Separating within and between effects in family studies: an application to the study of blood pressure in children

Early Life Factors and Polycystic Ovary Syndrome in a Swedish Birth Cohort

Polycystic ovary syndrome (PCOS) is a medical condition with important consequences for women's well-being and reproductive outcomes. Although the etiology of PCOS is not fully understood, there is increasing evidence of both genetic and environmental determinants, including development in early life. We studied a population of 977,637 singleton women born in in Sweden between 1973 and 1995, followed sometime between the age 15 and 40. The incidence of PCOS was measured using hospital register data during 2001-2012, complemented with information about the women's, parents' and sisters' health and social characteristics from population and health care registers. Cox regression was used to study how PCOS is associated with intergenerational factors, and a range of early life characteristics. 11,594 women in the study sample were diagnosed with PCOS during the follow-up period. The hazard rate for PCOS was increased 3-fold (HR 2.98, 95% CI 2.43-3.64) if the index woman's mother had been diagnosed with PCOS, and with 1.5-fold (HR 1.51, 95% CI 1.39-1.63) if their mother had diabetes mellitus. We found associations of PCOS with lower (<7) one-minute Apgar score (HR 1.19, 95% CI 1.09-1.29) and with post-term birth (HR 1.19, 95% CI 1.13-1.26). Furthermore, heavy (10+ cigarettes/day) maternal smoking (HR 1.30, 95% CI 1.18-1.44) and maternal obesity (HR 1.90, 95% CI 1.62-2.36) were strongly associated with PCOS. This study finds support for the heritability and fetal origins of PCOS. Risk of PCOS could be reduced by further emphasizing the importance of maternal and early life health.

Associations of perinatal characteristics with endometriosis: a nationwide birth cohort study

BACKGROUND

Perinatal characteristics are associated with subsequent risk of several chronic diseases. Previous studies regarding endometriosis were based on small samples and retrospective data and were limited by unmeasured confounding bias, leading to conflicting and inconclusive findings. We investigated the associations of maternal and birth characteristics with risk of endometriosis among Swedish women of reproductive age.

METHODS

This total-population register-based cohort study consisted of 628 312 singleton women born in Sweden between 1973 and 1987, who were followed for diagnosed endometriosis from age 15 years until the end of 2012. Multivariable Cox regression was applied to examine associations with perinatal characteristics. Residual unmeasured confounding was assessed through within-family and E-value analyses.

RESULTS

During follow-up, 8262 women received an endometriosis diagnosis. There were clear dose-response/linear associations of endometriosis with lower maternal education, endometriosis in the mother [adjusted hazard ratio (aHR): 2.24, 95% confidence interval (CI): 2.04-2.46], maternal smoking during pregnancy (aHR: 1.18, 95% CI: 1.04-1.33 for moderate smoker and aHR: 1.36, 95% CI: 1.18-1.57 for heavy smoker vs non-smoker), lower birthweight, and lower birthweight-for-gestational age (aHR: 0.93 per standard deviation increase, 95% CI: 0.91-0.95). Within-family and E-value analyses suggested that these perinatal characteristics are robust predictors of the incidence of endometriosis. We also found that an estimated 26% of the association between maternal smoking and early-onset endometriosis could be explained by birthweight-for-gestational age.

CONCLUSION

This study finds support for fetal origins of endometriosis, in that exposure to adverse environment or restricted development during the perinatal period may increase the risk. Further research is needed to provide an understanding of the underlying mechanisms.

Association of Cesarean Delivery With Body Mass Index z Score at Age 5 Years

This study describes a within-family study of more than 16 000 siblings to investigate the potential link between cesarean delivery and childhood obesity.

Age at Menarche and Cardiometabolic Health: A Sibling Analysis in the Scottish Family Health Study

BACKGROUND

Previous studies of age at menarche and cardiometabolic health report conflicting findings, and only a few could account for childhood characteristics. We aimed to estimate the associations of age at menarche with cardiovascular risk factors in unrelated women and within sister groups, under the assumption that within-sibship estimates will be better adjusted for shared genetics and early life environment.

