No excess risk of adverse birth outcomes in populations living near special waste landfill sites in Scotland

 SMJ 2003 48(4): 105-107

Sara E Morris, Andrew OW Thomson, Lars Jarup, Cornelis de Hoogh, David J Briggs, Paul Elliott. The Small Area Health Statistics Unit (SAHSU), Department of Epidemiology and Public Health, Imperial College London, Norfolk Place, London W2 1PG, UK.

Correspondence and reprint requests should be addressed to Lars Jarup, Small Area Health Statistics Unit, Department of Epidemiology and Public Health, Imperial College London, Norfolk Place, London W2 1PG, UK (e-mail: l.jarup@imperial.ac.uk).

Funding and acknowledgements The Small Area Health Statistics Unit is funded by a grant from the Department of Health, Department of the Environment, Food and Rural Affairs, Environment Agency, Health and Safety Executive, Scottish Executive, Welsh Assembly for Government and Northern Ireland Department of Health, Social Services and Public Safety.  We thank the Information and Statistics Division of the Scottish Health Service for providing data on congenital anomalies, births, stillbirths, and hospital admissions. We thank the Scottish Environment Protection Agency for providing data on landfill and for their help in resolving discrepancies.

Competing interests None.

Abstract

A recent study showed small excess risks of low birth weight, very low birth weight and certain congenital anomalies in populations living near landfill sites in Great Britain. The objective of the current study was to investigate the risk of adverse birth outcomes associated with residence near special waste landfill sites in Scotland.

We studied risks of adverse birth outcomes in populations living within 2 km of 61 Scottish special waste landfill sites operational at some time between 1982 and 1997 compared with those living further away. 324,167 live births, 1,849 stillbirths, and 11,138 congenital anomalies (including terminations) were included in the study.  Relative risks were computed for all congenital anomalies combined, some specific anomalies and prevalence of stillbirth and low and very low birth weight (< 2500 g and < 1500 g).

For all anomalies combined, relative risk of residence near special waste landfill sites was 0.96 (99% confidence interval 0.89 to 1.02) adjusted for confounders. Adjusted risks were 0.71 (0.36 to 1.42) for neural tube defects, 1.03 (0.85 to 1.26) for cardiovascular defects, 0.84 (0.58 to 1.22) for hypospadias and epispadias (with no excess of surgical corrections), 0.78 (0.27 to 2.23) for abdominal wall defects (1.32 (0.42-4.17) for hospital admissions), 1.22 (0.28 to 5.38) for surgical correction of gastroschisis and exomphalos and 1.01 (0.96 to 1.07) and 1.01 (0.90 to 1.15) for low and very low birth weight respectively. There was no excess risk of stillbirth.

In conclusion, we found no statistically significant excess risks of congenital anomalies or low birth weight in populations living near special waste landfill sites in Scotland.

Keywords   Special waste landfill sites, congenital anomalies, low birth weight

Introduction

Waste disposal by landfill is common in many countries[i], accounting for over 80% of municipal waste in the UK[ii]. Potentially toxic chemicals in landfill include volatile organic compounds[iii], pesticides, solvents and heavy metals[iv]. A diverse mixture of gases may be produced, dispersing either through the atmosphere or via subsurface pathways³. Dusts may also be released, especially during site operation and restoration.  Human exposure could result from contact with contaminated air or soil^{3 [v]}, leaching or runoff [vi], or dispersion by animals or birds, although the evidence for any significant exposures is largely lacking.[vii]

            Excess risks of certain congenital anomalies and low birth weight near landfill sites have been reported by several investigators[viii] [ix] [x] [xi] [xii]. Recently, Elliott et al[xiii] reported small excess risks of low and very low birth weight (<2500 g and <1500 g respectively) and certain congenital anomalies in populations living within 2 km of landfill sites in England, Scotland and Wales. For all congenital anomalies combined, there was a 7% excess risk near special waste sites. The current study focuses on the risk of anomalies within 2km of special waste sites in Scotland.

