Addressing issues in sparseness, ecological bias and formulation of the adjacency matrix in bayesian spatio temporal analysis of disease counts

Next, to address the sparseness and excess zeros commonly encountered in the analysis of rare outcomes such as birth defects, I compared a few models, including an extension of the usual

School of Mathematical Sciences Queensland University of Technology

Addressing Issues in Sparseness, Ecological Bias and Formulation of the Adjacency Matrix in Bayesian Spatio-temporal Analysis of Disease Counts

Arul Earnest

B.Soc.Sc (Hons) in Statistics, National University of Singapore

MSc in Medical Statistics, London School of Hygiene and Tropical Medicine,

University of London

A thesis submitted for the degree of Doctor of Philosophy in the Faculty of Science and Technology, Queensland University of Technology according to QUT requirements

Principal Supervisor: Professor Kerrie Mengersen

Associate Supervisors: Associate Professor Geoff Morgan

Professor Tony Pettitt

KEYWORDS

Spatial, autoregressive, disease mapping, CAR model, birth defects, ecological bias,

neighbourhood weight matrix, forecasting, priors, Bayesian, MCMC, joint modeling

ABSTRACT

The main objective of this PhD was to further develop Bayesian spatio-temporal models

(specifically the Conditional Autoregressive (CAR) class of models), for the analysis of

sparse disease outcomes such as birth defects The motivation for the thesis arose from

problems encountered when analyzing a large birth defect registry in New South Wales

The specific components and related research objectives of the thesis were developed

from gaps in the literature on current formulations of the CAR model, and health service

planning requirements Data from a large probabilistically-linked database from 1990 to

2004, consisting of fields from two separate registries: the Birth Defect Registry (BDR)

and Midwives Data Collection (MDC) were used in the analyses in this thesis

The main objective was split into smaller goals The first goal was to determine how the

specification of the neighbourhood weight matrix will affect the smoothing properties of

the CAR model, and this is the focus of chapter 6 Secondly, I hoped to evaluate the

usefulness of incorporating a zero-inflated Poisson (ZIP) component as well as a

shared-component model in terms of modeling a sparse outcome, and this is carried out in

chapter 7 The third goal was to identify optimal sampling and sample size schemes

designed to select individual level data for a hybrid ecological spatial model, and this is

done in chapter 8 Finally, I wanted to put together the earlier improvements to the CAR

For the first objective, I examined a series of neighbourhood weight matrices, and

showed how smoothing the relative risk estimates according to similarity by an

important covariate (i.e maternal age) helped improve the model’s ability to recover the

underlying risk, as compared to the traditional adjacency (specifically the Queen)

method of applying weights

Next, to address the sparseness and excess zeros commonly encountered in the analysis

of rare outcomes such as birth defects, I compared a few models, including an extension

of the usual Poisson model to encompass excess zeros in the data This was achieved via

a mixture model, which also encompassed the shared component model to improve on

the estimation of sparse counts through borrowing strength across a shared component

(e.g latent risk factor/s) with the referent outcome (caesarean section was used in this

example) Using the Deviance Information Criteria (DIC), I showed how the proposed

model performed better than the usual models, but only when both outcomes shared a

strong spatial correlation

The next objective involved identifying the optimal sampling and sample size strategy

for incorporating individual-level data with areal covariates in a hybrid study design I

SLAs, followed by selecting all cases in the SLAs chosen, along with an equal number

of controls, provided the lowest AMSE

The final objective involved combining the improved spatio-temporal CAR model with

population (i.e women) forecasts, to provide 30-year annual estimates of birth defects at

the Statistical Local Area (SLA) level in New South Wales, Australia The projections

were illustrated using sixteen different SLAs, representing the various areal measures of

socio-economic status and remoteness A sensitivity analysis of the assumptions used in

the projection was also undertaken

By the end of the thesis, I will show how challenges in the spatial analysis of rare

diseases such as birth defects can be addressed, by specifically formulating the

neighbourhood weight matrix to smooth according to a key covariate (i.e maternal age),

incorporating a ZIP component to model excess zeros in outcomes and borrowing

strength from a referent outcome (i.e caesarean counts) An efficient strategy to sample

individual-level data and sample size considerations for rare disease will also be

presented Finally, projections in birth defect categories at the SLA level will be made

TABLE OF CONTENTS

2.2.3 Areal-level indices of socio-economic status 14

2.3 Definition and classification of birth defects 16

2.4 Spatial and temporal trends of birth defects in New South Wales,

Australia 18

3.3.5 Common risk factors for caesarean section rates/ spatial variation 31

5.5 Specifying the hyperprior distribution 78

STATEMENT OF ORIGINAL AUTHORSHIP

"The work contained in this thesis has not been previously submitted to meet

requirements for an award at this or any other higher education institution To the best of

my knowledge and belief, the thesis contains no material previously published or written

by another person except where due reference is made”