METHODS AND RESULTS

Our study included 7770 women, from 5984 sibships, participating in the GS:SFHS (Generation Scotland: Scottish Family Health Study). We used fixed- and between-effects linear regression to estimate the associations within sister groups and between unrelated individuals, respectively. Within sibships, the mean difference between sisters with early menarche (≤11 years) and sisters with menarche at 12 to 13 years was 1.73 mm Hg (95% confidence interval [CI], -0.41 to 3.86) for systolic blood pressure, 1.26 mm Hg (95% CI, -0.02 to 2.55) for diastolic blood pressure, -0.06 nmol/L (95% CI, -0.11 to -0.02) for high-density lipoprotein, 0.20 nmol/L (95% CI, 0.08-0.32) for non-high-density lipoprotein, -0.34% (95% CI, -1.98 to 1.30) for glucose, 1.60 kg/m² (95% CI, 0.92-2.28) for body mass index, and 2.75 cm (95% CI, 1.06-4.44) for waist circumference. There was weak evidence of associations between later menarche (14-15 or ≥16 years) and lower body mass index, waist circumference, and blood pressure. We found no strong evidence that estimates from within- and between-sibship analyses differed (all P values >0.1). The associations with other cardiovascular risk factors were attenuated after adjustment for adult body mass index.

CONCLUSIONS

Our results suggest that confounding by shared familial characteristics is unlikely to be a major driver of the association between early menarche and adverse cardiometabolic health but do not exclude confounding by individual-level characteristics.

Vanishing twin syndrome among ART singletons and pregnancy outcomes

STUDY QUESTION

Among babies born by ART, do singleton survivors of a vanishing twin have lower birth weight than other singletons?

SUMMARY ANSWER

Vanishing twin syndrome (VTS) was associated with lower birth weight among ART singletons; a sibship analysis indicated that the association was not confounded by maternal characteristics that remain stable between deliveries.

WHAT IS KNOWN ALREADY

Previous studies indicate that ART singletons with VTS have increased risk of adverse pregnancy outcomes, compared with other ART singletons. The potential contribution of unmeasured maternal background characteristics has been unclear.

STUDY DESIGN, SIZE AND DURATION

This was a Norwegian population-based registry study, including 17 368 mothers with 20 410 ART singleton deliveries between January 1984 and December 2013.

PARTICIPANTS/MATERIALS, SETTING, METHODS

The study population included 17 291 ART singletons without VTS, 638 ART singletons with VTS and 2418 ART singletons with uncertain vanishing twin status. We estimated differences in birth weight and gestational age comparing ART singletons with VTS first to all ART singletons without VTS, and subsequently to their ART siblings without VTS, using random- and fixed-effects linear regression, respectively. The corresponding comparisons for the associations with preterm birth and small for gestational age (SGA) were conducted using random-and fixed-effects logistic regression. The sibling analysis of preterm birth included 587 discordant siblings, while the sibling analysis of SGA included 674 discordant siblings.

MAIN RESULTS AND THE ROLE OF CHANCE

ART singletons with VTS had lower birth weight when compared to all ART singletons without VTS, with an adjusted mean difference (95% CI) of -116 g (-165, -67). When we compared ART singletons with VTS to their ART singletons sibling without VTS, the adjusted mean difference was -112 g (-209, -15). ART singletons with VTS also had increased risk of being born SGA, with an adjusted odds ratio (OR) (95% CI) of 1.48 (1.07, 2.03) compared to all ART singletons without VTS, and 2.79 (1.12, 6.91) in the sibship analyses. ART singletons with VTS were also more likely to be born preterm, although this difference did not reach statistical significance.

LIMITATIONS REASONS FOR CAUTION

We did not have information on maternal socio-economic status, but this factor is accounted for in the sibship analyses. We also had no information on whether fresh or frozen embryos were replaced.

