Journal of Toxicology and Environmental Health
ISSN: 0098-4108 (Print) (Online) Journal homepage: http://www.tandfonline.com/loi/uteh19
Factors affecting levels of DDT and metabolites in human breast milk from kwazulu H. Bouwman , A. J. Reinecke , R. M. Cooppan & P. J. Becker To cite this article: H. Bouwman , A. J. Reinecke , R. M. Cooppan & P. J. Becker (1990) Factors affecting levels of DDT and metabolites in human breast milk from kwazulu, Journal of Toxicology and Environmental Health, 31:2, 93-115, DOI: 10.1080/15287399009531440 To link to this article: http://dx.doi.org/10.1080/15287399009531440
Published online: 20 Oct 2009.
Submit your article to this journal
Article views: 5
View related articles
Citing articles: 25 View citing articles
Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=uteh19 Download by: [NUS National University of Singapore]
Date: 07 November 2015, At: 21:23
FACTORS AFFECTING LEVELS OF DDT AND METABOLITES IN HUMAN BREAST MILK FROM KWAZULU H. Bouwman
Journal of Toxicology and Environmental Health 1990.31:93-115.
Research Institute for Environmental Diseases of the Medical Research Council, Pretoria, Republic of South Africa A. J. Reinecke Department of Zoology, Potchefstroom University for CHE, Potchefstroom, Republic of South Africa R. M. Cooppan Research Institute for Diseases in a Tropical Environment of the Medical Research Council, Durban, Republic of South Africa P. J. Becker Institute for Biostatistics of the Medical Research Council, Pretoria, Republic of South Africa
Concentrations of p,p'-DDT, p,p'-DDE, and p,p'-DDD have been determined in breast milk of mothers residing in two different areas of KwaZulu. Annual intradomiciliary application of DDT was used for the interruption of malaria transmission in one area, while the other served as the control. Milk from mothers living in DDT-treated dwellings had significantly higher mean levels of DDT and metabolites (mean ΣDDT 15.83 mg kg-1 in milk fat) than those from the control area (0.69 mg kg-1). The highest recorded ΣDDT value was 59.3 mg kg-1 (milk fat). Primiparous mothers from the malarious area had significantly more ΣDDT and metabolites (ΣDDT 24.82 mg kg-1) than multiparous mothers from the same area (mean 12.21 mg kg-1). Parity was the best predictor of DDT in breast milk of the exposed group. The percentage DDT and the ΣDDT increased significantly with an increase in parity. The same, but not significant, trend was also found for the control group. It was hypothesized that the increase in percentage DDT that occurred with higher parities was due to the uptake of DDT and elimination via milk. This process was faster than the uptake and endogenous formation of DDE. This work was supported by the Medical Research Council and published with their permission. Prof. C. H. J. Schutte of the MRC is thanked for his assistance and encouragement. The KwaZulu Department of Health (in particular Mr. S. Ngxongo, Dr. G. M. Short, and Dr. V. Friedlund) and the South African Department of Health and Population Development (Mr. M. J. Botha) are thanked for their invaluable assistance. Mrs. J. Nkomokazi and Mr. G. Ngcobo are thanked for their excellent technical assistance. Request for reprints should be sent to H. Bouwman, Department of Zoology, Potchefstroom University for CHE, Potchefstroom, 2520, Republic of South Africa.
93 Journal of Toxicology and Environmental Health, 31:93-115, 1990 Copyright © 1990 by Hemisphere Publishing Corporation
94
H. BOUWMAN ET AL.
Designing predictive models using multiple regression was not very successful. The recorded levels do not represent an appreciable health risk to the mothers. From the literature it was deduced that at the recorded levels, a well-founded risk to the infants, particularly the firstborns, exists in sprayed areas.
Journal of Toxicology and Environmental Health 1990.31:93-115.
