Public health interest often centres around the causal effect of an exposure on a particular outcome. For ethical, practical, or financial reasons, randomized controlled trials (RCTs) to investigate such an effect may not be possible and inferences must be drawn from observational data. These may be distorted by reverse causation, measurement errors, or the presence of socioeconomic and behavioural confounding factors that are difficult to measure even if known. Mendelian randomization is an instrumental variable (IV) method that permits estimation of the exposure-outcome causal effect in the presence of confounding. The assumptions required, i.e. all variables are continuous and all relationships between variables are linear, are often violated in epidemiological applications where the outcome (e.g. disease status) is typically binary and hence cannot have a linear relationship with the exposure. This is especially relevant to case-control data where the exposure-outcome relationship is usually logistic. The aim of this project is to develop theoretically sound methods for dealing with non-linear models and to investigate the practical implications of weakening, or even violating, some of the more restrictive conditions, given the type of data that we expect to arise in these settings. Theory must inform practice but applications should drive the theoretical investigation. This is not always the case and we view this as one of the main strengths of our proposal.

IV methods for binary outcomes are often used in econometrics to estimate causal effects when the exposure is subject to measurement error. We will investigate the adaptability of these approaches to epidemiological settings. Relevant methods also exist for RCT data when the effect for a particular subgroup is of interest. We will clarify when these are useful in Mendelian randomization applications and what conditions are required. The causal literature has shown that bounds on the average causal effect can be calculated without making restrictive distributional or relational assumptions. We will extend these to measures of casual effect that are more likely in epidemiological settings (e.g. causal odds ratios) and identify the data conditions under which they will be useful in practice. In order to get more precise estimates, data are often combined from different epidemiological studies. This pooling of data raises all the classic issues associated with meta-analysis and existing meta-analytic methods will be extended to take account of these. Finally, models that allow for more complex biological pathways will be developed to cater for multiple exposure and disease endpoints.

Introduction

Relations between cognition and substance abuse in adolescence

A number of studies have indicated that specific aspects of cognitive function may play a role in risk of substance abuse in young people1. For example, studies have reported that children at high risk for substance abuse, who have not yet started using substances (for example, children of substance dependent parents) may have specific cognitive deficits, including lower academic achievement, lower verbal and visuospatial ability, and reduced attention span and executive function. These findings have been extended in adoption studies, in which cognitive impairments were found in children of substance abusing parents, even though these children were adopted at birth and raised in an environment without substance abuse. These findings suggest that potential cognitive impairments that may predispose to later substance abuse may have a biological component and can not be entirely explained by rearing circumstances. Tarter et al. (2004) have hypothesized that cognitive tasks tapping into prefrontal cortex (PFC) dysfunction in particular may contribute to the liability to substance dependence2. However, there is a lack of studies that examine the relationships between cognitive functioning and the development of substance involvement over time, and take account of the influences of other potentially important factors, such as a youngster's personality, behavioural problems and family-related factors.

Personality, behavioural problems and family-related factors

Certain aspects of personality, including impulsivity, sensation-seeking motivation and low locus of control, have been related to increased risk of substance abuse in adolescence. However, few studies have examined how personality and cognition may work in conjunction to increase adolescents' risk of substance abuse. One cross-sectional study, which was limited to girls, reported that personality (i.e., temperament) acted as a mediator in the relation between executive function and substance use in adolescence3. It is however, also conceivable that cognition can act as a mediator in the relationship between personality and risk of substance abuse. In further disentangling these relationships, a longitudinal study design would provide major advantages in resolving the temporal relationships between cognition, personality and the development of substance abuse. Another study found that specific combinations of cognitive impairment and personality (i.e., disinhibition) posed the greatest risk for the development of substance abuse4. It would be important to replicate these findings and extend them by also examining other personality traits, such as sensation seeking and low locus of control. Relationships could first be explored for each trait individually and subsequently for combinations of personality traits. Furthermore, it would be important to take into account other aspects of a youngster's behaviour such as symptoms of conduct disorder and attention-deficit-hyperactivity disorder.

Family-related factors can also play an important role in adolescent substance abuse. For example, it is relatively well-established that young people who have a poor relationship with their parents or are poorly monitored by their parents have increased risk of substance abuse5. Young people's family background can also influence cognitive function. For example, Silveri et al. (2004) reported that parental substance abuse has a pronounced impact on young people's emotional and cognitive development6. Yet, few studies have examined to what extent the relationship between substance abuse and cognitive function in adolescence is influenced by family factors.