WIDER IMPLICATIONS OF THE FINDINGS

The reduction in birth weight and increased risk of SGA in ART singletons with VTS may suggest the presence of harmful intrauterine factors with long-term health impact. While vanishing twins are not routinely observed in naturally conceived pregnancies, loss of a twin is potentially a risk factor for the surviving foetus in any pregnancy. This could be further explored in large samples of naturally conceived pregnancies with the necessary information.

STUDY FUNDING/COMPETING INTEREST(S)

The authors of this study are supported in part by the UK Medical Research Council, US National Institute of Environmental Health Sciences and the Norwegian Research Council. The authors have no conflicts of interest.

TRIAL REGISTRATION NUMBER

N/A.

Birth characteristics and all-cause mortality: a sibling analysis using the Uppsala birth cohort multigenerational study

This paper investigates the association between perinatal health and all-cause mortality for specific age intervals, assessing the contribution of maternal socioeconomic characteristics and the presence of maternal-level confounding. Our study is based on a cohort of 12,564 singletons born between 1915 and 1929 at the Uppsala University Hospital. We fitted Cox regression models to estimate age-varying hazard ratios of all-cause mortality for absolute and relative birth weight and for gestational age. We found that associations with mortality vary by age and according to the measure under scrutiny, with effects being concentrated in infancy, childhood or early adult life. For example, the effect of low birth weight was greatest in the first year of life and then continued up to 44 years of age (HR between 2.82 and 1.51). These associations were confirmed in within-family analyses, which provided no evidence of residual confounding by maternal characteristics. Our findings support the interpretation that policies oriented towards improving population health should invest in birth outcomes and hence in maternal health.

No association of maternal gestational weight gain with offspring blood pressure and hypertension at age 18 years in male sibling-pairs: a prospective register-based cohort study

BACKGROUND

Maternal gestational weight gain (GWG) is associated with birth weight, obesity, and possibly blood pressure (BP) and hypertension in the offspring. These associations may however be confounded by genetic and/or shared environmental factors. In contrast to previous studies based on non-siblings and self-reported data, we investigated whether GWG is associated with offspring BP and hypertension, in a register-based cohort of full brothers while controlling for fixed shared effects.

METHODS

By using Swedish nation-wide record-linkage data, we identified women with at least two male children (full brothers) born 1982-1989. Their BP was obtained from the mandatory military conscription induction tests. We adopted linear and Poisson regression models with robust variance, using generalized estimating equations to analyze associations between GWG and BP, as well as with hypertension, within and between offspring sibling-pairs.

RESULTS

Complete data on the mothers' GWG and offspring BP was obtained for 9,816 brothers (4,908 brother-pairs). Adjusted regression models showed no significant associations between GWG and SBP (β = 0.03 mmHg per 1-kg GWG difference, [95% CI -0.08, 0.14], or DBP (β = -0.03 mmHg per 1-kg GWG difference [95% CI -0.11, 0.05]), or between GWG and offspring's risk of hypertension (relative risk = 1.0 [95% CI 0.99, 1.02], neither within nor between siblings.

CONCLUSIONS

In this large sibling-pair study, we did not find any significant association between GWG and offspring BP or the risk of hypertension at 18y, when taking genetic and environmental factors shared within sibling pairs into account. Further large sibling studies are required to confirm a null association between GWG and other cardiovascular risk factors.