Arul Earnest

26 th February 2010

ACKNOWLEDGEMENTS

I would like to thank my principal supervisor, Professor Kerrie Mengersen, from

Queensland University of Technology (QUT), for her unlimited guidance and

supervision throughout the course of my PhD candidature I am indebted to her for

introducing the field of Bayesian statistics to me My appreciation also goes out to

Professor Tony Pettitt for facilitating the smooth flow of my PhD studies I would also

like to express my gratitude to my associate supervisor, Associate Professor Geoff

Morgan, from the Northern Rivers University Department of Rural Health (University of

Sydney) for constantly providing input on my PhD, in particular the epidemiological,

study design and clinical implication aspects of the thesis I have certainly enjoyed the

numerous thought-provoking discussions we had in his office in Lismore I am equally

indebted to Professor John Beard, director of Ageing and Lifecourse at the World Health

Organisation, who was my previous supervisor I would like to credit him with

providing me with the opportunity to start on this PhD studies, and also for his generous

advice and guidance on the manuscripts resulting from this thesis My sincere gratitude

goes to Dr Lee Taylor and Dr David Muscatello from the New South Wales Department

of Health for providing me with useful advice on the data upon which this thesis is built

on, and also valuable opinion on the practical applications resulting from this thesis I

CHAPTER 1 INTRODUCTION

1.1 Primary research aims and motivation

This thesis aims to answer questions related to the small area analysis of sparse disease

counts in a geographical region The first question relates to the formulation of the

Conditional Autoregressive (CAR) model, a commonly used statistical model in the

analysis of geographically aggregated data Specifically, I wanted to evaluate whether

the formulation of the neighbourhood weight matrix has any impact on the smoothing

properties of the CAR model In addition, I wished to examine whether there were any

differences between the adjacency and distance-based methods of assigning neighbours

in terms of recovering the underlying relative risk estimates

The second hypothesis relates to the modeling or estimation of a sparse outcome, such as

birth defects The questions I wished to answer were: “Can we better estimate the

outcome with a sparse count by jointly modelling it with another related outcome that

may share some latent risk factors?” and “Can we improve on the estimates by

incorporating a component (zero-inflated Poisson component through a mixture model)

to model the excess zeros in the data?”

The third broad question relates to a CAR regression model, and includes both

performed an extensive simulation analysis to evaluate 13 different scenarios, including

various sampling schemes and variations in sample size

The fourth aim of this thesis was to provide a method for forecasting sparse outcomes at

a small-area level, which took into account spatial correlation in the data, optimal

neighbourhood weight matrix formulation, consideration of excess zeros in the data, as

well as population (women) forecasts for the next 30 years at the Statistical Local Area

(SLA) level in New South Wales, Australia Sensitivity analysis based on different

population scenarios was also assessed

The motivation for this thesis came about from the challenges faced when analyzing

birth defects from a large registry in New South Wales (NSW), Australia, as part of an

Australian Research Council (ARC) linkage grant The first challenge faced was

sparseness of the disease outcome, especially when individual birth defects were

mapped in geographical locations, or even when defects were analysed in broader

groupings, according to the International Classification of Disease- British Pediatric

Association (ICD9-BPA) coding system The problem was compounded when there

were a large number of areas with zero counts of particular defects Secondly, one had to

The impetus to fine-tune the CAR model was primarily driven by gaps in literature,

identified after an extensive literature review on single-disease and multiple-disease

CAR models was performed Both spatial only and spatio-temporal models were

evaluated and compared The review revealed that most of the models were applied to

outcomes that were not rare, and applied to data across broad time intervals, thus

ensuring that there were enough cases in each time point Disease mapping studies

involving birth defects were few, and none of them actually accounted for spatial

correlation in the data Almost all the spatial studies used the simpler formulation of the

Queen adjacency method of assigning neighbours, which I suspect was done out of

convenience The various formulations of the CAR models also failed to incorporate

sparseness in the data, implicitly or explicitly

1.2 Content and scope of thesis

This section details what is covered in this thesis and areas which are not within the

scope of this manuscript This section also provides the links between the various

chapters

In chapter 1, the motivation for undertaking the study is stated, along with the main aims

of this thesis The content, scope and structure of the thesis are also presented in this

chapter The source of data used in the analysis is described in chapter 2 Here, I also

provide the definition and classification of birth defects A description of the current

state of birth defects in New South Wales, in terms of spatial and temporal trends is

given in this chapter

A comprehensive literature review is provided in chapter 3 Summarized components of

the literature review are included in subsequent chapters, which are structured as

manuscripts to be submitted for publication Firstly, I summarise spatial analytical

studies in relation to birth defects, to identify gaps in literature Secondly, I provide a

review on selected risk factors, with a view to inform the analytical models for a

subsequent analysis that combines both areal with individual risk factors (i.e chapter 8)