Genetic influences

A large literature of family, adoption and twin studies has indicated that genetic influences play a role in substance use and abuse as well as in cognition. Specific genetic variants have been studied in relation to cognitive function and substance abuse. A considerable literature indicates that the neurotransmitter dopamine plays a role in many if not all addictive phenotypes7 as well as in cognitive processes mediated by the PFC8. There has been a flurry of studies examining the role of the COMT gene in executive function. COMT encodes the enzyme catechol-o-methyltransferase (COMT) and is critical for dopamine metabolism in the PFC. For example, a study based on children participating in the ALSPAC study found that the COMT Val158Met polymorphism was associated with IQ as well as executive function (particularly cognitive tasks tapping into PFC function)9, although the effects may be stronger for IQ than for certain specific cognitive tasks10. Studies to date, however, have not tended to take into account potential differences between tasks tapping into affective (or 'hot') versus non-emotional (or 'cold') cognitive processes11. Furthermore, findings to date have largely been limited to the COMT Val158Met polymorphism, while there is evidence that the study of COMT haplotypes may be more promising in accounting for phenotypic variability12;13.

COMT has also been associated with substance abuse, although there are negative findings as well14. Some of the discrepancy may be related to personality, for example, in a sample of polysubstance abusers, the Val158Met polymorphism was found to be associated with addiction through impulsivity or dyscontrol15. When explaining the relations between cognition and substance abuse it will be important to include genetic information. This will contribute towards clarification of these relations and may also provide some insight into the reasons for the underlying discrepancies in genetic studies of substance abuse.

Aims of the proposed study:

We propose to study the relations between cognition and substance use/abuse in a large and representative UK-based sample of adolescents. The specific aims are to:

1. Examine the relations between aspects of cognition, specifically IQ and PFC functioning, and the development of substance involvement over time;

2. Examine the influences of personality, problem behaviour and family-related factors (including relations with parents, parental monitoring and parental substance abuse) on these relationships;

3. Investigate whether genes that have been related to both cognition and substance abuse (e.g., COMT) moderate the association between both phenotypes.

Sample and statistical analysis

The Avon Longitudinal Study of Parents and Children (ALSPAC) is an ongoing UK-based longitudinal study, comprising 14,541 mothers recruited in 1991 who have been followed with their offspring from before birth up to age 16 (www.alspac.ac.uk). DNA samples are available for ~10,000 children and COMT has already been typed in this cohort. Typing of other genes that may play a role in the relationship between cognition and substance abuse is currently taking place.

We would be interested in studying the following variables:

Concept Person Source Timepoints

Substances

Alcohol use Child Child/ parent Childhood/ adolescence

Cigarette use Child Child/ parent Childhood/ adolescence

Cannabis use Child Child/ parent Childhood/ adolescence

Cognition

IQ Child Child Childhood/ adolescence

Verbal ability Child Child Childhood/ adolescence

Working memory Child Child Childhood

Attention Child Clinic Childhood/ adolescence

Executive function Child Clinic Childhood/ adolescence

Inhibition/ impulsivity Child Clinic Childhood/ adolescence

Mediating/ moderating factors

Personality

Temperament Child Parent Childhood

Sensation seeking Child Child Adolescence

Locus of control Child Child Childhood/ adolescence

Family-related variables

Parental substance abuse Parent Parent Childhood/ adolescence

Parent-child relations Child/ Parent Parent/ child Childhood/ adolescence

Parental monitoring Parent Parent/ child Childhood/ adolescence

Child problem behaviour

Conduct problems Child Parent/ child Childhood/ adolescence

Attention-Deficit- Child Parent/ child Childhood/ adolescence

Hyparactivity Problems

Genetic influences

COMT Child

Relations between variables will be evaluated using linear/ logistic regression analysis and path analysis. Latent class analyses will be used to define clusters of cognitive and other variables with different predictive relationships with substance abuse. Latent growth curve modelling will allow the study of predictors of change and individual trajectories over time.

