Kristin Gustavson

Attrition and generalizability in longitudinal studies: findings from a 15-year population-based study and a Monte Carlo simulation study

BackgroundAttrition is one of the major methodological problems in longitudinal studies. It can deteriorate generalizability of findings if participants who stay in a study differ from those who drop out. The aim of this study was to examine the degree to which attrition leads to biased estimates of means of variables and associations between them.MethodsMothers of 18-month-old children were enrolled in a population-based study in 1993 (N=913) that aimed to examine development in children and their families in the general population. Fifteen years later, 56% of the sample had dropped out. The present study examined predictors of attrition as well as baseline associations between variables among those who stayed and those who dropped out of that study. A Monte Carlo simulation study was also performed.ResultsThose who had dropped out of the study over 15 years had lower educational level at baseline than those who stayed, but they did not differ regarding baseline psychological and relationship variables. Baseline correlations were the same among those who stayed and those who later dropped out. The simulation study showed that estimates of means became biased even at low attrition rates and only weak dependency between attrition and follow-up variables. Estimates of associations between variables became biased only when attrition was dependent on both baseline and follow-up variables. Attrition rate did not affect estimates of associations between variables.ConclusionsLong-term longitudinal studies are valuable for studying associations between risk/protective factors and health outcomes even considering substantial attrition rates.

Longitudinal associations between relationship problems, divorce, and life satisfaction: Findings from a 15-year population-based study

Relationship problems are negatively associated with life satisfaction. Bottom-up theories assume that relationship quality affects life satisfaction while top-down theories assume that global personality dispositions affect evaluations of relationship quality. Only bottom-up theories imply that the negative association between relationship problems and life satisfaction will be removed when the relationship is ended and that divorce thus may be a positive event for persons from troubled relationships. In this study associations between relationship problems, divorce, and life satisfaction were examined among 369 heterosexual couples. Relationship problems predicted life satisfaction 15 years later in both men and women. This association was significantly stronger among not-divorced than among divorced couples. Among couples with severe relationship problems those who divorced had higher life satisfaction at 15-year follow-up than those who remained together while the reverse was true among less troubled couples. The findings thus support bottom-up theories of life satisfaction.

The Impact of Smoking in Adolescence on Early Adult Anxiety Symptoms and the Relationship between Infant Vulnerability Factors for Anxiety and Early Adult Anxiety Symptoms: The TOPP Study

Cigarette smoking is increased in people with trait anxiety and anxiety disorders, however no longitudinal data exist illuminating whether smoking in adolescence can influence the developmental trajectory of anxiety symptoms from early vulnerability in infancy to adult anxiety expression. Using The Tracing Opportunities and Problems in Childhood and Adolescence (TOPP) Study, a community-based cohort of children and adolescents from Norway who were observed from the age of 18months to age 18–19years, we explored the relationship between adolescent smoking, early vulnerability for anxiety in infancy (e.g. shyness, internalizing behaviors, emotional temperaments) and reported early adult anxiety. Structural equation modeling demonstrated that adolescent active smoking was positively associated with increased early adulthood anxiety (β = 0.17, p<0.05), after controlling for maternal education (proxy for socioeconomic status). Adolescent anxiety did not predict early adult smoking. Adolescent active smoking was a significant effect modifier in the relationship between some infant vulnerability factors and later anxiety; smoking during adolescence moderated the relationship between infant internalizing behaviors (total sample: active smokers: β = 0.85,p<0.01, non-active smokers: ns) and highly emotional temperament (total sample: active smokers: β = 0.55,p<0.01,non-active smokers: ns), but not shyness, and anxiety in early adulthood. The results support a model where smoking acts as an exogenous risk factor in the development of anxiety, and smoking may alter the developmental trajectory of anxiety from infant vulnerability to early adult anxiety symptom expression. Although alternative non-mutually exclusive models may explain these findings, the results suggest that adolescent smoking may be a risk factor for adult anxiety, potentially by influencing anxiety developmental trajectories. Given the known adverse health effects of cigarette smoking and significant health burden imposed by anxiety disorders, this study supports the importance of smoking prevention and cessation programs targeting children and adolescence.

