Dioxins in the Australian population: levels in human milk
Technical Report No. 10
Dr Fiona Harden, Dr Jochen Müller and Leisa Toms
Department of the Environment and Heritage, May 2004
ISBN 0 642 55002 6
- Dioxins in the Australian population: levels in human milk - Part A (PDF - 499 KB)
- Dioxins in the Australian population: levels in human milk - Part B (PDF - 820 KB)
The objective of this project was to investigate the levels of dioxin-like compounds in pooled human milk samples collected throughout Australia. These samples included polychlorinated dibenzo-p-dioxins and furans (PCDD/PCDFs) and polychlorinated biphenyls (PCBs). This study was carried out as part of the National Dioxins Program for the Australian Government Department of the Environment and Heritage. The study focused on donor cohorts with different potential exposure to dioxins and dioxin-like compounds in Australia (i.e. urban/industrial/rural exposure). The study was carried out in the following stages:
- obtaining appropriate ethical approval
- selection of the regions
- selection of cohorts suitable to provide information on the levels of dioxins and dioxin-like compounds in breast milk throughout Australia
- contact with local agencies to support the study
- identification of volunteers who fulfil the selection criteria of the individual cohorts
- collection of samples and completion of the questionnaire with individual donors
- pooling of samples
- analysis of dioxins and dioxin-like compounds in a certified laboratory
- data analysis and interpretation
- report writing
- data dissemination for public knowledge and peer review.
The results of this study provide a measure of the levels of polychlorinated PCDDs, PCDFs and PCBs in pooled human breast milk collected throughout Australia in 2002/03. Breast milk samples were collected from primiparae mothers recruited from a variety of sources. In order to allow direct comparison with previous World Health Organization (WHO) studies, volunteering mothers were selected using the following criteria:
- A primipara mother with a baby aged two to eight weeks (mothers of IVF babies were included)
- Exclusively breastfeeding
- Willing to provide a minimum of 100 ml (preferably 150 ml) of expressed milk. This volume was to be collected over a six week period (2-8 weeks post-partum)
- Healthy pregnancy, mother and child
- A resident of the area for the past five years.
In total, 173 samples were collected from 12 regions of Australia during the period March 2002 and September 2003. Of these, 16 were excluded because they were later found to have violated the inclusion/exclusion criteria. The remaining 157 samples were analysed as pooled samples and there were 17 pooled samples in total. These 157 samples covered the following sampling regions:
- Sydney (two pools)
- Melbourne (four pools)
- Adelaide (two pools)
- Rural inland NSW (Dubbo)
- Rural inland Queensland (Dalby)
- Rural Victoria (Bendigo, Ballarat, Lakes Entrance)
In addition to these samples, a further 24 .historical. samples collected in 1993 by the Key Centre for Applied and Nutritional Toxicology, were analysed as three pools of eight samples. In total, 20 pools of breast milk were analysed.
All pooled samples were sent to the Australian Government Analytical Laboratories, Sydney, and two duplicate samples (i.e. 50% volume of two of the pools) were sent to the State Laboratory of NRW, Münster, Germany. Both laboratories are accredited for dioxin analysis.
PCDD/PCDFs and PCBs were detected in all pooled samples. For samples collected during 2002/03, the mean and median levels, expressed as TEQ (middle bound), were 9.0 and 8.9 pg TEQ g-1 lipid, respectively. Lipid content was measured in all pooled samples and gave an average lipid concentration of 3.7±0.5%. No systematic differences were observed in the levels of dioxin-like chemicals in breast milk samples collected from different regions of Australia during 2002/03. A higher level of dioxin-like chemicals was detected in the Brisbane pool (15.2 pg TEQ g-1 lipid). This pool, however, showed a lower concentration of lipids in the breastmilk and since the concentrations of dioxin-like chemicals are normalised to the lipid content the elevation is likely to be related to an abnormally low lipid concentration in that sample and not an elevated level of contaminants. It is unclear why the lipid content was lower but it may have been due to preparation of the milk prior to analysis or analytical problems with the lipid determination in this sample. Samples obtained from Brisbane/Queensland for the WHO breast milk study as well as the results from the SE-Qld region for the blood study have not shown elevated levels of dioxin-like chemicals and, therefore, this result is not thought to reflect higher exposure of Brisbane women to these compounds.
For samples collected in 1993, the mean and median levels, expressed as TEQ (upper bound), were 16 and 16.4 pg TEQ g-1 lipid, respectively. Lipid content was measured in all pooled samples and the average concentration was found to be 3.9±0.7%.
A comparison of the samples collected from Melbourne women in 1993 with those collected for the present study showed clearly that the levels of these chemicals decreased over the ten year time period. It should, however, be noted that comparison of the two sample populations is complicated because details of maternal parity and infant age at date of collection was not made available for the older samples. Despite these limitations, a clear decrease in the levels of these compounds over time was observed. The concentration decreased by almost a factor of two from 1993 to 2003, from 16±1.4 to 9.1±1.3 pg g-1 lipid. Consistently, PCDD/PCDFs as well as PCBs decreased by about 60% during this period. This reflects the worldwide trend over recent decades of declining levels of dioxin-like compounds in the environment and humans. This was observed in the 3rd round WHO exposure studies, where on average, the decline between the 2nd round in 1993 and the 3rd round in 2003 was about 40% (Malisch and van Leeuwen, 2003). Similarly, a decline of 70% was observed in a study conducted in New Zealand (Bates et al, 2001).
In summary, the levels of PCDD/PCDFs and PCBs in the breast milk of Australian women are both similar across all regions of Australia and low by international standards. Consistent with worldwide trends, the levels of dioxin-like compounds have decreased over a ten-year period from 1993 to 2003 by approximately 60%. It should be noted that it is the advice of the WHO and the National Health and Medical Research Council (NHMRC) in Australia, that breast milk is the best food for babies. Breast milk may contain low levels of dioxins because of its fat content, but all babies are exposed to dioxins even if they are not breastfed. Alternative foods for babies, such as infant formula, may also contain dioxins because they may also contain fat. Several studies around the world in areas where dioxin levels are known to be high have still shown that breastfed babies are healthier than those fed infant formula.
(Department of the Environment and Heritage fact sheet http://www.deh.gov.au/industry/chemicals/dioxins/factsheet3.html)