References

1. N. Z. Weinberg, J Am Acad Child Adolesc Psychiatry 36, 1177-86 (1997).

2. R. E. Tarter, L. Kirisci, M. Habeych, M. Reynolds, M. Vanyukov, Drug Alcohol Depend 73, 121-32 (2004).

3. P. R. Giancola and A. C. Mezzich, J Child Psychol Psychiatry 44, 857-66 (2003).

4. L. Kirisci, R. E. Tarter, M. Vanyukov, M. Reynolds, M. Habeych, Drug Alcohol Depend 76, 125-33 (2004).

5. K. H. Shelton et al., (In Press (J Youth Adolesc)).

6. M. M. Silveri , G. K. Tzilos, P. J. Pimentel, D. A. Yurgelun-Todd, Ann N Y Acad Sci 1021, 363-70 (2004).

7. A. I. Leshner and G. F. Koob, Proc Assoc Am Physicians 111, 99-108 (1999).

8. E. M. Tunbridge, P. J. Harrison, D. R. Weinberger, Biol Psychiatry 60, 141-51 (2006).

9. J. H. Barnett et al., Am J Psychiatry 164, 142-9 (2007).

10. J. H. Barnett, L. Scoriels, M. R. Munafo, Biol Psychiatry 64, 137-44 (2008).

11. A. Kerr and P. D. Zelazo, Brain Cogn 55, 148-57 (2004).

12. S. Shifman et al., Am J Hum Genet 71, (2002).

13. A. G. Nackley et al., Science 314, 1930-3 (2006).

14. M. B. van den Bree, Curr Psychiatry Rep 7, 125-32 (2005).

15. D. J. Vandenbergh, L. A. Rodriguez, I. T. Miller, G. R. Uhl, H. M. Lachman, Am J Med Genet 74, 439-42 (1997).

Childhood overweight and obesity has emerged as an epidemic. In 2003-2004, over 33% of US children and adolescents were overweight or obese, and over 17% were obese. As the cohort born since 1980 moves into adulthood and middle age, we will see increasing incidence of diabetes, heart disease, kidney failure, and related metabolic disorders. Considerable research effort has been expended to identify the causes of childhood obesity, a necessary first step in suggesting potential prevention strategies to mitigate or reverse this growing problem. However, to date only particular subsets of the problem have been addressed.

Some of the identified factors related to obesity are clearly nested within others. For example, characteristics of the built environment vary by race-ethnicity, and obesigenic factors are more prevalent in disadvantaged communities; this may increase the risk of obesity and low physical activity levels in children. Likewise, metabolic processes are nested within individuals, who are further located within neighborhoods; as a result, insulin resistance may also follow proximity to healthy neighborhoods and their amenities.

Disentangling the web of causation of childhood obesity is a formidable task, and both available data and standard epidemiologic analyses may be inadequate to capture multiple and interacting levels of causation. Furthermore, the presence of feedback loops precludes standard approaches to causality, which are based on acyclicity and SUTVA assumptions. For example, obese children are more likely to be socially isolated, spending more time watching television, and less time engaged in sports and clubs than non-obese children. Causality is difficult to assess in this situation: obese children may be shunned by peers, making them more likely to spend isolated recreation time watching television, and less likely to participate in sports and clubs. Alternatively, participation in clubs and sports may prevent obesity by giving children opportunities for social interaction and physical activity.

In response to RFA-HD-08-023 (Innovative Computational and Statistical Methodologies for the Design and Analysis of Multilevel Studies on Childhood Obesity), we propose a novel multilevel study design and analysis plan to address this problem. In particular, we propose a three-stage strategy based on simulation, confirmation, and manipulation.

We propose the following specific aims:

1. Generate a series of simulated datasets based on knowledge obtained from the literature.

The simulations will be created using both a statistical simulation approache and an agent-based modeling approach, whereby social interactions follow explicit rules. The characteristics [heterogeneity] of individuals and environments will be specified using existing literature. We will focus on the following levels:

(1) Proximal determinants of obesity, which are in feedback and through which other determinants operate.

* Energy expenditure, which is a weakly correlated, negative, and linear determinant of adiposity

* Energy intake, which is a strongly correlated, positive, and nonlinear determinant of adiposity

* Energy expenditure and intake jointly determine BMI through complex feedback

(2) Distal demographic determinants of energy intake and expenditure:

* Household income, parental education level, gender, and race-ethnicity

(3) Family/household behavioral and biologic characteristics of energy intake and expenditure.

* Preparing meals at home; eating meals together; parental adiposity; genetic factors (e.g., FTO)

(4) Social network characteristics as they relate to energy intake and expenditure

* Homophily: on race, gender, parental education, adiposity; centrality, prestige

* Network formation: p* models for obesity as an explanation of friendship; sport/club affiliation

(5) Characteristics of neighborhood

* SES, school funding, and the built environment (supermarkets, fast food, and recreation outlets)

Characteristics 3 through 5 will further incorporate rules of social interaction between heterogeneous agents with each other and with the social and build environments. Institutions may themselves be agents. In order to be realistic, the construction of these simulated datasets will require a thorough understanding of the mechanisms and processes by which obesity and obesegenic environments are produced, not merely an ability to predict the shape of the obesity outcome.

2. Test the agreement between the simulated models and actual (observed) patterns of obesity.

Tests will be against available data, based on the characteristics cited in Aim 1, as well as overall U.S. and regional overweight and obesity rates. Data will come from the literature as well as representative datasets (NHANES, Add Health, ALSPAC). Tests must be limited to what data are available; ALSPAC will be used primarily to understand the longitudinal dynamics of obesity, while Add Health will focus on the role of social networks. Agreement between the regression-based and agent-based models may indicate that agent-based models are unnecessary.

3. Perform pseudo-experiments using the simulated data that manipulate individual characteristics and/or rules of agent interaction.

The true significance of simulated data comes not only from demonstrating how properly specified properties and rules create realistic patterns of obesity, but in the ability to artificially manipulate these properties and rules to provide counterfactual contrasts. We may then ask questions that are impossible to answer (or would yield implausible results) using the traditional multilevel approach, e.g.: what would be the effect on childhood eating habits of adding a supermarket to a "food desert", or eliminating the stigmatizing effects of overweight on friendship formation and club participation?

