Combining Mendelian randomization with the sibling comparison design.

Mendelian randomization (MR) is a popular epidemiologic study design that uses genetic variants as instrumental variables (IVs) to estimate causal effects, while accounting for unmeasured confounding. The validity of the MR design hinges on certain IV assumptions, which may sometimes be violated due to dynastic effects, population stratification, or assortative mating. Since these mechanisms act through parental factors it was recently suggested that the bias resulting from violations of the IV assumptions can be reduced by combing the MR design with the sibling comparison design, which implicitly controls for all factors that are constant within families. In this article, we provide a formal discussion of this combined MR-sibling design. We derive conditions under which the MR-sibling design is unbiased, and we relate these to the corresponding conditions for the standard MR and sibling comparison designs. We proceed by considering scenarios where all three designs are biased to some extent, and discuss under which conditions the MR-sibling design can be expected to have less bias than the other two designs. We finally illustrate the theoretical results and conclusions with an application to real data, in a study of low-density lipoprotein and diastolic blood pressure using data from the Swedish Twin Registry.

Identification of risk factors for incident cervical insufficiency in nulliparous and parous women: a population-based case-control study.

BACKGROUND: Cervical insufficiency is one of the underlying causes of late miscarriage and preterm birth. Although many risk factors have been identified, the relative magnitude of their association with risk in nulliparous versus parous women has not been well demonstrated, especially for incident cervical insufficiency (ICI). The aim of this study was to investigate and compare the magnitude of the association of ICI with predictive factors in nulliparous and parous women, and to further investigate various aspects of obstetric history for parous women. METHODS: Pregnant women with a first diagnosis of cervical insufficiency were compared to a random sample of control pregnancies from women with no diagnosis by using Swedish national health registers. Demographic, reproductive, and pregnancy-specific factors were compared in case and control pregnancies, and relative risks presented as odds ratios (OR), stratified by nulliparous/parous. Independent associations with ICI were estimated from multivariable logistic regression. Associations with obstetric history were further estimated for multiparous women. RESULTS: A total of 759 nulliparous ICI cases and 1498 parous cases were identified during the study period. Multifetal gestation had a strong positive association with ICI in both groups, but of much larger magnitude for nulliparous women. The number of previous miscarriages was also a much stronger predictor of risk in nulliparous women, especially for multifetal pregnancies. History of preterm delivery (<37 weeks' gestation) was an independent predictor for parous women, and for those whose most recent delivery was preterm, the association with ICI increased with each additional week of prematurity. A previous delivery with prolonged second stage of labor or delivery of a very large infant were both inversely associated with risk of ICI in the current pregnancy. CONCLUSIONS: The differences in importance of predictive risk factors for incident cervical insufficiency in nulliparous and parous women can help resolve some of the inconsistencies in the literature to date regarding factors that are useful for risk prediction. Stratifying on parity can inform more targeted surveillance of at-risk pregnancies, enable the two groups of women to be better informed of their risks, and eventually inform screening and intervention efforts.

Generalizability and effect measure modification in sibling comparison studies.

Sibling comparison studies have the attractive feature of being able to control for unmeasured confounding by factors that are shared within families. However, there is sometimes a concern that these studies may have poor generalizability (external validity) due to the implicit restriction to families that are covariate-discordant, i.e., those families where at least two siblings have different levels of at least one of the covariates (exposure or confounders) under investigation. Even if this selection mechanism has been noted by many authors, previous accounts of the problem tend to be brief. The purpose of this paper is to provide a formal discussion of the implicit restriction to covariate-discordant families in sibling comparison studies. We discuss when and how this restriction may impair the generalizability of the study, and we show that a similar generalizability problem may in fact arise even when all families are covariate-discordant, e.g. even if the exposure is continuous so that all siblings have different exposure levels. We show how this problem can be solved by using a so-called marginal between-within model for estimation of marginal exposure effects. Finally, we illustrate the theoretical conclusions with a simulation study.

Early-Life Adversity Due to Bereavement and Inflammatory Diseases in the Next Generation: A Population Study in Transgenerational Stress Exposure.

