Kjell Haug

Self-selection and bias in a large prospective pregnancy cohort in Norway.

Self-selection in epidemiological studies may introduce selection bias and influence the validity of study results. To evaluate potential bias due to self-selection in a large prospective pregnancy cohort in Norway, the authors studied differences in prevalence estimates and association measures between study participants and all women giving birth in Norway. Women who agreed to participate in the Norwegian Mother and Child Cohort Study (43.5% of invited; n = 73 579) were compared with all women giving birth in Norway (n = 398 849) using data from the population-based Medical Birth Registry of Norway in 2000-2006. Bias in the prevalence of 23 exposure and outcome variables was measured as the ratio of relative frequencies, whereas bias in exposure-outcome associations of eight relationships was measured as the ratio of odds ratios. Statistically significant relative differences in prevalence estimates between the cohort participants and the total population were found for all variables, except for maternal epilepsy, chronic hypertension and pre-eclampsia. There was a strong under-representation of the youngest women (<25 years), those living alone, mothers with more than two previous births and with previous stillbirths (relative deviation 30-45%). In addition, smokers, women with stillbirths and neonatal death were markedly under-represented in the cohort (relative deviation 22-43%), while multivitamin and folic acid supplement users were over-represented (relative deviation 31-43%). Despite this, no statistically relative differences in association measures were found between participants and the total population regarding the eight exposure-outcome associations. Using data from the Medical Birth Registry of Norway, this study suggests that prevalence estimates of exposures and outcomes, but not estimates of exposure-outcome associations are biased due to self-selection in the Norwegian Mother and Child Cohort Study.

Self-reported smoking status and plasma cotinine concentrations among pregnant women in the Norwegian Mother and Child Cohort Study

Introduction:Underreporting of smoking in epidemiologic studies is common and may constitute a validity problem, leading to biased association measures. In this prospective study, we validated self-reported tobacco use against nicotine exposure assessed by plasma cotinine in the Norwegian Mother and Child Cohort Study (MoBa).Methods:The study was based on a subsample of 2,997 women in the MoBa study who delivered infants during the period 2002–2003. Self-reported tobacco use (test variable) and plasma cotinine concentrations (gold standard) were assessed at approximately gestational week 18.Results:Daily smoking was reported by 9% of the women, occasional smoking by 4%, and nonsmoking by 86% of the women. Sensitivity and specificity for self-reported smoking status were calculated using a cotinine cut-off estimated from the study population (30 nmol/l). Plasma cotinine concentrations ≥30 nmol/l were found in 94% of self-reported daily smokers, 66% of occasional smokers, and 2% of nonsmokers. After the numbers of self-reported nonsmokers with cotinine concentrations above the cut-off limit were added, the daily smoking prevalence increased from 9 to 11%. The sensitivity and specificity for self-reported daily smoking, using 30 nmol/l as the cut-off concentration, were 82 and 99%, respectively.Discussion:These findings suggest that self-reported tobacco use is a valid marker for tobacco exposure in the MoBa cohort.

Maternal smoking and birthweight: effect modification of period, maternal age and paternal smoking

Objective. To study the effect on birthweight of maternal smoking, and its modification by study period, maternal age and paternal smoking.

Epidemiology of Ulcerative Colitis in Western Norway

The incidence of ulcerative colitis (UC) in the three counties that compose Western Norway was registered in a prospective study during the years 1984 and 1985. Both inpatients and outpatients were included. Five hundred and fifty general practitioners and 12 hospitals participated in the study. A total of 239 patients were diagnosed in this area with 807,000 inhabitants, giving a mean annual incidence of 14.8 per 100,000. For patients between 30 and 35 years of age the age-specific incidence rate was 31.2. The M/F sex ratio was 1.10. Familial occurrence of inflammatory bowel disease was found in 11% of the patients. Patients with rectal involvement only constituted 35% of the patients.

Long-term sickness absence and disability pension with psychiatric diagnoses: A population-based cohort study

Sickness absence certified with psychiatric diagnoses is increasing in many Western countries. A substantial proportion of the sickness absentees never return to work, but ends up with a permanent disability pension (DP). This study investigated the incidence of long-term sickness absence (LTSA) with different psychiatric diagnoses, and tested predictors of the transition to permanent DP. A special objective was to explore previously reported gender differences, using a population-based cohort study. The population at risk of LTSA was 106,674 occupational active men and 89,356 women in a Norwegian county in 1994; 314 women and 203 men with LTSA >8weeks, certified with a psychiatric diagnosis, were followed to the end of 1999, with DP as the endpoint. Diagnoses on sickness certificates, age, gender and income were used as explanatory variables in Cox regression analysis. Annual incidence of LTSA with a psychiatric diagnosis was 7.0/1000/year for women and 3.8/1000 for men; 72% of the women and 50% of the men had a diagnosis indicating depression. During follow-up, 32% of the men and 25% of the women obtained DP. Increasing age, male gender, low income and a diagnosis of psychosis or “other” increased the DP risk. Separate analysis for men indicated a different effect of age and a larger role of serious mental illness. The study verified that women more often than men had LTSA with psychiatric diagnoses, especially with depression. Men were at higher risk of transition to DP, and the study suggested some possible explanations.