Chapter 4 introduces the CAR model I provide readers with an understanding of the

context upon which the CAR model is applied and describe the two main fields of

application: namely disease mapping and geographical correlation studies The

mathematical properties of the CAR model are also described, along with a brief section

on the adjacency matrix, which introduces a subsequent chapter which examines in

detail the impact of various neighbourhood weight matrices on the smoothing properties

of the CAR model (i.e chapter 6)

In the same chapter 4, I also discuss the strengths and limitations of the various types of

CAR models commonly used In addition, I examine the properties of spatial and

spatio-temporal models, including specific comparisons about the nature of data (sparseness of

outcome) used in the studies reviewed, along with the priors and model selection

techniques The results from these comparisons inform the modelling strategy adopted in

subsequent chapters Comparisons were made within the multivariate (i.e models

examining more than one disease outcome simultaneously) classes of models, and the

results used in chapter 7

The CAR model predominantly uses the Bayesian framework of analysis To help

readers familiarize with the context of Bayesian modeling, an introduction to Bayesian

(MAUP) and Bayesian model convergence diagnostics are discussed briefly, as these do

not relate to the main objective of the thesis

In chapter 6, I examine in detail, the effect of various choices of neighbourhood weight

matrices (ranging from adjacency to distance-based functions, as well as weights based

on key covariates) on the smoothing properties of the CAR model Addressing the issue

of sparse disease count is the focus of chapter 7, where I investigate the performance of

a CAR model with a zero-inflated Poisson extension, in terms of its ability to recover the

underlying risk surface of specific birth defects, such as Spina Bifida and Trisomy 21 I

also demonstrate how the model can be strengthened by incorporating a shared

component, via jointly modeling birth defects with a referent outcome (caesarean

counts)

Chapter 8 discusses in detail the major drawback of ecological analysis (i.e potential

ecological bias) and reviews the literature for suggested strategies to incorporate

individual-level data with areal level data, in order to minimize this potential bias

Through extensive simulation studies, I investigate the performance of various sampling

strategies, along with modifications in sample sizes, and examine how they fare for

be seen in the next 30 years from 2001 for sixteen randomly selected SLAs The

strengths and limitations of this thesis, as well as areas for future research, are the focus

of the discussion in chapter 10

In this thesis, I have excluded discussions on other seemingly related models such as

multi-level models and statistical models to analyse point process data, as my main

focus is the CAR model The aims and objectives as well as the nature of data utilised by

the other models are generally different from studies which use the CAR model, as I will

briefly describe here Multi-level models, or random effects model as they are

commonly known, are often used to study variables which can vary at more than one

level The levels can be nested hierarchicaly, and the models can be formulated within

both the frequentist and Bayesian frameworks Gelman provides details on the theory

behind these models, as well as various formulations and applications of multi-level

models(1) In the context of our spatial analysis, the CAR convolution prior (to be

discussed later in the thesis) is a more specific formulation of a multi-level model, where

the variance of the relative risk estimates is partitioned into both spatially structured and

spatially unstructured random effects

As for point-process models, one basic goal is to determine whether cases occur at

models have been used to fit stochastic epidemic models to study measles epidemics in

one study(5) Gelfand and colleagues have also used spatiotemporally varying

coefficient models to study and make predictions of climate data, such as precipitation

and temperature, which are measured at fixed locations(6) The fundamental difference

between these models and CAR models is that for the latter, data is available at an

aggregate level, as opposed to fixed locations or at continuous geographical scales

1.3 Structure of thesis

The thesis is structured in the following way It consists of a series of chapters that are

either published or submitted for publication and unpublished Chapter 6 “Addressing

the Neighbourhood Weight Matrix” has been published in the International Journal of

Health Geographics Chapter 7 “Modelling Sparse Disease Counts” has been accepted

for publication in the Health and Place Journal Chapter 8 “Strategies for Combining

Areal with Individual Data” and chapter 9 “Forecasting Birth Defects at the Small Area

Level, NSW” have been submitted to the Statistics in Medicine journal and the BMC