Emerging evidence suggests that trauma experienced in childhood has negative transgenerational implications for offspring mental and physical health. We aimed to investigate whether early-life adversity experienced as bereavement is associated with chronic inflammatory health in offspring. The study population included 3 generations of Swedish families with a base population of 453,516 children (generation 3) born in 2001-2012. Exposure was defined as the middle generation's (generation 2) experiencing bereavement in childhood due to the death of a parent (generation 1). Outcomes in generation 3 included 2 diagnoses of inflammatory diseases, including asthma, allergic diseases, eczema, and autoimmune diseases. Survival analysis was used to identify causal pathways, including investigation of mediation by generation 2 mood disorders and socioeconomic status (SES). We found that early-life bereavement experienced by women was associated with early-onset offspring asthma (hazard ratio = 1.15, 95% confidence interval: 1.08, 1.23); mediation analysis revealed that 28%-33% of the association may be mediated by SES and 9%-20% by mood disorders. Early-life bereavement experienced by men was associated with autoimmune diseases in offspring (hazard ratio = 1.31, 95% confidence interval: 1.06, 1.62), with no evidence of mediation. In conclusion, adversity experienced early in life may contribute to an increased risk of inflammatory diseases which is partly mediated by mood disorders and SES.

Patterns of red-cell transfusion use in obstetric practice in sweden 2003-2017: A nationwide study.

BACKGROUND: There is a paucity of data on patterns of red-cell transfusions in obstetrical care, but some studies have suggested an increase in transfusion rates during the last decade. The purpose of this study was to investigate maternal characteristics, temporal trends and hospital variations in red-cell use in a large contemporary obstetric cohort in Sweden. STUDY DESIGN AND METHODS: Nationwide observational cohort study of maternal red-cell transfusions for all deliveries in Sweden between 2003 and 2017. RESULTS: The proportion of deliveries that received red-cell transfusions was stable during the study period, although the number of red-cell units administered per delivery declined. Among transfused women, most received a low-volume transfusion of 1 or 2 units. Red-cell transfusion was more common among the nulliparous, for instrumental and caesarean deliveries, and with increased maternal age. We saw large variations in transfusion rates between hospitals in Sweden, despite adjusting for age and parity. CONCLUSIONS: In comparison to other high-resource countries we see a high proportion of deliveries with maternal red-cell transfusions. However, we do not see an increase in red-cell use over time.

Maternal anxiety, depression and asthma and adverse pregnancy outcomes - a population based study.

To evaluate associations between maternal anxiety or depression and adverse pregnancy outcomes, taking possible familial confounding and interaction with asthma into account, we conducted a cohort study of all singleton births in Sweden 2001-2013. We retrieved information about pregnancy, diagnoses of anxiety/depression, asthma, and prescribed medication from the Swedish Medical Birth, National Patient, and Prescribed Drug Registers. We estimated associations with regression models, performed cousin and sibling comparisons, and calculated interactions. In 950 301 identified pregnancies; 5.9% had anxiety/depression and 4.0% had asthma. Anxiety/depression was associated with adverse pregnancy outcomes (e.g. preeclampsia, adjusted Odds Ratio 1.17 (95% Confidence Interval 1.12, 1.22), instrumental delivery (1.14 (1.10, 1.18)), elective (1.62 (1.57, 1.68)) and emergency (1.32 (1.28, 1.35)) caesarean section (CS)). Their children had lower birth weight (-54 g (-59, -49)) and shorter gestational age (-0.29 weeks (-0.31, -0.28)). Associations were not confounded by familial factors and asthma did not modify the effect of anxiety/depression for outcomes other than elective CS, p < 0.001. In women with anxiety/depression diagnosis, untreated women had higher odds of elective CS compared to women on medication (1.30 (1.17, 1.43)). In conclusion, anxiety/depression should be considered when evaluating pregnant women's risk of complications such as preeclampsia and non-vaginal deliveries.

Carryover Effects in Sibling Comparison Designs.

A convenient way of dealing with confounding is the sibling comparison design, where the outcome in exposed individuals is compared with the outcome in their unexposed siblings. The standard analysis of sibling comparison designs assumes that the exposure and outcome of an individual do not affect the exposure and outcome of his/her siblings, sometimes referred to as an absence of sibling carryover or contagion effects. Unfortunately, there are many situations where carryover effects are likely to be present. In this article, we explore the consequences of carryover effects for sibling comparison designs. We show, using causal diagrams, when and why carryover effects lead to bias, and we investigate the sign and magnitude of this bias under various scenarios.

Breast cancer risk in opposite-sexed twins: influence of birth weight and co-twin birth weight.