Methods

The reference and study areas were classified as in Elliott et al¹³; the reference area was defined by postcodes that were not within 2km of any type of landfill site, special waste or otherwise, whether operating or closed, in Scotland.  Postcodes were classified hierarchically year on year by the operational status of the special waste landfill sites within 2km, with sites still operating taking precedence over those closed earlier in the study period, which took precedence over those opening later.  The ‘still operating’ and ‘closed’ groups form our ‘exposed’ area.

Health and denominator data. We used national postcoded registers held by the UK Small Area Health Statistics Unit: congenital anomaly and terminations data for Scotland, 1988-1994; hospital admissions data (Scottish Hospital In-Patient Statistics), 1993-1995; national births and stillbirths data, 1983-1998. A deprivation score was obtained by assigning postcodes to tertiles of the Great Britain distribution of the Carstairs’ deprivation index[xiv] based on 1991 census statistics at Enumeration District level. 

Cases were coded to the International Classification of Diseases (ICD) version 9 from 1988 to 1994, and to version 10 thereafter. Outcomes were all congenital anomalies combined (ICD9 740-59, ICD10 Q00-Q99), neural tube defects (ICD9 740.0-740.2, 741.0-741.9, 742.0; ICD10 Q00.0-Q00.2, Q05.0-Q05.9, Q01.0-Q01.9); cardiovascular defects (ICD9 745.0-747.9; ICD10 Q20.0-Q28.9); abdominal wall defects (ICD9 756.7; ICD10 Q79.2-Q79.4); hypospadias and epispadias (ICD9 752.6; ICD10 Q54.0-Q54.9, Q64.0); surgical correction of hypospadias and epispadias (M731, M732) and surgical correction of gastroschisis and exomphalos (T281). Multiple anomalies were counted under each outcome (once only for all anomalies combined).

Surgical corrections were analysed by date of birth, not date of surgical procedure — for hypospadias and epispadias, we included only procedures carried out before the age of 3, and for gastroschisis and exomphalos, in the first year of life only. Low and very low birthweight were defined as < 2500 g and < 1500 g respectively. Each birth date was lagged by one year to allow for the relevant maternal exposure period in the year before birth.  The relevant denominators and years of analysis are shown in Table I.

Statistical methods. Risks for the population within 2 km of special waste landfill sites relative to the reference population (> 2 km) were calculated by indirect standardisation, using model predictions from Poisson regression of data from the reference area. These provided more stable estimates of reference rates than did stratification (because of the small numbers of cases—or even zero counts—in some cells for the rarer events). The regression function included the following variables: year of birth, sex (birth weight and stillbirth only) and deprivation, using a descending stepwise selection procedure, starting from the fullest model including all possible interactions.  This was repeated without deprivation and models were constrained to differ only in terms of deprivation. We used stratification to adjust for deprivation for hospital admissions because of the sparseness of data.

Thus, deprivation unadjusted models were year only for all anomalies, neural tube defects, cardiovascular defects, abdominal wall defects and very low birth weight; sex only for low birth weight; sex and year for stillbirth and no adjustment for hypospadias and epispadias. Deprivation adjusted models included deprivation as a single additional term in each case except low birth weight, where the deprivation-sex interaction was also included.

We calculated Poisson 99% confidence intervals around the relative risk estimates, assuming a common relative risk for all landfill sites. This provides consistency with the analyses in Elliott et al¹³. To the extent that the model assumptions fail to hold (for example, because of data anomalies, unmeasured confounding or sampling variability in the rates) some degree of over-dispersion and a widening of the confidence intervals might be expected.

Results

There were 61 special waste landfill sites included in the study (figure 1).  The area within 2 km of these landfill sites tended to be more urban and more deprived than that beyond 2 km (reference area)— 46% (versus 39%) of the total births were in the most deprived tertile of Carstairs score. Compared with the reference area, the area within 2 km had a higher proportion of births to mothers under 20 years of age (8.6% versus 7.4%) and, among women aged 15-44, included (1991 census) a higher proportion of women of Indian, Pakistani or Bangladeshi origin (14.3% versus 5.5%) and a slightly higher proportion of black women (5.0% versus 4.3%).