Health Services Research journal respectively These chapters have been included in the

same format as they were submitted for publication This explains the variations in the

way the chapters are presented, the different sub-headings used in the various chapters,

and the distinct format of the bibliographies required by the various journal The rest of

the chapters consist of unpublished works I have included the bibliographies separately

at the end of each chapter for the published works, and one overall bibliography for the

rest of the unpublished chapters at the end of the thesis

1.4 List of publications and conferences arising from thesis

Arul Earnest, Geoff Morgan, Kerrie Mengersen, Louise Ryan, Richard Summerhayes,

John Beard Evaluating the effect of neighbourhood weight matrices on smoothing

properties of Conditional Autoregressive (CAR) models International Journal of Health

Geographics, November 2007, Volume 29;6: pp 54-65

Arul Earnest, John Beard, Geoff Morgan, Douglas Lincoln, Richard Summerhayes,

Deborah Donoghue, Therese Dunn, David Muscatello, Kerrie Mengersen Small Area

Estimation of Sparse Disease Counts using Shared Component Models- Application to

Birth Defect Registry Data in New South Wales, Australia Health and Place Journal

(Accepted for publication 23 February 2010)

Arul Earnest, John Beard, Geoff Morgan, Deborah Donoghue, Therese Dunn, David

Muscatello, Danielle Taylor, Kerrie Mengersen Sampling and sample size strategies for

including individual with areal-level covariates in the spatial analysis of a sparse disease

outcome Submitted to Statistics in Medicine Journal, Oct 2009

Arul Earnest, Kerrie Mengersen, Geoff Morgan, John Beard Forecasting Birth Defects

Arul Earnest Evaluating the effect of neighbourhood weight matrices on smoothing

properties of Conditional Autoregressive models Contributed talk for Spring Bayes

27-29 September 2006, Queensland University of Technology

Arul Earnest, John Beard, Geoff Morgan, Douglas Lincoln, Richard Summerhayes,

Deborah Donoghue, Therese Dunn, David Muscatello, Kerrie Mengersen Modelling

Sparse Disease Counts Using the Shared Component Model Poster presentation at the

International Society for Bayesian Analysis, 9th World Meeting, Hamilton Island,

Australia, July 20-25 2008

Arul Earnest, John Beard, Geoff Morgan, Douglas Lincoln, Richard Summerhayes,

Deborah Donoghue, Therese Dunn, David Muscatello, Kerrie Mengersen Modelling

Sparse Disease Counts Using the Shared Component Model Poster presentation at the

National Healthcare Group Annual Scientific Congress 7-8 November 2008, Singapore

The poster won the first prize in the best poster competition for the Quality/ Health

Services Research section

CHAPTER 2 DATA

2.1 Summary

The aim of this chapter is to provide readers with an understanding of the sources of data

used in subsequent analyses in this thesis Selected birth defects are also described,

along with the classification or grouping of birth defects A background description of

current spatial and temporal trends of birth defects in New South Wales is provided as a

precursor to subsequent work in this area It is clear from existing official health

department reports that birth defects do indeed exhibit clear spatial relationships as well

as a time gradient

2.2 Sources of data

2.2.1 Birth defects

De-identified birth defect records were obtained from the NSW Birth Defects Register

(BDR) The register has been operational since 1990, and in the early years, reporting of

defects was done on a voluntary basis Since 1998, doctors, hospitals and laboratories

have been required by law to report all birth defects These defects included those

observed during pregnancy, at birth or up to one year of life Each birth defect is

recorded as a separate record, so the total number of congenital abnormalities reported is

2.2.2 Births and maternal characteristics

Information on births in NSW from 1990 to 2004 was obtained from the NSW

Midwives Data Collection (MDC), which is a population-based register just like the

BDR Covering all births in NSW (including public, private and home-births), the MDC

is dependent on the attending midwife or doctor to complete and submit a notification

form whenever a birth occurs(7).The registry includes all livebirths and stillbirths of at

least 20 weeks gestation or at least 400 grams birth weight I also obtained maternal

demographic information (e.g residential address at time of birth, maternal age at

delivery, maternal smoking during pregnancy, maternal diabetes, delivery in private

versus public hospital), pregnancy, labour, delivery and perinatal outcomes from the

MDC

Each of the birth records in NSW within the study period was geocoded (i.e given a

longitude and latitude) based on the mother’s residential address at the time of birth

This geocoding was done by Mr Richard Summerhayes from the Northern Rivers

University Department of Rural Health using geocoding software developed by the

NSW Health and Australian National University Further details on the software called

FEBRYL, can be found in this reference(8) Each record was then assigned to the 2001

Census Collectors Districts (CCDs) within which they fell in There are 11,706 CCDs in

also probabilistically linked to the MDC, and this was carried out by the Department of

