Applied Sciences

Hodgson S, Lutz PW, Shirley MD, Bythell M, Rankin J. Exposure misclassification due to residential mobility during pregnancy. International Journal of Hygiene and Environmental Health 2015, 218(4), 414-421.

Copyright:

© 2015 The Authors. Published by Elsevier GmbH. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/)

DOI link to article:

http://dx.doi.org/10.1016/j.ijheh.2015.03.007

Date deposited:

12/06/2015

This work is licensed under a Creative Commons Attribution 4.0 International License

Newcastle University ePrints – eprint.ncl.ac.uk

E d

S M a

b

c

d

e

a

A R R A

K E M R E E

I

u i i m s t c

m N

l m

h 1

International Journal of Hygiene and Environmental Health 218 (2015) 414–421

Contents lists available at ScienceDirect

International Journal of Hygiene and Environmental Health

jou rn al hom ep age: www.elsev ier .com/ locate / i jheh

xposure misclassification due to residential mobility uring pregnancy

usan Hodgsona,b,∗, Peter W.W. Lurzc,d, Mark D.F. Shirleyc, ary Bythell e, Judith Rankinb,e

MRC-PHE Centre for Environment & Health, Department of Epidemiology & Biostatistics, Imperial College London, United Kingdom Institute of Health & Society, Newcastle University, Newcastle upon Tyne, United Kingdom School of Biology, Newcastle University, Newcastle upon Tyne, United Kingdom Royal (Dick) School of Veterinary Studies, University of Edinburgh, Edinburgh, United Kingdom Regional Maternity Survey Office, Newcastle upon Tyne, United Kingdom

r t i c l e i n f o

rticle history: eceived 26 September 2014 eceived in revised form 13 March 2015 ccepted 18 March 2015

eywords: nvironmental exposure aternal exposure

esidential mobility xposure assessment xposure error

a b s t r a c t

Objectives: Pregnant women are a highly mobile group, yet studies suggest exposure error due to migra- tion in pregnancy is minimal. We aimed to investigate the impact of maternal residential mobility on exposure to environmental variables (urban fabric, roads and air pollution (PM10 and NO2)) and socio-economic factors (deprivation) that varied spatially and temporally. Methods: We used data on residential histories for deliveries at ≥24 weeks gestation recorded by the Northern Congenital Abnormality Survey, 2000–2008 (n = 5399) to compare: (a) exposure at conception assigned to maternal postcode at delivery versus maternal postcode at conception, and (b) exposure at conception assigned to maternal postcode at delivery versus mean exposure based on residences throughout pregnancy. Results: In this population, 24.4% of women moved during pregnancy. Depending on the exposure vari-

able assessed, 1–12% of women overall were assigned an exposure at delivery >1SD different to that at conception, and 2–25% assigned an exposure at delivery >1SD different to the mean exposure throughout pregnancy. Conclusions: To meaningfully explore the subtle associations between environmental exposures and health, consideration must be given to error introduced by residential mobility.ors. P

© 2015 The Authntroduction

Epidemiological studies carried out at the ecological level, or sing routinely collected health data, often assign exposure to an

ndividual’s residence at a single time point, such as birth, hospital- sation or death. This approach fails to account for individuals who

ight have migrated into or out of the population or for periodic pells away from a residence where levels of exposure are likely

o be different from those experienced at home. Such migrations ould result in exposure error or misclassification, reduced study∗ Corresponding author at: MRC-PHE Centre for Environment & Health, Depart- ent of Epidemiology & Biostatistics, Imperial College London, St. Mary’s Campus, orfolk Place, London W2 1PG, United Kingdom. Tel.: +44 020 7594 2789.

E-mail addresses: susan.hodgson@imperial.ac.uk (S. Hodgson), urzpww@gmail.com (P.W.W. Lurz), mark.shirley@ncl.ac.uk (M.D.F. Shirley),

arybythell@gmail.com (M. Bythell), judith.rankin@newcastle.ac.uk (J. Rankin).

ttp://dx.doi.org/10.1016/j.ijheh.2015.03.007 438-4639/© 2015 The Authors. Published by Elsevier GmbH. This is an open access artic

ublished by Elsevier GmbH. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).

power, and may result in biased risk estimates (Armstrong, 1998; Blair et al., 2007; Khoury et al., 1988).