A related approach involves the application of structural equation models to the simulated and longitudinal datasets (ALSPAC, Add Health). [Palloni's Monte Carlo method for SEM]

We note that only existing ALSPAC data will be used for the proposed work; there is no need for further data collection.

Background

Hypertensive disorders of pregnancy (Hypertension identified at least twice after the 20th week of gestation with proteinuria present at least twice after the 20th week of gestation (pre-eclampsia) or without proteinuria (gestational hypertension)) are by far the commonest complication of pregnancy and are associated with increased risk of maternal and perinatal mortality, small for gestational age and preterm birth in the short time and with increased risk of cardiovascular disease in mother and offspring in the long term. Hypertensive disorders of pregnancy (HDP) affect 10-20% of pregnancies in developed countries.

Despite the common occurrence of HDP, their important impact on maternal and child health in the short and long-term and the many decades of research into this condition, two recent reviews noted that the pathophysiology of these disorders, how to predict women who are at high risk and how to prevent HDP remained unclear. They note that in part this reflects the failure to appreciated the complex nature of the syndrome which is likely to be affected by a number of pathophysiological pathways; the limited ability in studies to date to explore possible differences in pathophysiology and consequences for potentially different sub-types and the lack of large studies with adequate maternal and fetal/offspring data. With respect to exploring pathophysiology and genetic pathways it has been noted that too few studies have examined maternal and fetal genotype together when taking a candidate gene approach and that no studies to date have examined within one study population a wide array of variants that relate to all of the plausible pathophysiolgogical pathways that are likely to be involved in these conditions. It was noted that studies that restrict samples to first pregnancies are problematic, because although pre-eclampsia is more common in nulliparous women, different subtypes may be more common in women who have experienced previous pregnancies and it is important to try to understand the whole spectrum of this condition.

Classical twin studies have reported heritabilities of around 20% for HDP and have lead some to conclude that the genetic contribution to HDP was smaller than hitherto believed. However, examining levels of association in twin mothers fails to recognise the potential importance of paternal genes expressed in the fetus. Two large studies using record linkage between Norwegian population registers and containing over 1 million participants examined a wide range of familial relationships and concluded that heritability was 30% with fetal paternal genotype. The plausible pathophysiological mechanisms for HDP, including maternal-fetal immunological maladaptation, maternal endothelial dysfunction, oxidative stress, inflammatory response and glucose-insulin metabolism mechanisms, have high heritabilities (greater than 50%).

A number of genome wide association studies (GWAS) of pre-eclampsia or HDP are currently underway. Only a minority of these include genetic information on fetus/offspring as well as mother. Other studies with genome wide data are potentially able to complete GWAS for pre-eclampsia with some having fetal genome wide data only and others maternal data only.

The 50K Cardiochip has densely typed loci for all pathways that are plausibly linked to HDP and cardiovascular disease and therefore provides an efficient way of establishing the many genetic variants that are likely to be associated with HDP. This includes loci that have to date been identified as relating to HDP. Brendan Keating (co-applicant) produced this chip and is responsible for updates/revisions to later versions.

Our aim is to submit a grant proposal to BHF in March 2009 that will have the following objectives:

1. Identify and collate all GWAS of HDP and also potential additional studies with genome wide data and the possibility to define cases of HDP. To combine data from all such studies in order to identify key loci in mothers and fetus associated with HDP.

2. Add any new loci from (1) to an update version of the cardiochip

3. Use the newly updated cardiochip on all ALSPAC mothers and offspring in order to determine the association of genetic variants (in mothers or fetus) associated with HDP, cardiovascular disease and plausible mechanisms for HDP, with patterns of change in blood pressure during pregnancy in ALSPAC mothers and with later blood pressure in ALSPAC offspring ages 7-17 and mothers at the 17 year postnatal clinic.

Methods

PILOT

Before submitting the grant application we would like to complete a small pilot study in ALSPAC. This would involve using the current version of the cardiochip on 1000 ALSPAC samples (500 mothers and their 500 offspring). Debbie Lawlor is able to identify 500 mothers who have had obstetric data abstracted and for whom there is extracted DNA on both mother and offspring. From this 'eligible' sample she will identify 300 mothers-offspring pairs where mother had severe HDP (pre-eclampsia and/or systolic BPgreater than 200 on at least 2 occasions after 20 weeks gestation) and a random sample of 200 mother-offspring pairs where there is no evidence of HDP or existing hypertension in mother. We would like permission to ship DNA from these mother-offspring pairs (1000 samples in total: 500 mothers/500 offspring) to the University of Pennsylvania ASAP (DAL can provide IDs immediately) where Brendan Keating will provide genotyping with the cardiochip. The cardiochip will be provided for free for this pilot and MRC CAiTE will cover shiping costs. Analyses will be completed by DAL, Brendan Keating & Nic Timpson.