Most, but not all, studies report a positive association between birth weight, as an indirect marker of prenatal hormone exposure, and offspring breast cancer risk, particularly premenopausal breast cancer. Females from opposite-sexed twin pairs may also be prenatally exposed to androgens from their twin brothers. A Swedish study of opposite-sexed twins with a small sample size found a very strong positive association between female birth weight and breast cancer risk. In this case-control study, nested within a cohort of female opposite-sexed twins, we included 543 breast cancer case subjects diagnosed in the period from 1972 to 2008 and 2715 matched control subjects. Conditional logistic regression estimated the breast cancer risk associated with birth weight and other birth characteristics, including gestational age and co-twin birth weight. All statistical tests were two-sided. There was no association between birth weight (odds ratio = 1.01; 95% confidence interval = 0.70 to 1.46) or twin brother's birth weight and risk of breast cancer, which suggests the previously reported strong positive association may have been a chance finding.

Sibling comparison designs: bias from non-shared confounders and measurement error.

Twins, full siblings, and half-siblings are increasingly used as comparison groups in matched cohort and matched case-control studies. The "within-pair" estimates acquired through these comparisons are free from confounding from all factors that are shared by the siblings. This has made sibling comparisons popular in studying associations thought likely to suffer confounding from socioeconomic or genetic factors. Despite the wide application of these designs in epidemiology, they have received little scrutiny from a statistical or methodological standpoint. In this paper we show, analytically and through a series of simulations, that the standard interpretation of the models is subject to several limitations that are rarely acknowledged.Although within-pair estimates will not be confounded by factors shared by the siblings, such estimates are more severely biased by non-shared confounders than the unpaired estimate. If siblings are less similar with regard to confounders than to the exposure under study, the within-pair estimate will always be more biased than the ordinary unpaired estimate. Attenuation of associations due to random measurement error in exposure will also be higher in the within-pair estimate, leading within-pair associations to be weaker than corresponding unpaired associations, even in the absence of confounding. Implications for the interpretation of sibling comparison results are discussed.

Twinship influence on morbidity and mortality across the lifespan.

BACKGROUND: Studies in twins may be questioned with respect to their representativeness of the general population, not least considering the potential importance of the fetal environment for future health and disease. To better understand the influence twinning may have on health, we investigated long-term health outcomes of twins, their singleton siblings and singletons from the population. METHODS: Morbidity and mortality in twins was contrasted to that of their singleton siblings. These singletons from families with twins were then compared with singletons of the population to further reveal potential twin family influences on health. Familial relations were identified through the Swedish Multi-Generation Register. Among individuals born between 1932 and 1958, the number of twins and their singleton siblings identified were 49,156 and 35,277, respectively. Outcomes were incident overall cancer, cardiovascular disease (CVD) and death, identified in national registers. Standardized survival functions were estimated using Cox proportional hazards regression and the corresponding cumulative risks plotted against age. RESULTS: Cumulative risks of cancer, CVD and death in twins did not differ from singletons of families with twins, who in turn were found to be similar to singletons of families without twins. As could be expected from these findings, no differences in risks were found when twins were compared with singletons of the population. CONCLUSIONS: Despite their adverse intrauterine experience, twins do not seem to fare worse than singletons with respect to adult morbidity and mortality. The findings indicate that the unique experience of twinning does not lead to adverse long-term health outcomes.

Familial aggregation of schizophrenia: the moderating effect of age at onset, parental immigration, paternal age and season of birth.

AIMS: An abundance of evidence has firmly established the familial aggregation of schizophrenia. The aim of this study was to examine how age at onset, parental characteristics and season of birth modify the familiality in schizophrenia. METHODS: A population-based cohort was created by linking the Swedish Multi-Generation and Hospital Discharge Registers. Among 5,075,998 full siblings born between 1932 through to 1990, 16,346 cases of schizophrenia were identified. Familial aggregation was measured by the sibling recurrence-risk ratio, defined as the risk of schizophrenia among full siblings of schizophrenia patients compared with the risk among siblings of unaffected people. RESULTS: We found a statistically significantly lower recurrence-risk ratio in siblings of later onset cases (7.2; 95% confidence interval (95% CI) 6.7-7.9) than of early onset cases (10.8; 95% CI 9.4-12.2). A lower recurrence-risk ratio was observed among offspring to fathers above 40 years (6.3; 95% CI 5.3-7.3) as compared with offspring of younger fathers (8.6; 95% CI 8.0-9.3). Further, among offspring to parents born outside Sweden the recurrence-risk ratio was statistically significantly lower (maternal immigrants 4.8; 95% CI 4.0-5.7, paternal immigrants 5.7; 95% CI 4.6-6.9) than among offspring to parents born in Sweden. CONCLUSIONS: The familial aggregation of schizophrenia was reduced by higher age at onset, advancing paternal age and immigrant status of parents.