Many environmental epidemiological studies of birth outcomes assign a measure of exposure based on maternal residential loca- tion at delivery because this information is readily available. The relatively short period between exposure and disease manifesta- tion should mean that studies on congenital anomalies are less prone to migration bias, as there is less time in which the pop- ulation can migrate. However, there is now a significant body of literature showing that pregnant women are a highly mobile group, with 10–30% of women moving residence during pregnancy (Bell and Belanger, 2012; Canfield et al., 2006; Fell et al., 2004; Hodgson et al., 2009; Khoury et al., 1988; Shaw and Malcoe, 1992; Zender et al., 2001).

Theoretical papers on the implications of residential mobility

during pregnancy on the ability to detect environmental terato- gens (Khoury et al., 1988) and impacts of differential mobility (Schulman et al., 1993) remain relevant, and a study showing the impact of mobility on real-life exposure scenarios and onle under the CC BY license (http://creativecommons.org/licenses/by/4.0/).

iene and Environmental Health 218 (2015) 414–421 415

e i o e i v u 2 a 2 i a p

M

p n n a o b a c g p g g c c a e s v s b f t t d g s w

a d t N u N A c ( w o t a p o l p e d e

Table 1 Social and environmental variables assigned to maternal residential postcodes to explore the impact of residential mobility during pregnancy on characterisation of exposure.

Socio-economic status 1. Index of Multiple Deprivation at Super Output Area level

Data source Office for National Statistics Variable type Continuous and quintile, area level Spatial resolution Super Output Area Temporal resolution n/a (data for 2007 used for whole study period)

2. Index of Multiple Deprivation at Local Authority level Data source Office for National Statistics Variable type Continuous and quintile, area level Spatial resolution Local Authority Temporal resolution n/a (data for 2007 used for whole study period)

Land cover 3. % Continuous Urban Fabric within 500 m buffer of postcode

Data source CORINE land cover 2000v8a

Variable type Continuous and dichotomous, individual level Spatial resolution 100 m Temporal resolution n/a (data from 2000 used for whole study period)

4. % Discontinuous Urban Fabric within 500 m buffer of postcode Data source CORINE land cover 2000v8a

Variable type Continuous and quintile, individual level Spatial resolution 100 m Temporal resolution n/a (data from 2000 used for whole study period)

Roads 5. Total length (m) of roads (motorways, A and B roads) within 500 m buffer of postcode

Data source Strategi 2011b

Variable type Continuous and quintile, individual level Spatial resolution 1 m Temporal resolution n/a (data from 2011 used for whole study period)

Air pollution 6. Annual background PM10

Data source DEFRA Ambient Air Quality Assessment (UKAAQA)c

Variable type Continuous and quintile, individual level Spatial resolution 1 km grid square Temporal resolution Annual mean, 2001–2008

7. Daily NO2 Data source DEFRA Automatic Urban and Rural Networkd

Variable type Continuous and quintile, individual level Spatial resolution Nearest monitor (for those living within 15 km of a

monitor) Temporal resolution Daily mean (averaged over first trimester),

2000–2008

a http://www.eea.europa.eu/data-and-maps/data/corine-land-cover-2000- clc2000-100-m-version-8-2005.

b www.ordnancesurvey.co.uk/oswebsite/docs/user-guides/strategi-user- guide.pdf.

S. Hodgson et al. / International Journal of Hyg

nvironmental risk factors likely to confer small, but important ncreases in risk, is overdue. In this paper we investigate the impact f residential mobility during pregnancy on the measurement of xposure to a range of environmental factors previously explored n aetiological research (for example area-level measures of depri- ation (Dibben et al., 2006; Janevic et al., 2010), land cover (e.g. rban/rural classifications) (Hillemeier et al., 2007; Langlois et al., 010), road density/proximity to roads (Yorifuji et al., 2011) and ir pollutants (Dugandzic et al., 2006; Hansen et al., 2009; Xu et al., 011)), and quantify the exposure error likely to be introduced

nto a study reliant on maternal residential location at delivery s a proxy for residential location at conception and throughout regnancy.