Brendan Keating requires 1ug total of DNA (50 or 100ng/ul) not from cell-lines for this work.

FULL GRANT PROPOSAL

The aims of the full grant proposal are detailed above.

With respect to ALSPAC we would request funds to complete genotyping using the cardiochip on all ALSPAC mothers and offspring with DNA. Our plan would be for this genotyping on the complete sample to be completed at University of Pennsylvania with a newly updated cardiochip. Funds would be requested for shipping samples; completion of the genotyping; merging of large amount of new genetic data with the existing ALSPAC data; a statistician to complete analyses under supervision of applicants. If funded for this complete project we would again require 1ug of DNA (50 or 100ng/ul) from whole blood samples and not cell lines for each sample from mothers and offspring.

This proposal has been discussed with Sue Ring who feels that is is feasible.

(No outline received).

Obesity and overweight have been found as risk factors for asthma and asthma severity, although it is difficult to isolate exposure to obesity per se from sedentary activity and dietary factors associated with obesity(Beuther and Sutherland 2007). Furthermore, sedentary lifestyle and possibly even dietary indiscretion are thought to be mediators of this association. Unfortunately, both asthma and obesity have been increasing in prevalence(Chinn and Rona 2001). Little information exists about the potential effects of maternal obesity on the risk of asthma in their offspring. The lung branch of the National Health Lung and Blood Institute (NHLBI) has suggested studies explore in utero exposure to obesity as a possible causative factor in childhood asthma(Weiss and Shore 2004). Biologic plausibility exists for this association; obesity and asthma are both conditions with an increase in certain inflammatory markers. It has been hypothesized that obesity during pregnancy may cause immunologic changes (increase in TNF alpha, IL 4 and IL5) that can be transferred to the fetus resulting in increasing risk of atopy among offspring(L.-G. Hersoug 2007). Leptin levels known to be increased in obesity also exist in fetal lung and are associated with fetal lung maturation(Torday, Sun et al. 2002).

A study from the Fragile Families and Child Wellbeing study found an association between maternal pre-pregnancy weight and self-reported physician-diagnosed incident asthma in offspring up to age 3 years old(Reichman and Nepomnyaschy 2007). We propose to study the association of maternal weight (and weight gain during pregnancy) with incident asthma (and atopy) among offspring up to age 7 1/2 years old in the Avon Longitudinal Study of Parents and Children (ALSPAC). This analysis differs from the one study in this area, in that maternal weight at 18 weeks will be the exposure rather than pre-pregnancy weight, and the outcome of incident asthma will extend to 7 1/2 years rather than just 3 years. One advantage of ALSPAC is the phenotype of asthma is further validated in a subgroup of children through bronchial hyperresponsiveness testing with methocholine challenge. A disadvantage is that maternal weight was self-reported. Since we are relating asthma to in utero exposures, it is unlikely that follow-up beyond age 7 1/2 will be useful at this point, especially if there is no association in this age group.

Table 1 (see attached) shows the variables requested for this analysis. The main exposure will be self-reported weight and height at the week 18 pregnancy visit, maternal weight change during pregnancy, and self-reported physical activity level during pregnancy. The outcome of incident asthma will be defined in 3 separate ways and then also as a composite of these measures. First, asthma up to 3 years of age will be based on maternal responses about wheeze in their child up to that time. Second, asthma at age 7 1/2 years will be based on questions about symptoms of wheeze, treatment for asthma and a history of a physician diagnosis of asthma. Third, asthma will be defined by bronchial hyperresponsiveness (BHR) as tested with methacholine challenge. BHR testing was performed in those with a baseline FEV1 of at least 70%. A positive test indicative of BHR will be defined as a greater than 20 % reduction in FEV1 from the postsaline value prior to delivery of the maximum dose of methacholine (6.1 micro-mol). Atopy will be defined as a separate outcome based on positive responses to the pin prick skin testing. Atopic responses will be defined as those with a 2 mm or greater wheal reaction to selected common antigens and a negative reaction to the control.

Potential confounders for these analyses include maternal physical activity (hours per week participated in 11 specific exercise activities), maternal nutrition information, maternal age, maternal smoking, maternal history of asthma, maternal history of allergy/atopy, maternal housing inadequacy, maternal financial difficulties, maternal education, maternal psychopathology, and maternal social support, complications during pregnancy, pre-eclampsia, gestational diabetes, types of delivery (ie, vaginal, c-section), child's birth weight, child's gestational age at birth, child's gender, child length of time breastfed, BMI of child, child's sedentary behavior and TV viewing, child's exposure to pets, and child's exposure to environmental tobacco smoke.