INTRODUCTION The concentration of DDT in human breast milk has been the subject of many studies (Wickstrom et al., 1983; Stacey et al v 1985; Rogan et al., 1986a; Mussalo-Rauhamaa et al., 1988). However, none of these have established the factors influencing these levels of milk from mothers living in an area where malaria control is the only possible source of DDT. The concentration of DDT in breast milk must be established to determine its risk to the mother and infant. Hyporeflexia in infants has been associated with DDE at levels as low as 4 mg I"1 (milk fat) in breast milk (Rogan et al., 1986b). The need for a special approach to evaluate the health risk of chemicals during infancy has been identified by the WHO (1986). The northern parts of KwaZulu in Natal are malaria endemic. The dwellings of the more than 300,000 residents are treated annually with DDT at a rate of 2 g m~2 on the inner wall surface (Sharp et al., 1988). DDT is taken up by the population via food, air, and skin contact, but the relative contribution of each route is not known. Fish from the Pongolo River, which forms a major part of dietary protein, contributes, at its worst, only 0.312 /xg kg" 1 d" 1 of total DDT (EDDT) (Bouwman et al., 1990). This is less than the ADI (acceptable daily intake) of 5 ng kg"1d~1 (Klaassen et al., 1986). The concentration of EDDT in serum of 71 people (all members of 12 families) from this area was 140.9 /xg I"1 before application. Ten days after treatment the level increased to 174.6 jug I"1 (H. Bouwman, unpublished observations, 1988). The aim of this study was to determine the influence of factors such as parity, maternal age, and the age of the infant on levels of DDT, DDE, and EDDT (total DDT) in breast milk of exposed and nonexposed groups of lactating mothers. Particular attention was also given to the change in percentage DDT. The possible health implications to the infant will also be discussed. MATERIALS A N D METHODS Subjects Milk samples were collected at two hospitals during 1986/1987. The mothers attending clinics at Mseleni Hospital (test group) live in an area where DDT is used for malaria control (test area). Mseleni is situated on the shore of Lake Sibaya and is removed from major routes used by migrant workers from Mozambique. In a pilot study it was found that migrants from Mozambique, where hardly any malaria control is prac-
Journal of Toxicology and Environmental Health 1990.31:93-115.
DETERMINANTS OF DDT LEVELS IN BREAST MILK
95
ticed, had very low levels of DDT and this influenced results (H. Bouwman, unpublished observations, 1987). Dwellings are usually constructed of mud, branches, and thatch. Homesteads consist of 3 to 7 such structures and house from 4 to 22 people. The major source of income is from migrant labor in mines or on farms. The diet consists of staples such as maize and rice with meat from fish, goats, chickens, or cattle. This is supplemented by fruit, nuts, and roots. The control group was drawn from the Port Shepstone area in southern Natal where malaria has never occurred. No DDT has been used there since 1976 (van Dyk et al., 1982) for any purpose. The population from this area is also stable, but more westernized than the Mseleni group. Dwellings were also constructed from mud, but more cement structures were in evidence. Homesteads consist of two to four dwellings. Income was generated from labor in transport, plumbing, carpentry, and the tourist industry. The diet was essentially the same as for the Mseleni population, a large proportion of which was obtained from retailers. There were no vegetarians or smokers in this or the exposed group. Sample Collection Samples were taken over a period of 12 mo (November 1986 to November 1987). Informed consent was obtained from all the mothers on routine visits to the clinics and from mothers in the delivery wards. A questionnaire was completed for each participant. Information obtained included maternal and infant age, parity, previous residence, and occupational exposure to DDT or other pesticides. This was done in the subject's own language, Zulu. Maternal weight was not determined. The mothers were provided with clean 100-ml glass beakers and asked to manually express about 10 ml breast milk. The milk was immediately transferred to 10-ml blood collecting tubes and kept on ice until frozen on the same day. Milk from mothers unable to produce more than 4 ml was pooled and used to establish recovery. Sample Preparation Samples were extracted using a method previously described (Bouwman et al., 1989). Briefly, this involved the denaturing of the milk protein using mercaptoethanol and solubilizing the fats using deoxycholic acid. Liquid-liquid extraction with hexane was then done and the fat content determined after evaporation of the solvent. The extract was then cleaned up using silica gel. The purified extract was analyzed using gas chromatography (electron-capture detection) with aldrin as an internal standard. The lower limit of detection was 0.5 pg per component. This method gave recoveries ranging from 107% for DDT at high levels to 74%