Associations between birth characteristics and eating disorders across the life course: findings from 2 million males and females born in Sweden, 1975-1998

Birth characteristics predict a range of major physical and mental disorders, but findings regarding eating disorders are inconsistent and inconclusive. This total-population Swedish cohort study identified 2,015,862 individuals born in 1975-1998 and followed them for anorexia nervosa, bulimia nervosa, and eating disorder not otherwise specified until the end of 2010. We examined associations with multiple family and birth characteristics and conducted within-family analyses to test for maternal-level confounding. In total, 1,019 males and 15,395 females received an eating disorder diagnosis. Anorexia nervosa was independently predicted by multiple birth (adjusted hazard ratio = 1.33, 95% confidence interval: 1.15, 1.53) for twins or triplets vs. singletons) and lower gestational age (adjusted hazard ratio = 0.96, 95% confidence interval: 0.95, 0.98) per extra week of gestation, with a clear dose-response pattern. Within-family analyses provided no evidence of residual maternal-level confounding. Higher birth weight for gestational age showed a strong, positive dose-response association with bulimia nervosa (adjusted hazard ratio = 1.15, 95% confidence interval: 1.09, 1.22, per each standard-deviation increase), again with no evidence of residual maternal-level confounding. We conclude that some perinatal characteristics may play causal, disease-specific roles in the development of eating disorders, including via perinatal variation within the normal range. Further research into the underlying mechanisms is warranted. Finally, several large population-based studies of anorexia nervosa have been conducted in twins; it is possible that these studies considerably overestimate prevalence.

Maternal diabetes in pregnancy and offspring cognitive ability: sibling study with 723,775 men from 579,857 families

AIMS/HYPOTHESIS

The aim of this study was to investigate the association between maternal diabetes in pregnancy and offspring cognitive ability and also to assess whether the association was due to intrauterine mechanisms or shared familial characteristics.

METHODS

We linked national registers and conducted a prospective cohort study of singleton Swedish-born men to explore associations between maternal pregnancy diabetes and educational achievement at age 16 years, the age of completing compulsory education in Sweden (n = 391,545 men from 337,174 families, graduating in 1988-1997 and n = 326,033 men from 282,079 families, graduating in 1998-2009), and intelligence quotient (IQ) at the mandatory conscription examination at 18 years of age (n = 664,871 from 543,203 families).

RESULTS

Among non-siblings, maternal diabetes in pregnancy was associated with lower offspring cognitive ability even after adjustment for maternal age at birth, parity, education, early-pregnancy BMI, offspring birth year, gestational age and birthweight. For example, in non-siblings, the IQ of men whose mothers had diabetes in their pregnancy was on average 1.36 points lower (95% CI -2.12, -0.60) than men whose mothers did not have diabetes. In comparison, we found no such association within sibships (mean difference 1.70; 95% CI -1.80, 5.21).

CONCLUSIONS/INTERPRETATION

The association between maternal diabetes in pregnancy and offspring cognitive outcomes is likely explained by shared familial characteristics and not by an intrauterine mechanism.

Robustness of parameter and standard error estimates against ignoring a contextual effect of a subject-level covariate in cluster-randomized trials

In experimental research, it is not uncommon to assign clusters to conditions. When analysing the data of such cluster-randomized trials, a multilevel analysis should be applied in order to take into account the dependency of first-level units (i.e., subjects) within a second-level unit (i.e., a cluster). Moreover, the multilevel analysis can handle covariates on both levels. If a first-level covariate is involved, usually the within-cluster effect of this covariate will be estimated, implicitly assuming the contextual effect to be equal. However, this assumption may be violated. The focus of the present simulation study is the effects of ignoring the inequality of the within-cluster and contextual covariate effects on parameter and standard error estimates of the treatment effect, which is the parameter of main interest in experimental research. We found that ignoring the inequality of the within-cluster and contextual effects does not affect the estimation of the treatment effect or its standard errors. However, estimates of the variance components, as well as standard errors of the constant, were found to be biased.

Does maternal weight gain in pregnancy have long-term effects on offspring adiposity? A sibling study in a prospective cohort of 146,894 men from 136,050 families

BACKGROUND

A small number of relatively small studies have found greater gestational weight gain to be associated with greater offspring body mass index (BMI; in kg/m(2)), but whether this association is caused by intrauterine mechanisms or by shared genetic and environmental risk factors for adiposity is unclear.