Health, NSW The combined data was used in a subsequent analysis in the thesis,

involving the association between birth defects and individual maternal characteristics

along with areal covariates, such as socio-economic status of the area that the mother

was living in

In 1998 a 2% sample of Midwives Data Collection records (N=1703) was validated

against other hospital records(9) The excellent quality of this database is reflected in

high correlations, including a 99.1% agreement on gestational diabetes (kappa 0.87),

94.9% agreement for smoking in pregnancy (kappa 0.85), 96.5% agreement for

birthweight (kappa not calculated) and 84.8% agreement for gestational age (kappa

0.81) This study, and access to both BDR and MDC databases, was approved by the

New South Wales Population & Health Services Research Ethics Committee

2.2.3 Areal-level indices of socio-economic status

I used data from the Australian Bureau of Statistics (ABS) to describe the level of social

and economic well-being in areal levels of NSW This data was freely available on the

ABS website, and a technical paper can be found here (10) The following 4 indices

2 Index of Relative Social Disadvantage Higher values reflect lack of

disadvantage, which has a subtle difference from the index above The variables

that were used to compute this index included income, educational attainment,

unemployment, and dwellings without motor vehicles

3 Index of Economic Resources Variables such as income, expenditure and assets

of families, such as family income, rent paid, mortgage repayments, and dwelling

size went into computing this index

4 Index of Education and Occupation This index took into account the proportion

of people with a higher qualification or those employed in a skilled occupation

The data were available at the various Australian Standard Geographical Classification

(ASGC) levels, starting from the most basic Census Collection District (CCD) to the

Statistical Local Area (SLA) level There are problems associated with the simple

averaging up of the indices from CCD to SLA level, and I used an index that was

calculated at the SLA level and population-weighted This was performed by the ABS

Data on the four indices were standardised by the ABS to have a mean of 1000 units and

used to derive the final index score for each CD Further details on how these indices

were derived from variables obtained from the 2001 census is described in an

information paper available from the ABS(10) It should be noted that the indices

measure the socio-economic well-being of a region, and not the individual, and this

subtle difference is exemplified in a subsequent analysis presented later in the paper

2.3 Definition and classification of birth defects

A birth defect can be thought of as a physiological or structural abnormality that is

present at birth and is significant enough to be considered a problem According to the

US Centers for Disease Control and Prevention, most birth defects are thought to be

caused by a complex mix of factors including genetics, environment, and behaviors(11).

Much of the analysis for this thesis draws on data from the NSW Birth Defects Register

The Register uses the following definition for a birth defect: ‘Any structural defect or

chromosomal abnormality detected during pregnancy, at birth, or in the first year of life,

excluding birth injuries and minor anomalies such as skin tags, talipes, birthmarks, or

clicky hips(7)

the state from which the data for this study was drawn, relies on the BPA classification

system(13) that is basically organised by body system(7) Table 1 shows a list of birth

defects recorded by the Registry using this approach, together with a short description

In the United States, the Centers for Disease Control and Prevention uses a classification

system that is modified from the original BPA system(14) The key advantage of this

system is that it allows researchers to describe more specific details about the birth

defects and related conditions In particular, it describes the laterality of the defect (i.e

whether the defect was on the right or left part of the body) and provides greater

specificity for a defect One disadvantage of this approach is that the analysis of such

data becomes more challenging due to the sparseness of the defects as one becomes

more specific

Table 1 Description of selected birth defects from the NSW birth defect registry Defect Description

Anencephaly Absence of the cranial vault, with the brain tissue

completely missing or markedly reduced

Spina bifida Defective closure of the bony encasement of the spinal

cord, through which the spinal cord may protrude Encephalocele Protrusion of brain through a congenital opening in the

skull Hydrocephalus Dilatation of the cerebral ventricles accompanied by an

accumulation of cerebral fluid within the skull

Buphthalmos Enlargement and distension of the fibrous coats of the

turned outward

Polydactyly Presence of additional fingers or toes on hands or feet Syndactyly Attachment of adjacent fingers or toes on hands or feet Craniosynostosis Premature closure of the sutures of the skull

Exomphalos Herniation of the abdominal contents into the umbilical

cord

Gastroschisis A defect in the abdominal wall not involving the

umbilicus and through which the abdominal contents herniate

Cystic hygroma A sac, cyst or bursa distended with fluid

Centre for Epidemiology and Research NSW Department of Health New South Wales Mothers and Babies 2005 N S W Public Health Bull 2006; 18(S-1); pp 135

