus muscles with increased glycogen content. Glycogen phosphorylase activation during adrenaline stimulation was much higher in epitrochlearis than in soleus muscles with a similar content of glycogen.

Water chlorination and birth defects.

Chlorination of drinking water that contains organic compounds leads to the formation of by-products, some of which have been shown to have mutagenic or carcinogenic effects. As yet, too little is known about the possible teratogenic effects on the human fetus. We linked the Norwegian waterwork registry, containing 1994 data on chlorination practice and color (an indicator for natural organic matter), with the Medical Birth Registry for 1993-1995. The proportion of the population exposed to chlorination and a weighted mean color number in drinking water was computed for each municipality. Among 141,077 births, 2,608 (1.8%) had birth defects. In a comparison between exposed (high color; chlorination) and reference groups (low color; no chlorination), the adjusted odds ratio was 1.14 (0.99-1.31) for any malformation, 1.26 (0.61-2.62) for neural tube defects, and 1.99 (1.10-3.57) for urinary tract defects. This study provides further evidence of the role of chlorination of humic water as a potential cause of birth defects, in a country with relatively low levels of chlorination byproducts.

The problem of structural indeterminacy in multidimensional symptom report instruments. The case of SCL-90-R.

The factor structure of SCL-90-R items and scales was analyzed using both exploratory factor analysis (EFA) and confirmatory factor analysis (CFA). Results of CFA studies at the item-level failed to support the original nine-factor model, as well as several alternative models and EFA suggested very different dimensionality, depending on which criteria were used. Analyses at the scale-level (i.e. the nine original symptom dimensions) suggested that a one- or two-factor model was satisfactory according to descriptive goodness of fit criteria. However, using the likelihood ratio test, specification of four factors was necessary to avoid rejection. According to the likelihood ratio test in a multi-group analysis, a lack of factorial invariance across gender was indicated. Moreover, the factorial structure of the instrument was clearly different across levels of negative affectivity (NA); the dimensionality was substantially higher in the low-NA group as compared to the high-NA group. It is concluded that we are confronted with a profound structural indeterminacy problem and that factor analytic methods and model acceptance criteria alone are insufficient to solve this problem. The indeterminacy problem can be accounted for, at least in part, by the complex logical-semantical structure of SCL-90-R items and scales and the role of the NA trait as a structure generating factor.

Consanguinity and recurrence risk of stillbirth and infant death.

OBJECTIVES: The aim of this study was to estimate the recurrence risk for stillbirth and infant death and compare results for offspring of first-cousin parents with results for offspring of unrelated parents. METHODS: The study population consisted of all single births with a previous sibling born in Norway between 1967 and 1994. Altogether, 629,888 births were to unrelated parents, and 3466 births were to parents who were first cousins. The risk of stillbirth and infant death was estimated for subsequent siblings contingent on parental consanguinity and survival of the previous sibling. RESULTS: For unrelated parents, the risk of early death (stillbirth plus infant death) for the subsequent sibling was 17 of 1000 if the previous child survived and 67 of 1000 if the previous child died before 1 year of age. For parents who were first cousins, the risk of early death for the subsequent sibling was 29 of 1000 if the previous child survived and 116 of 1000 if the previous child died. CONCLUSIONS: The risk of recurrence of stillbirth and infant death is higher for offspring of first-cousin parents compared with offspring of unrelated parents.

Consanguinity and recurrence risk of birth defects: a population-based study.

Recurrence risks give insight into the causes of birth defects and are useful in genetic counseling. There are few population-based studies of recurrence of birth defects for subsequent sibs with consanguineous parents. The aim of this study was to estimate and compare the recurrence risk of birth defects for offspring of first cousins and nonconsanguineous parents. The study population consisted of all single births with a previous sib born in Norway between 1967 and 1995. Altogether 660,398 children had nonconsanguineous parents, and 3,583 had parents who were first cousins. For nonconsanguineous parents the risk of a birth defect for the subsequent sib was 15 per 1,000 births (95% confidence interval: 14.5-15.1) if the previous child did not have a birth defect and 33 (95% confidence interval: 30-37) if the previous child had a birth defect. For parents who were first cousins the risk of a birth defect for the subsequent sib was 36 per 1,000 (95% confidence interval: 30-42) if the previous child did not have a birth defect and 68 (95% confidence interval: 33-122) if the previous child had a birth defect. The risk of recurrence of birth defects is higher for subsequent sibs with first-cousin parents than for those with nonconsanguineous parents. This difference indicates the degree to which the increased homozygosity among offspring of consanguineous parents influences the risk of recurrence of birth defects.