OBJECTIVE

The objective was to examine the association of greater maternal weight gain (MWG; postnatal weight minus weight at the first antenatal clinic assessment) with greater offspring BMI and to explore whether any observed association is explained by intrauterine mechanisms.

DESIGN

This was a prospective cohort study that used record linkage data (n = 146,894 individuals from 136,050 families). To compare the within-sibling and between-nonsibling associations, we used fixed- and between-cluster linear regression models.

RESULTS

Associations of MWG with later offspring BMI differed by the mother's early-pregnancy overweight or obesity status (P for interaction <0.0001). MWG was positively associated with BMI at a mean age of 18 y in the offspring of normal-weight women but only between unrelated men (0.07; 95% CI: 0.06, 0.07) per 1-kg greater MWG; no within-sibling association (0.00; 95% CI: -0.02, 0.02) per 1-kg greater MWG was found. In contrast, in overweight and obese women we found a within-sibling association (0.06; 95% CI: 0.01, 0.12) and an association between unrelated men (0.02; 95% CI: 0.01, 0.03) per 1-kg greater MWG.

CONCLUSION

In normal-weight mothers, most of the association between MWG and later offspring BMI is explained by shared familial (genetic and early environmental) characteristics, whereas evidence indicates a contribution of intrauterine mechanisms in overweight and obese women.

Fetal, developmental, and parental influences on cystatin C in childhood: the Uppsala Family Study

BACKGROUND

The aim was to identify determinants (biomedical and social characteristics of children and their parents) of cystatin C levels in healthy children drawn from a population sample.

STUDY DESIGN

Cross-sectional study.

SETTING & PARTICIPANTS

425 pairs of consecutive full siblings born 1987-1995 in Uppsala were identified using the Swedish Medical Birth Registry and invited with their parents for examination in 2000-2001.

OUTCOME

Serum cystatin C level was log-transformed and analyzed using random-effects models.

MEASUREMENTS

The examination in parents and children consisted of a nonfasting blood sample, anthropometry, and questionnaires about lifestyle and socioeconomic position. Tanner stage was used for assessment of pubertal status.

RESULTS

In age-, height-, and body mass index-adjusted analyses, cystatin C level increased by 2.6% (95% CI, 0.3%-4.8%) higher in Tanner stage 2 vs 1 girls, and 1.6% (95%CI, 0.2%-3.1%) lower in boys than girls. For every 10% increase in maternal cystatin C level, offspring cystatin C level increased by 3.0% (95% CI, 2.2%-3.8%); the equivalent effect for paternal cystatin C level was 2.1% (95% CI, 1.3%-2.9%). Lower maternal education was associated with a 2.4% (95% CI, 0.3%-4.6%) higher cystatin C level in their offspring.

LIMITATIONS

Cross-sectional study design, missing cystatin C values for subset of parents, lack of urinary measurements, no gold-standard measurement of glomerular filtration rate.

CONCLUSIONS

There are intergenerational associations of cystatin C level in families in line with previous reports of heritability of kidney disease. Lower maternal education is associated with higher cystatin C levels in their children. Further studies of healthy children are needed to explore the biological mechanisms for these findings. If cystatin C is measured, these studies will need to record pubertal stages.

Association of maternal diabetes mellitus in pregnancy with offspring adiposity into early adulthood: sibling study in a prospective cohort of 280,866 men from 248,293 families

BACKGROUND

Maternal diabetes mellitus in pregnancy results in greater offspring adiposity at birth. It is unclear whether it is associated with greater adiposity into adulthood, and if so, whether this is via intrauterine mechanisms or shared familial characteristics.