aterials and methods

The Northern Congenital Abnormality Survey (NorCAS) is a rospective, population-based registry covering the former UK orthern health region, which includes north east England and orth Cumbria (Fig. 1). This region comprises a population of bout three million, with approximately 32,000 births each year ver the study period 2000–2008, of which approximately 826 irths each year (2.6%) included a major congenital anomaly nd were therefore recorded in NorCAS. Data are collected on ongenital anomalies occurring in late miscarriages (>20 weeks estation), in live births and stillbirths, and in terminations of regnancy for foetal anomaly after prenatal diagnosis at any estation. The NorCAS follows the European Surveillance of Con- enital Anomalies guidelines for inclusion on the register and lassification of anomalies (see http://www.eurocat-network.eu/ ontent/EUROCAT-Guide-1.3-Chapter-3.3-Jan2012.pdf) and codes nomalies according to the WHO International Classification of Dis- ases version 10. Cases are reported to the register from multiple ources to ensure a high case ascertainment, as described pre- iously (Boyd et al., 2005; Richmond and Atkins, 2005). For this tudy, data on all pregnancies with a congenital anomaly delivered etween 01 January 2000 and 31 December 2008 were extracted rom NorCAS, although this dataset was subsequently restricted to hose with a gestation at delivery of ≥24 weeks (a viable delivery), o allow better comparison with pregnancies resulting in a healthy elivery. If more than one baby in a multiple pregnancy has a con- enital anomaly, each case is included on NorCAS. However, for this tudy, the pregnancy was counted as the ‘case’ so each pregnancy as counted only once.

The NorCAS contains addresses for women at both booking ppointment (average gestational age 13 weeks in the UK) and elivery. To obtain more detailed information on residential his- ory, the NorCAS data were linked to the UK National Health Service ational Strategic Tracing Service records. Linkage was achieved sing several data fields, including the mother’s date of birth, ational Health Service number, surname and residential postcode. ddress at delivery was confirmed and updated as required. Date of onception was calculated from the date and gestation at booking available within the NorCAS), and address details at this date, as ell as any other residences during the index pregnancy (with dates

f when the women moved to and from this address) available from he National Strategic Tracing Service were extracted to provide ddress at conception, and enable residential history throughout regnancy to be established. All addresses were geocoded based n the address postcode centroid, the geographic centre of a col- ection of approximately 15 adjacent households making up the

ostcode. Within the study area the average distance between near- st neighbouring postcodes was 104 m, max 6.2 km, though this istance varied considerably between urban and rural areas (for xample, in Newcastle Local Authority (a predominantly urbanc http://uk-air.defra.gov.uk/data/pcm-data. d http://uk-air.defra.gov.uk/networks/network-info?view=aurn.

area) the average distance was 49 m, max 1.16 km, in contrast in Tynedale (rural authority) the average distance was 255 m, max 7.95 km). Grid references were obtained from the Office for National Statistics Postcode Directory (http://edina.ac.uk/ukborders/).

To establish the impact of residential mobility during pregnancy on exposure classification, we assigned to each woman’s postcode at delivery and conception a measure of exposure to a variety of environmental factors, and, based on residential history, a mea- sure of mean exposure throughout pregnancy weighted according to proportion of the pregnancy spent at each postcode. These vari- ables include typical environmental factors explored in aetiological epidemiological research. We deliberately chose factors that were (a) readily available, (b) varied in terms of their spatial and/or tem- poral resolution, and (c) able to be assigned at the individual and/or area level. These variables are described in Table 1.

For deprivation, we used the 2007 Index of Multiple Depriva-

tion, which comprises 38 indicators of deprivation spread across seven domains (income deprivation; employment deprivation; health deprivation and disability; education, skills and training

416 S. Hodgson et al. / International Journal of Hygiene and Environmental Health 218 (2015) 414–421

F ty Surv U

d m e c A 1 S c

p l a s w I c c c b f ( t a p p

S w a o

ig. 1. Map showing the geographic coverage of the Northern Congenital Abnormali rban and Rural Network NO2 monitors (black triangles).

eprivation; barriers to housing and services; living environ- ent deprivation; and crime) (Noble et al., 2008). This index was

xtracted at the area level for (1) lower layer Super Output Areas, a ensus based unit with a mean population of 1500, and (2) at Local uthority level, an administrative unit with a mean population of 40,000. We assigned to each postcode the deprivation score of the uper Output Area or Local Authority that contained that postcode entroid.