References

Beuther, D. A. and E. R. Sutherland (2007). Overweight, Obesity, and Incident Asthma: A Meta-analysis of Prospective Epidemiologic Studies. 175: 661-666.

Chinn, S. and R. J. Rona (2001). Can the increase in body mass index explain the rising trend in asthma in children? 56: 845-850.

L.-G. Hersoug, A. L. (2007). "The link between the epidemics of obesity and allergic diseases: does obesity induce decreased immune tolerance?" Allergy 62(10): 1205-1213.

Reichman, N. E. and L. Nepomnyaschy (2007). "Maternal Pre-Pregnancy Obesity and Diagnosis of Asthma in Offspring at Age 3 Years." Matern Child Health J.

Torday, J. S., H. Sun, et al. (2002). "Leptin mediates the parathyroid hormone-related protein paracrine stimulation of fetal lung maturation." Am J Physiol Lung Cell Mol Physiol 282(3): L405-10.

Weiss, S. T. and S. Shore (2004). Obesity and Asthma: Directions for Research. 169: 963-968.

Life course epidemiology is concerned with understanding how risk factors occurring during development and degenerative periods combine to affect disease risk.(1) Central to this approach is an understanding that many physiological processes reach a peak level (their highest functional level) early in life, remain relatively stable through mid-life, and then decline during older age. For example, there is evidence that greatest lung volume is usually attained in late adolescence / early adulthood in humans, that it then remains constant into the 6th/7th decade of life in healthy non-smokers and then begins to decline.(2) Similarly, peak bone mass, though achieved somewhat earlier in females compared to males, is achieved by the end of the second decade in both genders, remains constant through middle-age, before declining, particularly post-menopause in women.(3) Peak vascular and metabolic health appears to be reached by late adolescence and begins to decline in healthy individuals during their 60s/70s.(4, 5) Disease risk, related to sub-optimal levels of physiological function, may result from attaining lower than average peak levels of function (most likely due to risk factors early in life during developmental periods), earlier decline from the peak level or more rapid decline in older age.(2) The importance of the developmental period in determining the level of the peak function has been emphasised in relation to bone health by describing the actual value of peak bone mass achieved by late 20s as the "bone bank for the remainder of life".(3)

This underlying theory suggests that ideal life course studies will have repeat measurements of key physiological functions across life in large cohorts of individuals and will be able to examine the relative importance of level of peak function, age at decline and speed of decline in relation to future health risk as well as key risk factors for these different aspects of functional trajectories. To our knowledge, no existing study has such repeat life course measurements from early age into late adulthood and it will be some decades before existing birth cohorts that do have repeat infant and childhood measurements reach older age. One potential alternative to examining life course trajectories in markers of physiological function in a single cohort study is to use multiple cohorts that cover different life course stages. The aim of this study is to examine the extent to which this might be possible.

We have chosen blood pressure as our measurement of an important marker of physiological function since it is an important risk factor for cardiovascular disease and is likely to have been measured in most cohort studies. Textbook descriptions of age related changes in blood pressure are based on series of cross-sectional studies of different age groups.(6) These show marked increases in systolic blood pressure from birth to age 1 (from on average 70mmHg to 95mmHg), followed by steady increases of 1-2mmHg per year to late 20s after which the rise is more marked to age 80 when it begins to level off, but never appears to decline with increasing age.(6) A very different pattern is observed for diastolic blood pressure, which appears to have a more gentle increase in early life, rising by just 2mmHg in the first year of life and then 0.5-1mmHg per year to age 40, where they remain relatively stable until the 6th decade and then show a very gradual decline.(6) On top of these average age-related changes are gender differences with little evidence of a gender difference until adolescence, at which time increases in both systolic and diastolic blood pressure become more gentle in females compared to males, so that mean blood pressure is greater in males than females by age 18.(6) On average systolic blood pressure is 10mmHg in males than females at age 18.6 However, in middle-age the increase in systolic blood pressure is greater in females than males so that by age 70 year systolic blood pressure is on average 10mmHg higher in females than males.(6) Whether these age-related differences in blood pressure from cross-sectional studies are found in prospective studies of the same individuals with repeat measurements is unclear.

The specific objectives of this study are:

1. To develop statistical methods to model the average population age related trajectory of blood pressure by combining information from cohorts of different ages with repeat measures of blood pressure.

2. To apply these models to MRC funded cohorts in order to investigate the sociodemographic differences in such trajectories.

3. To apply latent class analysis to individual cohort to identify groups of individuals with different trajectories and to investigate the risk factors for such groups.

Methods

Cohort selection

Cohorts are selected on the basis of (i) receiving core funds from the UK Medical Research Council (the first objective of our grant was to identify such cohorts and because funding is from MRC public health network the focus is primarily on MRC funded cohorts (though we hope to extend this if possible in the future); (ii) general population cohort studies; (iii) have at least three examinations on study participants where blood pressure was assessed.