96
H. BOUWMAN ET AL.
for DDD at low levels. EDDT recovery was 98.6%. This method was developed to facilitate recovery of a fraction of the spiked compound that was not extracted after the spiked sample was frozen for longer than 24 h. Confirmation of peak identities was done by an independent laboratory. Concentration will mostly be referred to on a milk fat basis to facilitate comparison with literature. Results presented here were not corrected for recovery. Values less than the detection limits were treated as missing values and excluded from statistical analysis.
Journal of Toxicology and Environmental Health 1990.31:93-115.
RESULTS None of the mothers approached refused to participate, except on grounds of difficult or no lactation. Only five mothers, three from the control and two from the exposed group, fell in this category. Comparative information on characteristics of the two groups is presented in Table 1. There were no significant differences (at the 5% level of significance) in maternal age, parity, infant age, or percentage milk fat between groups. Maternal age was difficult to establish as in many cases this was not known. A calculated guess was made. The same was true for infant age, but here more accurate inference was possible. Mothers from the Murchison area practiced more bottle feeding than those mothers attending the Mseleni clinics. The percentage of mothers that practiced bottle feeding exclusively was included in the table for sake of completeness, but was not included in the total number of participants nor in subsequent statistics. DDE and DDT was detected in all samples of the exposed group. DDD was not detected in three samples of the exposed group. The control group was 100% positive for DDE, five samples was positive for DDD, and only 19 had detectable DDT residues (see Table 3). Chi-square analysis of two-by-two tables for frequency of samples with detectable amounts of DDT, DDE, and DDD was performed. The frequencies for TABLE 1. Comparative Summary of Some Characteristics of Mothers and Children Attending Baby Clinicsa
Number of samples Maternal age (yr)6 Parity6 Infant age (mo)6 Percent breast feeding Percent bottle feeding Percent breast + bottle feeding Percent milk fat6
Exposed
Control
132
88 25.5 (7.1) 3.0 (2.1) 7.7 (5.9) 71 4.6 24.1 3.76 (2.3)
25.5 (5.6) 2.7 (1.8) 8.6 (5.4) 89 1.5 8.3
3.95 (2.1)
^Standard deviations are shown in parentheses. 6 Does not differ (p > .05) between groups.
97
Journal of Toxicology and Environmental Health 1990.31:93-115.
DETERMINANTS OF DDT LEVELS IN BREAST MILK
detectable amounts of both DDT and DDD were significantly different between the two groups (p < .001). Standard deviations of the EDDT concentrations in breast milk of the exposed group were relatively large, sometimes as large as the mean value, indicating a wide distribution (Table 2). This was also reflected by the indicated minimum and maximum values. Data were skewed as indicated by the differences between the mean and median values. The DDT levels of the control group (Table 3) were much lower than for the exposed group. The standard deviations were of the same order as that of the respective mean values. Data were slightly skewed, as indicated by the difference between the mean and median. Differences were highly significant for all parameters relating to DDT between the two groups (p < .001). The natural logarithms of DDE and DDT concentrations were well correlated in the exposed group (Fig. 1) and poorly in the control group, which had a coefficient of determination of 13.2% and a regression of In DDT = 0.53 + 0.33 In DDE. TABLE 2. Summary Statistics for Pooled Data of the Exposed Group
Sample size Mean Median Standard deviation Minimum Maximum
Sample size Mean Median Standard deviation Minimum Maximum
Sample size Mean Median Standard deviation Minimum Maximum a
Whole milk: ftg
fc
1
Milkfat:mgkg- .