2.4 Spatial and temporal trends of birth defects in New South Wales, Australia

Across all states in Australia, there has been considerable variation in the reported rates

of birth defects over the past 20 years For instance, the rate of all malformations ranged

from 159.4 per 10,000 births (1981-1995) to about 175.2 per 10,000 births (1997) There

was also gross spatial variation in the reported rates between states in the period

1991-1997, with highest rates found in Victoria (229.2 per 10,000 births), followed by ACT

(222.3 per 10,000 births) and Queensland (194.3 per 10,000 births)(15)

However, since the criteria and source of notification varies by state, these trends may

comparisons, due to small numbers At this stage, we try not to be unduly concerned

about the reasons for spatial variation, except to note that defects exhibit both spatial and

temporal variation, even at the broader scales of analysis

In NSW, state-wide surveillance of birth defects is monitored through the Birth Defects

Register (BDR), which is administered through the NSW Department of Health The

overall rate of birth defects appears to have been stable between 1999 and 2004

However, when the defects were examined by individual diagnostic categories, there

was considerable year to year variation Ventricular septal defect, for instance, saw rates

ranging from 2.1 per 1,000 births in 2002, to 0.9 per 1,000 births in 2003 and 2.1 per

1,000 births in 2004(7, 16)

Within NSW, there was spatial variation in the reported birth defect rates for the 8

different administrative health areas between 1999 and 2005 For example, the NSW

Mothers and Babies Report 2005 found elevated rates of birth defects in the Hunter and

New England area(7) However, it should be noted that there are some issues to note

when making this sort of comparisons across regions The first involves mothers

residing near state borders and the possibly of them going interstate for treatment, where

they may nominate an interstate place of residence for the duration of treatment It is

sources of spatial variation at this stage of the project, but rather make the point that

spatial variation in reported rates of birth defects at a smaller aggregate level is inherent

in the data Health data in NSW can be grouped at various geographical scales, including

the Census Collection District (CCD) and the Statistical Local Area (SLA) The CCDs

are the smallest spatial unit, and there are 11,706 units in NSW These CCDs can be

aggregated up to broader groupings, including the SLA and Local Government Area

(LGA) In urban areas, the average number of dwellings per CCD is about 220, whereas

this number drops considerably for CCDs in rural areas There is also considerable

variation in the geographical size of these CCDs (i.e interquartile range of 1 km2 to 62

km2)

CHAPTER 3 LITERATURE REVIEW

3.1 Summary

The aim of this chapter is to provide a comprehensive literature review in a few selected

important topics Firstly, current research on the spatial analysis of birth defects is

summarized Next, I examine the prior evidence on the relationship between selected

risk factors such as maternal age, maternal smoking, maternal diabetes mellitus status

and socio-economic indicators (both areal and individual measures) and birth defects

The aim is to identify which particular defects are associated with the risk factors This

chapter also identifies risk factors that are common to birth defects and caesarean rates,

as well as describing spatial variation in caesarean rates The results are used in

subsequent chapters examining the risk factors for birth defects, as well as the joint

modeling of two related outcomes

The search strategy used to identify studies for discussion in this chapter is described

here I searched for all relevant research articles in MEDLINE, which contains

bibliographic citations and author abstracts from more than 4,000 biomedical journals

published in the United States and 70 other countries The PubMed on-line search

engine tool was used for this purpose In addition, I also went through the bibliographic

lists of relevant journal articles to identify additional pieces of research to include in my

that were performed in the laboratories, as well as those not published in English For

the section on risk factors for birth defects, I used the following search terms For

maternal age, as an example, I used “maternal age” and “birth defect”, “age” and “birth

defect”, as well as “risk factor” and “birth defect” more generally For caesarean rates,

the following key-words were used: “caesarean” and “risk factor”, “caesarean” and

spatial”, as well as “caesarean” and “geographic” I would like to add that this was not a

systematic review exercise, and hence I did not provide a summary of the results from

the literature Rather, the studies identified in the literature were used to justify the use

of the selected risk factors, in studying their association with birth defects in my thesis

3.2 Spatial analysis of birth defects

In many countries, information on birth defects is obtained and analysed from national

or regional birth defect registries These registries often have data on the mother’s

residential address This location data enables researchers to undertake various forms of

spatial analyses on the epidemiology of birth defects Application of spatial analyses

ranges from simple mapping of defects, to identifying clusters and exploring the

influence of environmental factors, such as air pollution, contaminated sites, disinfection

byproducts from water chlorination on the occurrence of birth defects Ecological

Few studies have examined the spatial distribution of birth defects and their association

with possible spatially varying risk factors High altitude has been implicated in at least

three studies in South America In the first, looking at 53 hospitals across Latin