To date 5 cohorts fulfiling these criteria have been identified of which ALSPAC is one

Statistical methods

Analyses will be undertaken within each individual cohort using Bayesian models and latent class analyses as appropriate.

Variables required from ALSPAC

Gender

Month & Year of birth

Whether singleton/twin/triplet

Mothers education

Fathers education

Head of household social class

Age at each clinic measurement 7, 8, 9, 10, 11, 13, 15

Systolic blood pressure at each clinic measurement 7, 8, 9, 10, 11, 13, 15

Diastolic blood pressure at each clinic measurement 7, 8, 9, 10, 11, 13, 15

History of blood pressure medication

Height and weight at age clinic measurement 7, 8, 9, 10, 11, 13, 15

Birth weight

Gestational age

References

(1) Kuh D, Ben-Shlomo Y, Lynch J, Hallqvist J, Power C. Life course epidemiology. J Epidemiol Community Health 2003 October;57(10):778-83.

(2) Strachan DP, Sheikh A. A life course approach to respiratory and allergic diseases. In: Kuh D, Ben-Shlomo Y, editors. A life course approach to chronic disease epidemiology. 2nd Edition. 2 ed. Oxford: Oxford University Press; 2004. p. 240-59.

(3) Mora S, Gilsanz V. Establishment of peak bone mass. Endocrinol Metab Clin North Am 2003 March;32(1):39-63.

(4) McGill HC, Jr., McMahan CA, Herderick EE, Malcom GT, Tracy RE, Strong JP. Origin of atherosclerosis in childhood and adolescence. Am J Clin Nutr 2000 November;72(5 Suppl):1370S-15S.

(5) Lawlor DA, Chaturvedi N. Treatment and prevention of obesity--are there critical periods for intervention? Int J Epidemiol 2006 January 5;35:3-9.

(6) Bazzano LA, He J, Whelton PK. Blood pressure in westernized and isolated populations. In: Lip GYH, Hall JE, editors. Comprehensive Hypertension.Philadelphia: Mosby, Elsevier; 2007. p. 21-30.

It is widely claimed that body mass index (BMI) is a poor measure of adiposity since it does not directly measure fat mass and does not give an indication of fat distribution. Measurements of central adiposity such as waist circumference or waist:hip circumference are said to be better predictors of adverse metabolic outcomes and more direct measurements of total or regional fat mass (such as obtained by DXA, MRI or CT scanning) are thought to better predict adverse outcomes than BMI. However, simple measurements of height and weight (that are used to derive BMI) are more feasible to complete in clinical and public health practice than methods that involve radiological examination, and BMI assessment is also likely to be easier and more reliable than measuring waist and hip circumference in clinical or public health practice. It is also possible that the best measure of adiposity (with respect to predicting adverse outcomes) differs by age and gender. To our knowledge very few studies have directly compared the magnitudes of association of BMI with other measurements of adiposity/fat distribution to determine the extent to which BMI may underestimate true associations with adverse outcomes. Several studies in adults have compared BMI and waist or waist to hip circumference but these have rarely made appropriate direct comparisons of standardised measurements. They have tended to include both measurements in multivariable models and conclude that the one remaining statistically significant is the better measurement. To our knowledge no study has attempted to make such comparisons across a range of cohorts of different ages. The aim of this project is to compare the magnitudes of association of different measurements of adiposity/fat distribution with metabolic and vascular outcomes and all-cause mortality (outcomes will vary and be appropriate to the cohort) in cohorts of different ages. Overall the project will use data from BWHHS, ALSPAC, Caerphilly, Speedwell, Boyd Orr (pending relevant permissions).

The specific aims of ALSPAC are:

1. To compare the magnitudes of association of BMI; waist circumference and DXA total and trunkal fat mass (all standardised), as measured at age 11 with each of the following outcomes:

a. Fasting insulin, glucose, lipids assessed at age 15

b. Blood pressure assessed at age 11, 13 and 15

We propose using anthropometry assessed at age 11 because a recent prosepective study (and meta-analysis presented at DoHAD conference) suggest clear associations of BMI assessed at this age and older in both girls and boys with future CHD risk. We think it important to assess associations with prospective outcomes but fasting blood assays for lipids, glucose and insulin are only available at age 15. Blood pressure will be assessed cross-sectionally (age 11) but also at 2 years and 4 years from the anthropometric assessment.

The final paper will have Amy Taylor as first author and Debbie Lawlor as last (supervising author) and co-authors from each study. We would be grateful if you could let us know who would be appropriate to include as co-authors from ALSPAC.

The physical activity recommendations suggest that children should engage in 60 minutes of moderate-to-vigorous physical activity (MVPA) on most days of the week to help prevent numerous health outcomes (1). It has recently been shown that engaging in 15 minutes of MVPA was negatively associated with obesity in the ALSPAC cohort at 12 years old (2). The physical activity recommendations and the previous ALSPAC findings do not address the potential health benefits associated with light physical activity (LPA). There are indications that LPA may be of benefit to health outcomes, with a recent study noting that light physical activity was independently associated with improved blood glucose regulation in adults (3). The implications of LPA in terms of preventing obesity have not yet been extensively studied.