METHODS AND RESULTS

A record-linkage prospective cohort study of 280,866 singleton-born Swedish men from 248,293 families was used to explore the intrauterine effect of maternal diabetes mellitus on offspring body mass index (BMI) in early adulthood. Maternal diabetes mellitus during pregnancy was associated with greater mean BMI at age 18 in their sons. The difference in BMI was similar within brothers and between nonsiblings. BMI of men whose mothers had diabetes mellitus during their pregnancy was on average 0.94 kg/m² greater (95% confidence interval [CI], 0.35 to 1.52) than in their brothers born before their mother was diagnosed with diabetes, after adjustment for birth year, maternal age, parity and education, birth weight, gestational age, and age at assessment of BMI. Early-pregnancy BMI was positively associated with son's BMI between nonsiblings, but there was no association within brothers. Adjustment of the maternal diabetes-offspring BMI association for maternal BMI did not alter the association either within brothers or between nonsiblings. Results were also robust to sensitivity analyses restricting the within-sibling analyses to siblings born within 3 years of each other.

CONCLUSION

Maternal diabetes mellitus has long-term consequences for greater BMI in offspring; this association is likely to be via intrauterine mechanisms, and is independent of maternal BMI in early pregnancy.

Gestational age differences in health and development among young Swedish men born at term

BACKGROUND

Although increased morbidity and mortality associated with pre-term birth and restricted fetal growth have been extensively studied, relatively little is known about variations in health outcomes among term births, because they are often assumed to be homogeneous.

METHODS

We examined variations in height, body mass index (BMI), systolic and diastolic blood pressure (SBP and DBP), and intellectual performance by gestational age and fetal 'growth' (birth weight for gestational age) among young Swedish men born at term (37-41 weeks of gestation). We also compared the magnitude of associations among 314,642 men from different families and among 72,212 full brothers from 35,215 families to assess whether the associations are explained by familial factors shared by siblings.

RESULTS

Gestational age in completed weeks was positively associated with height [0.11 cm, 95% confidence interval (CI): 0.09-0.13] and intellectual performance (0.01, 95% CI: 0.00-0.02) and negatively associated with SBP (-0.28 mmHg, 95% CI: -0.33 to -0.24), after controlling for birth weight, maternal age at the men's birth, parity, family socio-economic position and family structure. The associations with height and SBP were observed also among brothers within families, suggesting that they are not explained by shared family characteristics. However, the positive association between gestational age and intellectual performance was no longer present within families. Birth weight for gestational age (z-score) was positively associated with height, BMI and intellectual performance and negatively associated with SBP. These associations were robust within families.

CONCLUSIONS

Among young men born at term, fetal growth and even gestational age are independently associated with adult size, BP and cognitive ability. The extent to which shared family characteristics explain the associations varies across outcomes.

Birth weight and cognitive ability in childhood among siblings and nonsiblings

OBJECTIVES

The purpose of this work was to examine whether the positive association between birth weight and childhood cognitive ability is seen within siblings from the same family, as well as between nonsiblings, and to determine whether these associations vary with age.

METHODS

We compared the association of birth weight with cognitive ability measured at ages 5 to 6, 7 to 9, and 11 to 12 years among a total of 5402 children from different families with that among 2236 to 3083 sibships from the National Longitudinal Study of Youth 1979-Children.

RESULTS

In the whole cohort, there were positive associations between birth weight and cognitive ability at all ages, with the association increasing with age from a 0.81-point increase at ages 5 to 6 years to 1.30 and 1.44 points at ages 7 to 9 and 11 to 12 years, respectively, per 1 SD of gestational age- and gender-adjusted birth weight z score. With adjustment for covariates, there was marked attenuation of these associations. Mean differences were 0.28 points in children aged 5 to 6 years, 0.67 points in those aged 7 to 9 years, and 0.52 points in those aged 11 to 12 years after adjusting for child's gender, race or ethnicity, year of birth, and age at test; maternal age, height, parity, education, smoking during pregnancy, and cognitive ability; and household income. Our family-based analyses that separated within- and between-family effects found that the between-family associations were much stronger than the within-family associations. However, adjustment for potential confounders attenuated the between-family associations, and there was no evidence for a difference in association comparing the between- and within-family associations.