For congenital anomalies, a total of 149,171 live births, stillbirths and terminations (19,468 within 2 km and 129,703 in the reference area) were included, while for stillbirths, there was a total of 326,016 live and stillbirths (41,991 within 2 km and 284,025 in the reference area) (Table I).

Table II shows the numbers of cases, rates and relative risk estimates comparing the area near special waste landfill sites with the reference area for each of the outcomes. The relative risk of all congenital anomalies combined was 0.96 (0.89-1.02) after adjusting for deprivation. No statistically significant excess risks were found for any of the specific anomalies. Results from subsidiary analyses that included only sites operating at the time (ie without the sites that were closed) were substantively unchanged (not shown).  Sites that opened during the study period were too few for reasonable analysis.

Discussion

We analysed a subset of the data reported in Elliott et al¹³ focussing on special waste sites in Scotland. We found no statistically significant excess risks of any congenital anomaly or of low and very low birth weight in populations living within 2km of a Scottish special waste site.

In the previous study¹³ the relative risk of all congenital anomalies near special waste sites in Great Britain was 1.07 (99% CI: 1.04-1.09), whereas the current sub-analysis for Scottish special waste sites showed no excess risk: 0.96 (0.89-1.02). Possible explanations for this difference include chance, the socio-demography of areas surrounding landfills or some aspect of the type or management of waste in Scottish special waste sites compared with the rest.  However, as noted in Elliott et al¹³, information on characteristics potentially affecting emissions (including type, volume and management of waste, geology and topography) was not available and classification into special and non-special waste sites does not necessarily reflect their hazard potential. Special waste sites may be subject to stricter management measures and design standards than non-special waste sites, and may handle only small quantities of special waste, while some hazardous waste may be disposed of in other sites.

The previous analysis of all special waste sites showed statistically significant excess risks of cardiovascular defects (1.11 (1.03-1.21)) and hypospadias and epispadias (1.11 (1.03-1.21)), whereas for analysis of the subset of Scottish special waste sites, the corresponding relative risks were 1.03 (0.85-1.26) and 0.84 (0.58-1.22) respectively. However, for the specific congenital anomalies, the numbers of cases in the sub-analyses were comparatively small, yielding a limited statistical power to detect any excess risks.

A number of limitations of our analysis need to be noted. We adjusted for calendar year to deal with fluctuating rates through time, but it is known that registration of congenital anomalies is incomplete[xv] and ascertainment artifacts could have biased our results if they were differential with respect to distance from landfill. On the other hand, we included data on terminations for “grounds E” to improve ascertainment especially for neural tube defects, and data on hospital admissions and surgical corrections to give an independent source of data for those specific anomalies. Post-coded databases on births, stillbirths, congenital anomalies and terminations, and hospital data on admissions and surgical procedures for Scotland were used to locate cases and estimate distance from special waste sites. Misclassification of potential exposure to landfill may have occurred because of inaccuracies in the postcode directories, in the precise location of the landfill sites (which were represented by a point coordinate) or if the mother moved home during the relevant period after conception[xvi].

Potential confounding was addressed in two ways. Firstly, we carried out the analysis without and with adjustment for the Carstairs’ index at small-area level. This may incompletely account for individual-level characteristics associated with risk of congenital anomaly, such as smoking[xvii], drug use[xviii], and infections during pregnancy [xix] (which may themselves be distributed differentially with respect to landfill sites). Maternal age is not available from the Scottish congenital anomaly database, however, Elliott et al ¹³ and the EUROHAZCON ¹¹ study reported that maternal age (for risk of abdominal wall defects) did not appear to be a strong confounder in their studies. It was not possible to adjust for maternal ethnicity in the analysis because the data are not available at the postcode level.

In summary, no statistically significant excess risks of congenital anomalies were found in populations living near Scottish landfill sites licensed to carry special waste. Further work to characterise the spatial distributon of cngenital anomalies with respect to landfill sites, and to explore pathways of potential exposure to chemicals emanating from landfill sites (http://www.doh.gov.uk/landup.htm) is underway.

References

Back to November Contents