Chronic Inflammatory Bowel Disease in Children in Western Norway

The incidence of Crohns disease (CD) and ulcerative colitis (UC) in children in western Norway was estimated in a prospective epidemiological study during the years 1984 and 1985. The total population in the area was 807,000 and the child population was 198,570 (1984). There were 27 new cases of chronic inflammatory bowel disease (IBD) in children aged 15 years or less, 10 new cases of CD, and 17 of UC. The mean annual incidence of CD in the child population was 2.5/100,000/year, whereas the incidence of UC in the child population was 4.3/100,000/year. Nearly all the children had abdominal symptoms. In this study, we found an incidence of CD in children that is the highest hitherto reported, to our knowledge. To the contrary, the incidence of UC was considerably lower than previously reported from northern Europe.

In utero exposure to maternal tobacco smoke and subsequent obesity, hypertension, and gestational diabetes among women in the MoBa Cohort

Background: Environmental factors influencing the developmental origins of health and disease need to be identified and investigated. In utero exposure to tobacco smoke has been associated with obesity and a small increase in blood pressure in children; however, whether there is a corresponding increased risk of conditions such as diabetes and hypertension during adulthood remains unclear. Objective: Our goal was to assess the association of self-reported in utero exposure to tobacco smoke with the prevalence of obesity, hypertension, type 2 diabetes mellitus (T2DM), and gestational diabetes mellitus (GDM) in women 14–47 years of age. Methods: We conducted a cross-sectional analysis of the Norwegian Mother and Child Cohort Study, which enrolled pregnant women in Norway from 1999 thorough 2008. Exposure to tobacco smoke in utero (yes vs. no) was ascertained on the baseline questionnaire (obtained at ~ 17 weeks’ gestation); the outcomes were ascertained from the Medical Birth Registry of Norway and the questionnaire. Our analysis included 74,023 women. Results: Women exposed to tobacco smoke in utero had 1.53 times the odds of obesity [95% confidence interval (CI): 1.45, 1.61] relative to those unexposed, after adjusting for age, education, and personal smoking. After further adjustment for body mass index, the odds ratio for hypertension was 1.68 (95% CI: 1.19, 2.39); for T2DM 1.14 (95% CI: 0.79, 1.65); and for GDM 1.32 (95% CI: 1.10, 1.58) among exposed compared with unexposed. Conclusions: Exposure to tobacco smoke in utero was associated with obesity, hypertension, and GDM in adult women. The possibility that the associations were attributable to unmeasured confounding cannot be excluded.

Epidemiology of Crohn's Disease in Western Norway

The incidence of Crohns disease in western Norway was estimated in a prospective epidemiologic study during the years 1984 and 1985. The total population in the area was 807,000. Both inpatients and outpatients were included. Five hundred and fifty general practitioners and 12 hospitals participated in the study. A total of 86 patients were diagnosed, giving a mean annual incidence of 5.3 per 100,000. For patients between 15 and 19 years of age-specific incidence rate was 16.0 The M/F sex ratio was 0.9. In the same period 240 patients with ulcerative colitis were diagnosed (mean annual incidence, 14.8 per 100,000), giving a ratio of 2.8 between ulcerative colitis and Crohns disease and an annual incidence of 20.1 per 100,000 for inflammatory bowel disease. Familial occurrence of IBD was found for 12% of the patients.

Smoking habits among pregnant women in Norway 1994-95

AIMS To investigate the smoking prevalence the last three months before pregnancy and at 18 weeks of gestation among women in Norway and to evaluate the impact of pre-pregnancy smoking habits, maternal age, level of education, civil status and parity on smoking cessation. MATERIAL AND METHODS A prospective, multicenter survey. The study population included 4 766 pregnant women who attended a routine ultrasound examination at 18 weeks of pregnancy in six Norwegian hospitals during the period from September 1994 to March 1995. Smoking habits before and during pregnancy were recorded. RESULTS The point prevalence of self-reported daily smoking among the women three months before the pregnancy was 34%. At 18 weeks of pregnancy, 21% of the women reported smoking daily (p<0.001). A multiple logistic regression analysis revealed that a low number of cigarettes smoked per day during the last three months before pregnancy was the best predictor for smoking cessation. Educational level, maternal age, parity and civil status were also statistically significant contributors to smoking cessation. Eighty percent of the women who were unable to stop smoking, reported a reduction in cigarette consumption. The mean number of cigarettes per day was reduced from 13.9 before pregnancy to 7.3 at 18 weeks of pregnancy (p<0.001). CONCLUSION In a national survey, 21% of the pregnant women reported smoking daily in the second trimester. Thirty-eight percent of the women who were daily smokers before the pregnancy stopped smoking in early pregnancy. A low cigarette consumption prior to the pregnancy was the best predictor for smoking cessation.

Secular trends in breastfeeding and parental smoking

To explore the association between smoking and breastfeeding, we obtained data from a retrospective questionnaire‐based national survey comprising a random sample (n= 34 799) of all mothers giving birth in Norway 1970‐91. Variables studied were postpartum smoking habits for both parents, duration of breastfeeding, infants year of birth and parental age. The response rate was 70% (n= 24 438). During the study period, the maternal postpartum smoking prevalence decreased from 38% to 26%. The proportion breastfeeding at 6 months increased from 15% to 44% among smokers, and from 30% to 72% among non‐smokers. In spite of a considerable increase in breastfeeding both among smokers and non‐smokers, the proportion of breastfeeding, non‐smoking women at 6 months was twice that of smoking women during the whole period. Furthermore, the duration of breastfeeding was shorter among young mothers and when the fathers were smoking. There was epidemiological evidence that the effect on breastfeeding of smoking might represent both biological and social mechanisms.