DDE*
DDD a
DDTa
129 315.5 209.9 298.6 18.5 1758.8
129 15.98 10.38 25.42 0 226.72
129 242.21 203.7 164.6 20.1 967.9
EDDTa
DDE6
DDD 6
129 574.89 427.5 454.16 67.6 2743.3
129 8.65 6.2 7.67 0.5 46.9
129 0.400 0.3 0.398 0 2.14
DDT6
EDDT6
%DDT
129 15.83 12.8 11.62 1.05 59.3
129 45.71 45.83 8.77 19.11 70
129 6.77 5.8 4.31 0.42 28.8 I"1.
98
H. BOUWMAN ET AL.
TABLE 3 . Summary Statistics for Data of the Control Croup
Journal of Toxicology and Environmental Health 1990.31:93-115.
Sample size Mean Median Standard deviation Minimum Maximum
Sample size Mean Median Standard deviation Minimum Maximum
Sample size Mean Median Standard deviation Standard error Minimum Maximum Range a
DDEa
DDD a
DDTa
88 20.74 18 16.03 0.1 72.7
88 0.03 0 0.14 0 0.93
88 1.30 0.8 2.07 0 13.1
EDDT3
DDE6
DDD 6
88 22.07 19.15 17.08 0.1 74.6
88 0.65 0.47 0.665 0 4.73
88 0 0 0 0 0.03
DDT6
EDDT 6
%DDT
88 0.04 0.02 0.05 0.0005 0 0.36 0.36
88 0.69 0.49 0.68 0.07 0 4.8 4.8
88 5.59 3.96 5.94 0.63 0 24 24
Whole milk:^g I"1. Milk fat: mg k g " 1 .
6
Three factors were examined to determine their influence on the levels of organochlorines in breast milk. These were parity, the age of the infant, and the age of the mother. Percentage milk fat was not correlated with the determined levels. The parity groups (Fig. 2) were compared with respect to EDDT in a one-way analysis of variance by multiple comparisons to find that the mean EDDT level (24.82 mg kg" ) of the primiparous mothers was significantly higher than mothers breast feeding their second or subsequent child. The coefficient of determination for the second-order polynomial regression (which gave the best fit) for the means of the parities was 95.6% [EDDT = 32.7 - 8.34(Parity) + 7.89(Parity)2], but was only 25.1% when all raw data were used. The slopes of the separate multiplicative regression analysis for DDE and DDT (graphs not shown) separately were different. For DDE it was -0.71 (r .534; p < .001) with the expression In DDE = 2.43 - 0.71
DETERMINANTS OF DDT LEVELS IN BREAST MILK
58
99
L. MULTIPLICATIVE
r
Corr Coerf. = 6.8846
46
f
s E 30
h
20
lL_
a
E
p < O.OuOl
CoeTf Deter.
= 78.26X
LnY=-e.64-1.09LnX
h
I
Journal of Toxicology and Environmental Health 1990.31:93-115.
D
h
16
IS DDT
26
(milk f a t : mg/1)
FIGURE 1. Relationship between levels of DDT and DDE in milk fat of breast milk of the exposed group. Relevant statistics are displayed for a multiplicative model (V - aXb).
L
2S [_-
| S4.B2 "f *
r-
t L
13.94'
ISSN: 0098-4108 (Print) (Online) Journal homepage: http://www.tandfonline.com/loi/uteh19
Factors affecting levels of DDT and metabolites in human breast milk from kwazulu H. Bouwman , A. J. Reinecke , R. M. Cooppan & P. J. Becker To cite this article: H. Bouwman , A. J. Reinecke , R. M. Cooppan & P. J. Becker (1990) Factors affecting levels of DDT and metabolites in human breast milk from kwazulu, Journal of Toxicology and Environmental Health, 31:2, 93-115, DOI: 10.1080/15287399009531440 To link to this article: http://dx.doi.org/10.1080/15287399009531440
Published online: 20 Oct 2009.
Submit your article to this journal
Article views: 5
View related articles
Citing articles: 25 View citing articles
Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=uteh19 Download by: [NUS National University of Singapore]
Date: 07 November 2015, At: 21:23
FACTORS AFFECTING LEVELS OF DDT AND METABOLITES IN HUMAN BREAST MILK FROM KWAZULU H. Bouwman
Journal of Toxicology and Environmental Health 1990.31:93-115.