The CORINE land cover classes are discriminated mainly by hysiognomic attributes (shape, size, colour and pattern) of

andscape objects (natural, modified, cultivated and artificial), s recorded on satellite images (de Lima, 2005). The smallest urfaces mapped correspond to 25 ha, and the scale of the output as fixed at 1:100,000, giving a location precision of 100 m.

n the Continuous urban fabric land class, most of the land is overed by buildings, roads and artificially surfaced areas which over almost all the ground. The Discontinuous urban fabric land lass is also characterised by most of the land being covered y structures, but here the buildings, roads and artificially sur- aced areas are associated with vegetated areas and bare soil www.eea.europa.eu/publications/COR0-part1/download). Con- inuous urban fabric (3) and Discontinuous urban fabric (4) were ssessed at the individual level; we assigned to each postcode the roportion of each land cover class within a 500 m buffer of the ostcode centroid.

For roads (5), assessed at the individual level, we used OS

trategi data to assign to each postcode the total metres of motor- ays, A roads (large-scale transport links within or between areas)nd B roads (which feed traffic between A roads and smaller roads n the network) within a 500 m buffer of the postcode centroid.

ey (NorCAS) (shaded area), and inset showing the locations of the DEFRA Automatic

For annual background particulate matter (particles less than 10 �m in diameter (PM10)) (6), we used DEFRA Ambient Air Quality data (background pollution maps at 1 km × 1 km resolu- tion) to assign to each postcode at conception and delivery the annual mean PM10 concentration for year of conception. To calcu- late the mean PM10 exposure through pregnancy, each postcode was assigned the annual mean(s) for the year(s) of residence, which were then weighted according to proportion of the pregnancy spent at each postcode.

Nearest monitor daily mean nitrogen dioxide (NO2) concentra- tions (7), a variable with limited spatial variability due to the small number of monitors across the study region (at conception, six sites in the north east provided data for 98.4% of the women, see Fig. 1), was assessed at the individual level (for those women living within 15 km of a monitor (an arbitrary cut-off)). We used DEFRA Auto- matic Urban and Rural Network data (the main network used for compliance reporting) to assign to each postcode at delivery and conception the mean NO2 exposure for the first trimester (first 90 days of each pregnancy), as well as mean exposure throughout pregnancy based on residential history.

The level of agreement between (a) exposure at conception assigned to postcode at delivery versus postcode of conception, and (b) exposure at conception assigned to postcode at delivery versus mean exposure throughout pregnancy based on residential history, was assessed by a range of measures. These included: (i) as continuous variables using Pearson correlation co-efficient (R),

(ii) as quintiles using Cohen’s kappa co-efficient (K) to take into account agreement occurring by chance, with quintiles based on equal percentiles at conception/mean exposure throughout preg- nancy, apart from continuous urban fabric which, due to granularity

S. Hodgson et al. / International Journal of Hygiene and Environmental Health 218 (2015) 414–421 417

Table 2 Agreement between exposures (a) at conception, assigned to postcode at delivery versus postcode at conception, and (b) at conception assigned to postcode at delivery versus mean through pregnancy based on residential history, for all women, non-movers and those who moved during pregnancy.

Variable All women Non-movers Movers

n R K Accuracy n R K Accuracy n R K Accuracy

1. Super Output Area Deprivation Score a) Delivery versus conception 5391 0.89 0.83 0.91 4078 1 1 1 1313 0.57 0.30 0.62 b) Delivery versus mean through pregnancy 5396 0.96 0.90 0.95 4078 1 1 1 1318 0.86 0.59 0.81

2. Local Authority Deprivation Score a) Delivery versus conception 5391 0.92 0.94 0.97 4076 1 1 1 1313 0.69 0.75 0.89 b) Delivery versus mean through pregnancy 5393 0.98 0.96 0.98 4076 1 1 1 1317 0.90 0.84 0.94

3. % Continuous Urban Fabric a) Delivery versus conception 5399 0.81 0.83 0.96 4080 1 1 1 1319 0.33 0.32 0.84 b) Delivery versus mean through pregnancy 5399 0.94 0.91 0.97 4080 1 1 1 1319 0.75 0.68 0.89

4. % Discontinuous Urban Fabric a) Delivery versus conception 5399 0.84 0.79 0.90 4080 1 1 1 1319 0.34 0.19 0.59 b) Delivery versus mean through pregnancy 5399 0.95 0.88 0.94 4080 1 1 1 1319 0.79 0.49 0.77

5. Metres roads within 500 m a) Delivery versus conception 5399 0.83 0.80 0.88 4080 1 1 1 1319 0.36 0.20 0.52 b) Delivery versus mean through pregnancy 5399 0.94 0.86 0.93 4080 1 1 1 1319 0.78 0.44 0.71