Considering the principles of energy balance - reducing or increasing the intensity of physical activity, while altering the duration of physical activity, to match the energy expended with 15 minutes of MVPA, should result in the same negative association with obesity in the ALSPAC children at 12 years old. In line with this hypothesis, a cross-sectional study involving children, determined that 45 minutes of moderate physical activity (MPA) and 15 minutes of vigorous physical activity (VPA) were both associated with lower body composition measures (4). However, different physical activity intensities may be negatively associated with obesity at varying levels of energy expenditure and this hypothesis requires further research.

Aim:

To determine the influence of the intensity and duration of physical activity with regard to obesity among a sample of 12 year old children participating in ALSPAC.

Objective 1:

To determine if LPA is independently associated with obesity in the ALSPAC children at 12 years old.

Objective 2:

To determine if the energy expended with 1 MET-hour (15 minutes of MVPA) is negatively associated with obesity, when achieved through LPA, MPA and VPA in the ALSPAC children at 12 years old.

Objective 3:

To determine if the energy expended with 2, 3 and 4 MET-hours is negatively associated with obesity, when achieved through LPA, MPA, MVPA and VPA in the ALSPAC children at 12 years old.

Dependent variable:

- Obesity (top 10% fat mass)

Independent Variables:

- LPA (objective 1)

- 30 minutes of LPA (objective 2)*

- 20 minutes of MPA (objective 2)*

- 15 minutes of MVPA (objective 2)*

- 10 minutes of VPA (objective 2)*

- 60 minutes, 120 minutes & 180 minutes of LPA (objective 3)*

- 40 minutes, 60 minutes & 80 minutes of MPA (objective 3)*

- 30 minutes, 45 minutes & 60 minutes of MVPA (objective 3)*

- 20 minutes, 30 minutes & 40 minutes of VPA (objective 3)*

*See appendix 1 for the calculation of the MET-hours

Covariates:

- Gender (model 1)**

- Social class, maternal education, smoked during pregnancy, maternal obesity, birth weight & gestational age (model 2)**

- TV viewing (38mo) & sleep pattern (30mo) (model 3)**

- Pubertal status (model 4)**

**Repeat models with the inclusion of sedentary behaviour and MVPA as covariates (objective 1)

Repeat models with the incusion of sedentary behaviour and either LPA, MPA, MVPA or VPA

(objectives 2 & 3)

Data Analysis:

- Descriptive statistics

- Logistic regression (obese: yes or no)

References:

(1) Haskell WL, Lee IM, Pate RR et al. Physical activity and public health: updated recommendation for adults from the American College of Sports Medicine and the American Heart Association. Med Sci Sports Exerc. 2007;39(8):1423-34

(2) Ness AR, Leary SD, Mattocks C et al. Objectively measured physical activity and fat mass in a large cohort of children. PLoS Med. 2007;4(3):e97

(3) Healy GN, Dunstan DW, Salmon J et al. Objectively measured light-intensity physical activity is independently associated with 2-h plasma glucose. Diabetes Care. 2007;30(6):1384-9

(4) Wittmeier KDM, Mollard RC & Kriellaars DJ. Physical Activity Intensity and Risk of Overweight and Adiposity in Children. Obes 2008; 16: 415-420

Appendix 1- Calculation of MET-hours:

MVPA (4 METs):

15 minutes of MVPA = 1 MET-hour

30 minutes of MVPA = 2 MET-hours

45 minutes of MVPA = 3 MET-hours

60 minutes of MVPA = 4 MET-hours

LPA (2METs):

30 minutes of LPA = 1 MET-hour

60 minutes of LPA = 2 MET-hours

90 minutes of LPA = 3 MET-hours

180 minutes of LPA = 4 MET-hours

MPA (3 METs):

20 minutes of MPA = 1 MET-hour

40 minutes of MPA = 2 MET-hours

60 minutes of MPA = 3 MET-hours

80 minutes of MPA = 4 MET-hours

VPA (6 METs):

10 minutes of VPA = 1 MET-hour

20 minutes of VPA = 2 MET-hours

30 minutes of VPA = 3 MET-hours

40 minutes of VPA = 4 MET-hours

Appendix 2 - Variables Required to match current data:

30, 60, 90 & 180 minutes of LPA; Accelerometry; Children; 11 year clinic

20, 40, 60 & 80 minutes of MPA; Accelerometry; Children; 11 year clinic

30, 45 & 60 minutes of MVPA; Accelerometry; Children; 11 year clinic

10, 20, 30 & 40 minutes of VPA; Accelerometry; Children; 11 year clinic.