Research Institute for Environmental Diseases of the Medical Research Council, Pretoria, Republic of South Africa A. J. Reinecke Department of Zoology, Potchefstroom University for CHE, Potchefstroom, Republic of South Africa R. M. Cooppan Research Institute for Diseases in a Tropical Environment of the Medical Research Council, Durban, Republic of South Africa P. J. Becker Institute for Biostatistics of the Medical Research Council, Pretoria, Republic of South Africa
Concentrations of p,p'-DDT, p,p'-DDE, and p,p'-DDD have been determined in breast milk of mothers residing in two different areas of KwaZulu. Annual intradomiciliary application of DDT was used for the interruption of malaria transmission in one area, while the other served as the control. Milk from mothers living in DDT-treated dwellings had significantly higher mean levels of DDT and metabolites (mean ΣDDT 15.83 mg kg-1 in milk fat) than those from the control area (0.69 mg kg-1). The highest recorded ΣDDT value was 59.3 mg kg-1 (milk fat). Primiparous mothers from the malarious area had significantly more ΣDDT and metabolites (ΣDDT 24.82 mg kg-1) than multiparous mothers from the same area (mean 12.21 mg kg-1). Parity was the best predictor of DDT in breast milk of the exposed group. The percentage DDT and the ΣDDT increased significantly with an increase in parity. The same, but not significant, trend was also found for the control group. It was hypothesized that the increase in percentage DDT that occurred with higher parities was due to the uptake of DDT and elimination via milk. This process was faster than the uptake and endogenous formation of DDE. This work was supported by the Medical Research Council and published with their permission. Prof. C. H. J. Schutte of the MRC is thanked for his assistance and encouragement. The KwaZulu Department of Health (in particular Mr. S. Ngxongo, Dr. G. M. Short, and Dr. V. Friedlund) and the South African Department of Health and Population Development (Mr. M. J. Botha) are thanked for their invaluable assistance. Mrs. J. Nkomokazi and Mr. G. Ngcobo are thanked for their excellent technical assistance. Request for reprints should be sent to H. Bouwman, Department of Zoology, Potchefstroom University for CHE, Potchefstroom, 2520, Republic of South Africa.
93 Journal of Toxicology and Environmental Health, 31:93-115, 1990 Copyright © 1990 by Hemisphere Publishing Corporation
94
H. BOUWMAN ET AL.
Designing predictive models using multiple regression was not very successful. The recorded levels do not represent an appreciable health risk to the mothers. From the literature it was deduced that at the recorded levels, a well-founded risk to the infants, particularly the firstborns, exists in sprayed areas.
Journal of Toxicology and Environmental Health 1990.31:93-115.
INTRODUCTION The concentration of DDT in human breast milk has been the subject of many studies (Wickstrom et al., 1983; Stacey et al v 1985; Rogan et al., 1986a; Mussalo-Rauhamaa et al., 1988). However, none of these have established the factors influencing these levels of milk from mothers living in an area where malaria control is the only possible source of DDT. The concentration of DDT in breast milk must be established to determine its risk to the mother and infant. Hyporeflexia in infants has been associated with DDE at levels as low as 4 mg I"1 (milk fat) in breast milk (Rogan et al., 1986b). The need for a special approach to evaluate the health risk of chemicals during infancy has been identified by the WHO (1986). The northern parts of KwaZulu in Natal are malaria endemic. The dwellings of the more than 300,000 residents are treated annually with DDT at a rate of 2 g m~2 on the inner wall surface (Sharp et al., 1988). DDT is taken up by the population via food, air, and skin contact, but the relative contribution of each route is not known. Fish from the Pongolo River, which forms a major part of dietary protein, contributes, at its worst, only 0.312 /xg kg" 1 d" 1 of total DDT (EDDT) (Bouwman et al., 1990). This is less than the ADI (acceptable daily intake) of 5 ng kg"1d~1 (Klaassen et al., 1986). The concentration of EDDT in serum of 71 people (all members of 12 families) from this area was 140.9 /xg I"1 before application. Ten days after treatment the level increased to 174.6 jug I"1 (H. Bouwman, unpublished observations, 1988). The aim of this study was to determine the influence of factors such as parity, maternal age, and the age of the infant on levels of DDT, DDE, and EDDT (total DDT) in breast milk of exposed and nonexposed groups of lactating mothers. Particular attention was also given to the change in percentage DDT. The possible health implications to the infant will also be discussed. MATERIALS A N D METHODS Subjects Milk samples were collected at two hospitals during 1986/1987. The mothers attending clinics at Mseleni Hospital (test group) live in an area where DDT is used for malaria control (test area). Mseleni is situated on the shore of Lake Sibaya and is removed from major routes used by migrant workers from Mozambique. In a pilot study it was found that migrants from Mozambique, where hardly any malaria control is prac-
Journal of Toxicology and Environmental Health 1990.31:93-115.