America(17), adjusted relative risks were found to be significantly higher among those

living in the highland, specifically for cleft lip (RR=1.57, 95%CI: 1.27-1.94), microtia

(RR=3.21, 95%CI: 2.35-4.79), preauricular tag (RR=2.09, 95%CI: 1.86-2.36), branchial

arch anomaly complex (RR=1.79, 95%CI: 1.23-2.61), constriction band complex

(RR=1.92, 95%CI: 1.11-3.31) and anal atresia (craniofacial defects) (RR=1.61, 95% CI:

1.01-2.57) On the other hand, risks were lower for spina bifida (RR=0.57, 95% CI:

0.37-0.78), anencephaly (RR=0.33, 95% CI: 0.20-0.54), hydrocephaly (RR=0.41, 95%

CI: 0.22-0.77) and pes equinovarus (neural tube defects) (RR=0.70, 95%CI: 0.51-0.96)

The second study also linked altitude with the risk of microtia, with a relative risk of

2.66 (p<0.01) comparing those living more than 1000m above sea-level versus those

living less than 500m(18), whilst the third study(19) found that cleft lip/ palate birth

prevalence rates were higher for those living at high altitude above sea-level (effect size

not provided)

In another study, researchers examined indicators of exposure to industrial activities(20)

and found significant associations between textile industry and anencephaly (RR=1.59,

relative risk of having the second infant with the same defect was higher among those

who lived in the same municipality (RR=11.6, 95%CI: 9.3-14.0) during both

pregnancies, as compared to those who moved to another municipality (RR=5.1, 95%CI:

3.4-6.7)(21) In contrast, the second study did not find any significant change in the

frequency of facial-cleft defects among mothers who changed municipality of residence

(RR=0.9, 95%CI: 0.6-1.5)(22)

In studies that examined the spatial variation (in particular geographical difference) in

risks of specific birth defects, neural tube defects seemed to be the most common defect

that was found to be spatially correlated(23-28), followed by clefts(19, 29),

anophthalmia and microphthalmia(30), where prevalence was found to be higher in rural

versus urban areas as well as diaphragmatic hernia and gastroschisis(27) Birth defects

were also found to vary by the level of aggregation: e.g across register areas and

hospital catchments, but not below this level(31) A study from England found variation

in reported rates by local register and hospital catchment area (p<0.001), but not by area

deprivation scores (p>0.1, effect size not provided)(32) Proximity of maternal residence

to landfill sites was found to be associated with certain birth defects such as neural tube

defects (RR=1.05, 95%CI: 1.01-1.10), hypospadias and epispadias (RR=1.07, 95%CI:

nonsyndromic cleft lip/ palate across public health region of residence across Texas,

USA(38)

3.3 Risk factors for birth defects

There have been numerous studies that have looked at the association between various

risk factors and the occurrence of congenital abnormalities at birth Study designs have

included case-control (including the use of sick controls), cohort studies, and more

commonly the use of birth defect registries to explore associations with risk factors

Variables of interest have included socio-demographic covariates, genetic, nutritional,

infectious, and other environmental factors It is not the aim of this thesis to undertake

an evaluation of the various risk factors, as this has been carried out by other authors

Here, I discuss specific risk factors that were commonly found in the literature and

available to us for analysis, and these include maternal age, maternal smoking, maternal

diabetes status and socio-economic status I also examine studies that have analysed the

defects spatially/ geographically, as this project’s interest is in studying risk factors from

the spatial perspective

3.3.1 Maternal age at delivery

Maternal age is the most commonly studied risk factor in most studies involving birth

Younger mothers have been shown to have a higher risk of giving birth to babies with

gastroschisis(39-43), chromosomal abnormalities(43, 44), cystic hygroma, autosomal

recessive disorders, monogenic disorders, ventricular septal disorders(43), anencephaly,

all ear defects, female genital defects, polydactyly, omphalocele(42) Older mothers, on

the other hand, are known to have a higher risk of having babies with various types of

atresia(30, 42), anophthalmia, microphthalmia(45), heart defects, right outflow tract

defects, males genital defects including hypospadias, craniosynostosis(42), trisomy 18,

trisomy 21, dysplasia of hip, chromosomal abnormalities(43, 46), pancreas, down

syndrome(47) and cleft lip/ palate(48)