Birth weight predicts risk of cardiovascular disease within dizygotic but not monozygotic twin pairs: a large population-based co-twin-control study.

BACKGROUND: The widely reported inverse association between birth weight and risk of cardiovascular disease (CVD) has sparked theories about early life determinants of adult disease. Within-twin-pair analysis provides a unique opportunity to investigate whether factors shared within twin pairs influence the association. METHODS AND RESULTS: In a population-based cohort of like-sexed twins with known zygosity born in Sweden from 1926 to 1958, disease-discordant twin pairs were identified through linkage to the National Inpatient and Cause of Death registers between 1973 and 2006. Co-twin-control analyses were performed on twins discordant for cardiovascular disease (n=3884), coronary heart disease (n=2668), and stroke (n=1372). Overall, inverse associations between birth weight and risk of cardiovascular diseases were seen within dizygotic but not monozygotic twin pairs. In dizygotic twins, the odds ratios for a 1-kg within-pair increase in birth weight were 0.74 (95% confidence interval, 0.56 to 0.98) for coronary heart disease and 0.57 (95% confidence interval, 0.37 to 0.88) for stroke. Conversely, no statistically significant associations were found within monozygotic twins (for coronary heart disease: odds ratio, 1.10; 95% confidence interval, 0.73 to 1.68; for stroke: odds ratio, 0.92; 95% confidence interval, 0.48 to 1.80). CONCLUSIONS: We found an association between birth weight and risk of cardiovascular disease within disease-discordant dizygotic but not monozygotic twin pairs. This indicates that the association between birth weight and cardiovascular disease could be a result of common causes, and that factors that vary within dizygotic but not monozygotic twin pairs may help identify them.

Birth weight-breast cancer revisited: is the association confounded by familial factors?

PURPOSE: The study aimed to investigate whether the association between birth weight and the risk of breast cancer can be confounded by familial factors, such as shared environment and common genes. MATERIALS AND METHODS: Eligible were all female like-sexed twins of the Swedish Twin Registry, born during the period 1926-1958 and alive in 1973. Data were obtained from birth records, and the final study population with reliable birth weight data was made up of 11,923 twins. Hazard ratios (HR) for breast cancer according to birth weight were estimated through Cox regression, using robust SE to account for the dependence within twin pairs. Paired analysis was done to account for potential confounding by familial factors. RESULTS: In the cohort analysis, a birth weight >or=3,000 g was associated with an increased risk of breast cancer diagnosed at or before 50 years [adjusted HR, 1.57; 95% confidence interval (95% CI), 1.03-2.42] but not with breast cancer with a later onset (adjusted HR, 0.80; 95% CI, 0.57-1.12). From >or=2,500 g, a 500-g increase in birth weight conferred a HR of 1.62 (95% CI, 1.16-2.27) for breast cancer diagnosed at or before 50 years. This risk remained in analysis within twin pairs (HR, 1.57; 95% CI, 1.00-2.48). CONCLUSION: In the present study, findings indicate that the association between birth weight and breast cancer risk, seen only in women diagnosed early (

Ethnic differences in the association of birth weight and blood pressure: the Georgia cardiovascular twin study.

BACKGROUND: African Americans (AAs) not only have higher blood-pressure levels, but also an increased risk of low weight at birth, compared with European Americans (EAs). In light of fetal programming theories, it has been suggested that ethnic differences in blood pressure originate in utero. However, most previous studies in biethnic samples have not found a significant inverse association between birth weight and blood pressure in AAs. METHODS: In 562 EA and 465 AA adolescent twins of the Georgia Cardiovascular Twin Study, we investigated the potential ethnic difference in the association of blood pressure and birth weight, with the ability to control for potential confounding by familial factors. RESULTS: Blood-pressure levels were significantly higher in AAs compared to EAs, independent of birth weight (P < .01). After adjustment for parental factors and body mass index, the difference in systolic blood pressure per kg birth weight was -1.1 mm Hg (95% confidence interval, -2.7 to 0.48, P = .17) in EAs, and -2.5 mm Hg (95% confidence interval, -4.7 to -0.40, P = .02) in AAs. A significant ethnic interaction was revealed in paired analysis, where the inverse association remained in AAs, but not in EAs. Associations with diastolic blood pressure were generally weaker and nonsignificant. CONCLUSIONS: We showed that low birth weight was associated with an elevated systolic blood pressure in AAs, independent of familial factors. The results also suggest that the association between birth weight and blood pressure may be more pronounced in AAs in adolescence.