DETERMINANTS OF DDT LEVELS IN BREAST MILK
95
ticed, had very low levels of DDT and this influenced results (H. Bouwman, unpublished observations, 1987). Dwellings are usually constructed of mud, branches, and thatch. Homesteads consist of 3 to 7 such structures and house from 4 to 22 people. The major source of income is from migrant labor in mines or on farms. The diet consists of staples such as maize and rice with meat from fish, goats, chickens, or cattle. This is supplemented by fruit, nuts, and roots. The control group was drawn from the Port Shepstone area in southern Natal where malaria has never occurred. No DDT has been used there since 1976 (van Dyk et al., 1982) for any purpose. The population from this area is also stable, but more westernized than the Mseleni group. Dwellings were also constructed from mud, but more cement structures were in evidence. Homesteads consist of two to four dwellings. Income was generated from labor in transport, plumbing, carpentry, and the tourist industry. The diet was essentially the same as for the Mseleni population, a large proportion of which was obtained from retailers. There were no vegetarians or smokers in this or the exposed group. Sample Collection Samples were taken over a period of 12 mo (November 1986 to November 1987). Informed consent was obtained from all the mothers on routine visits to the clinics and from mothers in the delivery wards. A questionnaire was completed for each participant. Information obtained included maternal and infant age, parity, previous residence, and occupational exposure to DDT or other pesticides. This was done in the subject's own language, Zulu. Maternal weight was not determined. The mothers were provided with clean 100-ml glass beakers and asked to manually express about 10 ml breast milk. The milk was immediately transferred to 10-ml blood collecting tubes and kept on ice until frozen on the same day. Milk from mothers unable to produce more than 4 ml was pooled and used to establish recovery. Sample Preparation Samples were extracted using a method previously described (Bouwman et al., 1989). Briefly, this involved the denaturing of the milk protein using mercaptoethanol and solubilizing the fats using deoxycholic acid. Liquid-liquid extraction with hexane was then done and the fat content determined after evaporation of the solvent. The extract was then cleaned up using silica gel. The purified extract was analyzed using gas chromatography (electron-capture detection) with aldrin as an internal standard. The lower limit of detection was 0.5 pg per component. This method gave recoveries ranging from 107% for DDT at high levels to 74%
96
H. BOUWMAN ET AL.
for DDD at low levels. EDDT recovery was 98.6%. This method was developed to facilitate recovery of a fraction of the spiked compound that was not extracted after the spiked sample was frozen for longer than 24 h. Confirmation of peak identities was done by an independent laboratory. Concentration will mostly be referred to on a milk fat basis to facilitate comparison with literature. Results presented here were not corrected for recovery. Values less than the detection limits were treated as missing values and excluded from statistical analysis.
Journal of Toxicology and Environmental Health 1990.31:93-115.