CONCLUSIONS

In these data, the positive association between birth weight and childhood cognitive ability at ages 5 to 12 years is explained largely by family characteristics rather than a specific intrauterine effect.

The Uppsala studies on developmental origins of health and disease

Large-scale epidemiological studies on developmental origins of health running in Uppsala benefit from the accessibility of well-archived and high quality Swedish records, high participation rates in the surveys, and a long tradition of interdisciplinary research and international collaboration. The UBCoS Multigen study is unique in being able to study intergenerational determinants of health and health inequalities as 'forward in time' processes, starting at the beginning of the last century, whilst the Uppsala Family and ULSAM studies contribute evidence on gene-environment interactions and specific mechanisms of developmental origins of circulatory diseases.

Associations of gestational age and intrauterine growth with systolic blood pressure in a family-based study of 386,485 men in 331,089 families

BACKGROUND

We conducted a family-based study to explore mechanisms underlying the associations of birth weight and gestational age with systolic blood pressure measured at 17 to 19 years of age.

METHODS AND RESULTS

A record linkage study of 386,485 singleton-born men from 331,089 families was undertaken. Birth weight was inversely associated with systolic blood pressure within siblings, with a mean difference (adjusted for age at examination, examination center, and year of examination) within siblings per 1-SD difference in birth weight of -0.21 mm Hg (95% CI, -0.33 to -0.08) and between nonsiblings of -0.12 (95% CI, -0.16 to -0.08). Gestational age was inversely associated with systolic blood pressure within siblings (-0.18 mm Hg; 95% CI, -0.25 to -0.11, per week of gestational age) and between nonsiblings (-0.26 mm Hg; 95% CI, -0.29 to -0.24). Adjustment for socioeconomic position and maternal characteristics did not alter these within- or between-family associations. Furthermore, the associations were not affected by adjustment for paternal height, body mass index, or systolic blood pressure.

CONCLUSIONS

Our present findings suggest that the inverse associations of birth weight and gestational age with systolic blood pressure are not explained by confounding resulting from family socioeconomic position or other factors that are shared by siblings. Variations in maternal metabolic or vascular health during pregnancy or placental implantation and function may explain these associations.

Using bivariate models to understand between- and within-cluster regression coefficients, with application to twin data

In the regression analysis of clustered data it is important to allow for the possibility of distinct between- and within-cluster exposure effects on the outcome measure, represented, respectively, by regression coefficients for the cluster mean and the deviation of the individual-level exposure value from this mean. In twin data, the within-pair regression effect represents association conditional on exposures shared within pairs, including any common genetic or environmental influences on the outcome measure. It has therefore been proposed that a comparison of the within-pair regression effects between monozygous (MZ) and dizygous (DZ) twins can be used to examine whether the association between exposure and outcome has a genetic origin. We address this issue by proposing a bivariate model for exposure and outcome measurements in twin-pair data. The between- and within-pair regression coefficients are shown to be weighted averages of ratios of the exposure and outcome variances and covariances, from which it is straightforward to determine the conditions under which the within-pair regression effect in MZ pairs will be different from that in DZ pairs. In particular, we show that a correlation structure in twin pairs for exposure and outcome that appears to be due to genetic factors will not necessarily be reflected in distinct MZ and DZ values for the within-pair regression coefficients. We illustrate these results in a study of female twin pairs from Australia and North America relating mammographic breast density to weight and body mass index.

Fetal, Developmental, and Parental Influences on Childhood Systolic Blood Pressure in 600 Sib Pairs

BACKGROUND

Little is known about the contribution of maternal and paternal factors to the inverse association between birth weight and later blood pressure in human offspring. A study of within- and between-family associations of birth weight with blood pressure, which collected data on both parents, would address this gap in our knowledge.