Some defects like chromosomal abnormalities are related to risk factors, associated with

both older and younger mothers, as we can see from above Neural tube defects also

display such a U-shaped relationship with maternal age (26, 28, 49, 50) On the other

end of the spectrum, other studies have shown that there is no significant relationship

between maternal age and defects such as cleft palate and lip(29, 51, 52) along with

ventricular septal defect(53), severe birth defects(54) and anencephalus and spina

bifida(55)

defects, persistent ductus arteriosus(58), isolated craniosynostosis(59), kidney

malformations(60), oral clefts(48, 61-66), limb reduction birth defects(67), deformities

of the foot(68), gastroschisis(40, 41, 69, 70), defects of the cardiovascular system(71),

hydrocephaly, polydactyly/ syndactyly/ adactyly(69), clubfoot(69, 72, 73) and defects in

general(47)

Most of the studies have involved the case-control study design and included data from

birth defect registries However, there were two meta-analyses (74, 75) that looked at the

combined (across studies) effect of maternal smoking on the occurrence of oral cleft

birth defects Both studies found a relationship between smoking and risk of oral clefts

Some studies (48, 62) managed to show a dose-response relationship between smoking

and birth defects, a sign of the causality criteria met for an observational study (48, 59,

62, 65)

The use of self-reported smoking as a risk factor variable in epidemiological studies has

been criticized because of the possibility of recall bias This is especially true in birth

defect studies, which use normal controls (i.e mothers with babies without any birth

defect who may not be as motivated to accurately recall past behaviours) In the

literature that we reviewed, we found some studies that used affected or sick controls

oral clefts, musculoskeletal malformations(78), conotruncal heart defects, limb

deficiencies(77, 79), Down’s syndrome(80), and nonsyndromic oral cleft(55)

3.3.3 Socio-economic indicators

There is growing evidence that socio-economic disadvantage is associated with higher

risk of a range of adverse health outcomes Epidemiological research in this field is often

hampered by difficulties in eliciting data on socio-economic status from study

participants This is particularly because such information may be sensitive to study

participants, or not be collected by routine data collection systems Consequently,

researchers often resort to using a wide variety of surrogate indicators of socio-economic

status, including occupation, income, race, education, health insurance type, etc

Gonzalez provides a systematic review of studies that looked at the relationship between

socio-economic status (as measured by education or occupation) and ischemic heart

disease (IHD), and they report a clear relationship between the two variables and risk of

IHD (81) A strong inverse relationship between socioeconomic status (SES) and risk of

cardiovascular disease and mortality has also been highlighted in another study (82)

Sometimes, socioeconomic information may not be collected from the individual

substitute for individual level information and any bias introduced by this approach is

likely to lead to conservative estimates of association (83)

These areal measures are often derived from data available from the census or other

ad-hoc independent surveys Composite indices measuring some form of disadvantage are

then formed using statistical techniques such as factor analysis or principal component

analysis The areal measures are postulated to measure contextual socio-economic

effects of a person’s residential area I reviewed literature on birth defects to examine the

various types of socio-economic status measured and their subsequent relationship to

specific birth defects As I will show in a subsequent chapter, the effect of

socio-economic status on the occurrence of birth defects seems to operate at both the

individual and areal-level, and complex statistical models are needed to incorporate this

hierarchical structure in the data (i.e data measured at disparate scales)

Among the various individual-level socio-economic indicators, occupation(20, 84-90)

seems to be the most commonly studied covariate, followed by education(26, 28, 48, 84,

85, 91, 92), race(23, 38, 50, 93-95), income(49, 70, 84) and insurance status(26, 47)

Areal-level measures of socio-economic status have been evaluated in a number of

studies of occurrences of birth defects(23, 30, 85, 96-101)

having babies with the defect For facial clefts, education (48, 85) and occupation(85,

87, 88) were the two main covariates implicated Areal-level measures were shown to be

related to neural tube defects in a number of studies (23, 97, 99) along with facial clefts

(85, 100, 101) Two particular studies evaluated both individually measured

socio-economic status, along with areal measures The results were mixed The first study

found a significant effect of lower individual socio-economic status and residence in a

SES-lower neighbourhood on the occurrence of neural tube defects (OR=1.7, 95%CI:

1.1-2.5) This was when we looked at maternal employment and neighbourhood

unemployment as an indicator of SES (99) The other study (101) found an increased

risk of spina bifida (OR=2.3, 95%CI: 1.0-5.5) and cleft palate (OR=2.3, 95%CI: 1.4-3.8)

with a household SES index, but not with an individual SES measure such as maternal

unemployment, with an OR=1.2, 95%CI: 0.7-2.2 for spina bifida and OR=1.2, 95%CI:

0.9-1.7 for cleft palate respectively

3.3.4 Maternal diabetes mellitus

There are two general types of diabetes(102) Diabetes type 1 is when the body produces

too little insulin that the body can’t make use of blood sugar for energy Type 2 diabetes

happens when the body makes too little insulin or is unable to use the insulin to produce