6. Annual PM10 a) Delivery versus conception 4396 0.95 0.91 0.96 3290 1 1 1 1106 0.81 0.65 0.85 b) Delivery versus mean through pregnancy 4396 0.88 0.64 0.84 3290 0.90 0.66 0.86 1108 0.81 0.55 0.80

7. NO2 a) Delivery versus conception 3373 0.95 0.98 0.98 2571 1 1 1 802 0.81 0.89 0.92 b) Delivery versus mean through pregnancy 3365 0.74 0.27 0.75 2571 0.76 0.26 0.75 794 0.67 0.28 0.73

R = Pearson correlation co-efficient, with exposures assessed as continuous variables. K rban A f expo

i o p c w a

m p I e p t n s

I

E

t ( f t (

R

7 t 2 m G S g t

= Cohen’s kappa co-efficient, with exposures as quintiles, apart from continuous u ccuracy = exposure at delivery assumed correct if within one standard deviation o

n the data, was explored as a dichotomous variable (i.e. exposed r not exposed to any continuous urban fabric within 500 m of ostcode), or (iii) assessed for accuracy, where exposure at con- eption assigned to delivery postcode was assumed ‘correct’ if it as within one standard deviation (SD) of the exposure assigned

t conception postcode/mean exposure throughout pregnancy. To explore the likelihood of introducing differential exposure

isclassification, independent sample t-tests were used to com- are mean exposure at conception for non-movers versus movers.

n addition, paired sample t-tests were used to compare mean xposure at conception assigned to postcode of delivery versus ostcode of conception, and mean exposure at conception assigned o postcode of delivery versus mean exposure throughout preg- ancy based on residential history. p Values <0.05 were taken as tatistically significant.

Data were linked in GIS ESRI ArcMap 10.0 and analysed using BM SPSS Statistics Version 20.

thical approval

The NorCAS, as part of the British Isles Network of Congeni- al Anomaly Registers, has National Information Governance Board now Health Research Authority) exemption from a requirement or consent for inclusion on the register under section 251 of he National Health Service Act (2006) and has ethics approval 09/H0405/48) to undertake studies involving the use of its data.

esults

NorCAS registered 7432 deliveries during 2000–2008. Of these, 231 (97.3%) were able to be linked to women represented in he National Strategic Tracing Service data, with the remaining 01 deliveries not able to be linked, likely due to missing or mis- atched data, or due to their mother not being registered with a

P and therefore not appearing in the National Strategic Tracing ervice dataset. Postcode at conception and delivery was able to be eocoded for 6972/7432 deliveries (93.8%). When further restricted o represent pregnancies with a gestational age at delivery of ≥24fabric which was explored as a dichotomous variable. sure assigned at conception/throughout pregnancy.

weeks (a viable delivery), 5399 (72.7%) pregnancies remained. Of these, 1319 women (24.4%) moved during pregnancy. With respect to the timing of moves, the mean number of days after gestation before the first move was 112 days (16 weeks); a little over half of the women who moved (686/1319; 52%) did so during their first trimester, 378 (28.7%) moved during their second trimester, and 255 (19.3%) moved during their third trimester. The mean and median moving distance amongst movers were short, at 19.26 and 1.85 km respectively, with 72.5% of women moving within 5 km.

When looking at all women, the majority of whom did not move, there was, as expected, good agreement between (a) expo- sure at conception assigned to postcode at delivery versus postcode at conception, and (b) exposure at conception assigned to post- code at delivery versus mean exposure throughout pregnancy based on residential history (Table 2). The level of agreement was similar when variables were assessed using Pearson correlation co- efficient (R), Cohen’s kappa co-efficient (K) or assessed for accuracy (i.e. within one standard deviation (SD)). For the air quality vari- ables PM10 (6) and NO2 (7), which exhibit temporal variability, the agreement between exposure at delivery and mean through- out pregnancy was weaker, likely due to the underlying temporal trends in pollution levels, which showed a decline over the time period studied. For women who moved during pregnancy, the agreement between exposures at conception assigned to postcode at delivery versus conception, or postcode at delivery versus mean exposure throughout pregnancy was much weaker.

The relatively good agreement, overall, between exposures at conception assigned to delivery versus conception postcode, and at delivery postcode versus residences throughout pregnancy, hides the fact that, at the individual level, substantial differences in expo- sure do occur.

Order now and get 10% discount on all orders above $50 now!!The professional are ready and willing handle your assignment.

ORDER NOW »»