Pathways from maternal distress and child problem behavior to adolescent depressive symptoms: a prospective examination from early childhood to adolescence.

Objective: The main aim of this study was to identify the pathways from maternal distress and child problem behaviors (i.e., internalizing and externalizing problems) across childhood and their impact on depressive symptoms during adolescence among girls and boys. Method: Data from families of 921 Norwegian children in a 15-year longitudinal community sample were used. Using structural equation modeling, the authors explored the interplay between maternal-reported distress and child problem behaviors measured at 5 time points from early (ages 1.5, 2.5, and 4.5 years) and middle (age 8.5 years) childhood to early adolescence (age 12.5 years), and their prediction of self-reported depressive symptoms during adolescence (ages 14.5 and 16.5 years). Results: The findings revealed paths from internalizing and externalizing problems throughout the development for corresponding problems (homotypic paths) and paths from early externalizing to subsequent internalizing problems (heterotypic paths). The findings suggest 2 pathways linking maternal-rated risk factors to self-reported adolescent depressive symptoms. There was a direct path from early externalizing problems to depressive symptoms. There was an indirect path from early maternal distress going through child problem behavior to depressive symptoms. In general, girls and boys were similar, but some gender-specific effects appeared. Problem behaviors in middle childhood had heterotypic paths to subsequent problems only for girls. Conclusion: The findings highlight the developmental importance of child externalizing problems, as well as the impact of maternal distress as early as age 1.5 years for the development of adolescent depressive symptoms. Findings also indicate a certain vulnerable period in middle childhood for girls. Note: See Supplemental Digital Content 1, at http://links.lww.com/JDBP/A45, for a video introduction to this article.

Reciprocal longitudinal associations between depressive symptoms and romantic partners’ synchronized view of relationship quality

Depressive symptoms are negatively associated with relationship satisfaction. The degree to which depression is associated longitudinally with relationship quality as assessed by both partners, versus with the depressed person’s unique view of the relationship, may have implications for interventions. Longitudinal associations between depressive symptoms and relationship satisfaction were examined among 260 couples. Individual measures of relationship satisfaction were decomposed into dyadic relationship quality and each partner’s additional unique view of the relationship. Depressive symptoms and dyadic relationship quality predicted changes in each other, which suggested that the longitudinal associations between depressive symptoms and relationship quality are not due to reporter biases or confounded by other individual factors, but rather that relationship quality and depressive symptoms affect each other over time.

Prenatal Exposure to Acetaminophen and Risk of ADHD

We identify a strong association between long-term acetaminophen use and offspring ADHD that is not explained by indications of use or familial ADHD. OBJECTIVES: To estimate the association between maternal use of acetaminophen during pregnancy and of paternal use before pregnancy with attention-deficit/hyperactivity disorder (ADHD) in offspring while adjusting for familial risk for ADHD and indications of acetaminophen use. METHODS: Diagnoses were obtained from the Norwegian Patient Registry for 112 973 offspring from the Norwegian Mother and Child Cohort Study, including 2246 with ADHD. We estimated hazard ratios (HRs) for an ADHD diagnosis by using Cox proportional hazard models. RESULTS: After adjusting for maternal use of acetaminophen before pregnancy, familial risk for ADHD, and indications of acetaminophen use, we observed a modest association between any prenatal maternal use of acetaminophen in 1 (HR = 1.07; 95% confidence interval [CI] 0.96–1.19), 2 (HR = 1.22; 95% CI 1.07–1.38), and 3 trimesters (HR = 1.27; 95% CI 0.99–1.63). The HR for more than 29 days of maternal acetaminophen use was 2.20 (95% CI 1.50–3.24). Use for <8 days was negatively associated with ADHD (HR = 0.90; 95% CI 0.81–1.00). Acetaminophen use for fever and infections for 22 to 28 days was associated with ADHD (HR = 6.15; 95% CI 1.71–22.05). Paternal and maternal use of acetaminophen were similarly associated with ADHD. CONCLUSIONS: Short-term maternal use of acetaminophen during pregnancy was negatively associated with ADHD in offspring. Long-term maternal use of acetaminophen during pregnancy was substantially associated with ADHD even after adjusting for indications of use, familial risk of ADHD, and other potential confounders.