METHODS AND RESULTS

The study examined families composed of mother, father, and 2 full sibs delivered between 38 and 41 weeks' gestation within 36 months of each other. A total of 1967 families meeting our inclusion criteria were contacted and 602 were examined (children 5 to 14 years old, 1998 to 2000). Birth weight and gestational age were available from obstetric records. Systolic blood pressure in childhood was inversely associated with birth weight within families (-2.3 mm Hg/kg, 95% CI -4.4 to -0.3) after adjustment for gestational age, sex, height, and weight at examination. The between-family effect (-1.5 mm Hg/kg, -3.1 to 0.0) was strengthened on adjustment for maternal and paternal height and weight, whereas adjustment for paternal and maternal systolic blood pressure at examination independently attenuated the effect.

CONCLUSIONS

The existence of an inverse association of birth weight with systolic blood pressure within families (adjusted for height and weight at examination) demonstrates that factors that vary between pregnancies in the same woman (including fetal genotype) can influence the later blood pressure of offspring. We conclude that this apparent fetal programming effect on blood pressure will not be eliminated solely by interventions aimed at modifying growth and cumulative nutritional status from conception through childhood or other fixed characteristics of future mothers.

Regression models for twin studies: a critical review

Twin studies have long been recognized for their value in learning about the aetiology of disease and specifically for their potential for separating genetic effects from environmental effects. The recent upsurge of interest in life-course epidemiology and the study of developmental influences on later health has provided a new impetus to study twins as a source of unique insights. Twins are of special interest because they provide naturally matched pairs where the confounding effects of a large number of potentially causal factors (such as maternal nutrition or gestation length) may be removed by comparisons between twins who share them. The traditional tool of epidemiological 'risk factor analysis' is the regression model, but it is not straightforward to transfer standard regression methods to twin data, because the analysis needs to reflect the paired structure of the data, which induces correlation between twins. This paper reviews the use of more specialized regression methods for twin data, based on generalized least squares or linear mixed models, and explains the relationship between these methods and the commonly used approach of analysing within-twin-pair difference values. Methods and issues of interpretation are illustrated using an example from a recent study of the association between birth weight and cord blood erythropoietin. We focus on the analysis of continuous outcome measures but review additional complexities that arise with binary outcomes. We recommend the use of a general model that includes separate regression coefficients for within-twin-pair and between-pair effects, and provide guidelines for the interpretation of estimates obtained under this model.

Morning cortisol does not mediate the association of size at birth with blood pressure in children born from full-term pregnancies

OBJECTIVE

It had been suggested that programming of the hypothalamo-pituitary-adrenal axis may underlie the associations of reduced size at birth with raised blood pressure in later life. We investigated whether morning salivary cortisol mediates the inverse association of birthweight with systolic blood pressure in children.

DESIGN

Subjects and measurements--a historical cohort study involving 1152 Swedish children aged 5-14 years, who took part in a family study comprised of mother, father, and two full-sibs delivered in 1987-1995 after 38-41 weeks gestation within 36 months of each other. Birthweight and gestational age were available from obstetric records. Blood pressure, weight, height and puberty stage were measured at a clinic. Cortisol was measured by radioimmunoassay in morning salivary samples taken at home, within 30 min of waking.

RESULTS

Morning cortisol showed a weak negative association with length of gestation in siblings, was not related to birthweight or to systolic or diastolic blood pressure. There was no change in the strength of the negative association between birthweight and systolic blood pressure on adjustment for cortisol (-1.4 mmHg/kg, 95% CI -2.7, -0.2; adjusted for age, sex, puberty stage, weight and height, and cortisol).

CONCLUSIONS

Morning cortisol was not associated with size at birth, and did not mediate the birthweight-blood pressure association in children born from full-term pregnancies. It is possible that basal cortisol levels are of more importance in explaining associations of size at birth with later blood pressure in older subjects, or in populations with more varied length of gestation. Alternatively, our results may be caused by misclassification of the hypothalamo-pituitary-adrenal activity.