Predicting Well‐Being and Internalizing Symptoms in Late Adolescence From Trajectories of Externalizing Behavior Starting in Infancy

This study aimed to examine the long-term prediction of well-being and internalizing symptoms from trajectories of externalizing behavior problems in 921 children from a population-based sample. We found that a high stable trajectory of externalizing behavior from infancy (age 1.5) to mid-adolescence (age 14.5) predicted lower scores on life satisfaction and flourishing for both girls and boys (age 18.5). The high stable trajectory also predicted higher levels of depressive symptoms in boys and anxiety symptoms in girls (age 18.5). The findings are noteworthy as they document how a person-oriented study of externalizing behavior problems starting in infancy can predict well-being and internalizing in late adolescence. The findings underline the importance of early health promotion and problem intervention efforts.

Bias in the study of prediction of change: a Monte Carlo simulation study of the effects of selective attrition and inappropriate modeling of regression toward the mean

BackgroundMedical researchers often use longitudinal observational studies to examine how risk factors predict change in health over time. Selective attrition and inappropriate modeling of regression toward the mean (RTM) are two potential sources of bias in such studies.MethodThe current study used Monte Carlo simulations to examine bias related to selective attrition and inappropriate modeling of RTM in the study of prediction of change. This was done for multiple regression (MR) and change score analysis.ResultsMR provided biased results when attrition was dependent on follow-up and baseline variables to quite substantial degrees, while results from change score analysis were biased when attrition was more strongly dependent on variables at one time point than the other. A positive association between the predictor and change in the health variable was underestimated in MR and overestimated in change score analysis due to selective attrition. Inappropriate modeling of RTM, on the other hand, lead to overestimation of this association in MR and underestimation in change score analysis. Hence, selective attrition and inappropriate modeling of RTM biased the results in opposite directions.ConclusionMR and change score analysis are both quite robust against selective attrition. The interplay between selective attrition and inappropriate modeling of RTM emphasizes that it is not an easy task to assess the degree to which obtained results from empirical studies are over- versus underestimated due to attrition or RTM. Researchers should therefore use modern techniques for handling missing data and be careful to model RTM appropriately.

Cohort Profile: The Tracking Opportunities and Problems Study (TOPP)–study of Norwegian children and their parents followed from infancy to early adulthood

Cohort Profile: The Tracking Opportunities and Problems Study (TOPP)–study of Norwegian children and their parents followed from infancy to early adulthood Wendy Nilsen,* Anne Kjeldsen, Evalill Bølstad Karevold, Anni Skipstein, Maren Sand Helland, Kristin Gustavson, Frøydis Enstad, Silje Baardstu, Espen Røysamb, Tilmann von Soest and Kristin S. Mathiesen Department of Mental Disorders, Division of Mental and Physical Health, Norwegian Institute of Public Health, Oslo, Norway, Work Research Institute, Oslo and Akershus University College of Applied Sciences, Oslo, Norway, Department of Child Development, Division of Mental and Physical Health, Norwegian Institute of Public Health, Oslo, Norway and Department of Psychology, University of Oslo, Norway

Relationship quality, divorce, and well-being: Findings from a three-year longitudinal study

Recent studies indicate that divorce sometimes is positive for subjective well-being, but we lack knowledge about the circumstances under which this is the case. Further, we lack knowledge about the possible different associations between divorce and life satisfaction (LS) vs. positive affect (PA). The current study is based on the Norwegian Mother and Child Cohort Study (MoBa). Change in LS and PA over a three-year period was examined in women who divorced and women who did not divorce (N > 37,000). Relationship quality and relationship status at follow-up were included in the models. The results indicated that divorce predicted higher LS and PA when initial relationship quality was poor. Furthermore, divorce predicted higher PA for women who, after divorce, reported being in a new romantic relationship. The associations between relationship quality and LS and PA were stronger among not-divorced than divorced women, indicating that relationship quality affects LS and PA.