RESULTS None of the mothers approached refused to participate, except on grounds of difficult or no lactation. Only five mothers, three from the control and two from the exposed group, fell in this category. Comparative information on characteristics of the two groups is presented in Table 1. There were no significant differences (at the 5% level of significance) in maternal age, parity, infant age, or percentage milk fat between groups. Maternal age was difficult to establish as in many cases this was not known. A calculated guess was made. The same was true for infant age, but here more accurate inference was possible. Mothers from the Murchison area practiced more bottle feeding than those mothers attending the Mseleni clinics. The percentage of mothers that practiced bottle feeding exclusively was included in the table for sake of completeness, but was not included in the total number of participants nor in subsequent statistics. DDE and DDT was detected in all samples of the exposed group. DDD was not detected in three samples of the exposed group. The control group was 100% positive for DDE, five samples was positive for DDD, and only 19 had detectable DDT residues (see Table 3). Chi-square analysis of two-by-two tables for frequency of samples with detectable amounts of DDT, DDE, and DDD was performed. The frequencies for TABLE 1. Comparative Summary of Some Characteristics of Mothers and Children Attending Baby Clinicsa
Number of samples Maternal age (yr)6 Parity6 Infant age (mo)6 Percent breast feeding Percent bottle feeding Percent breast + bottle feeding Percent milk fat6
Exposed
Control
132
88 25.5 (7.1) 3.0 (2.1) 7.7 (5.9) 71 4.6 24.1 3.76 (2.3)
25.5 (5.6) 2.7 (1.8) 8.6 (5.4) 89 1.5 8.3
3.95 (2.1)
^Standard deviations are shown in parentheses. 6 Does not differ (p > .05) between groups.
97
Journal of Toxicology and Environmental Health 1990.31:93-115.
DETERMINANTS OF DDT LEVELS IN BREAST MILK
detectable amounts of both DDT and DDD were significantly different between the two groups (p < .001). Standard deviations of the EDDT concentrations in breast milk of the exposed group were relatively large, sometimes as large as the mean value, indicating a wide distribution (Table 2). This was also reflected by the indicated minimum and maximum values. Data were skewed as indicated by the differences between the mean and median values. The DDT levels of the control group (Table 3) were much lower than for the exposed group. The standard deviations were of the same order as that of the respective mean values. Data were slightly skewed, as indicated by the difference between the mean and median. Differences were highly significant for all parameters relating to DDT between the two groups (p < .001). The natural logarithms of DDE and DDT concentrations were well correlated in the exposed group (Fig. 1) and poorly in the control group, which had a coefficient of determination of 13.2% and a regression of In DDT = 0.53 + 0.33 In DDE. TABLE 2. Summary Statistics for Pooled Data of the Exposed Group
Sample size Mean Median Standard deviation Minimum Maximum
Sample size Mean Median Standard deviation Minimum Maximum
Sample size Mean Median Standard deviation Minimum Maximum a
Whole milk: ftg
fc
1
Milkfat:mgkg- .
DDE*
DDD a
DDTa
129 315.5 209.9 298.6 18.5 1758.8
129 15.98 10.38 25.42 0 226.72
129 242.21 203.7 164.6 20.1 967.9
EDDTa
DDE6
DDD 6
129 574.89 427.5 454.16 67.6 2743.3
129 8.65 6.2 7.67 0.5 46.9
129 0.400 0.3 0.398 0 2.14
DDT6
EDDT6
%DDT
129 15.83 12.8 11.62 1.05 59.3
129 45.71 45.83 8.77 19.11 70
129 6.77 5.8 4.31 0.42 28.8 I"1.
98
H. BOUWMAN ET AL.
TABLE 3 . Summary Statistics for Data of the Control Croup
Journal of Toxicology and Environmental Health 1990.31:93-115.
Sample size Mean Median Standard deviation Minimum Maximum
Sample size Mean Median Standard deviation Minimum Maximum
Sample size Mean Median Standard deviation Standard error Minimum Maximum Range a
DDEa
DDD a
DDTa
88 20.74 18 16.03 0.1 72.7
88 0.03 0 0.14 0 0.93
88 1.30 0.8 2.07 0 13.1
EDDT3
DDE6
DDD 6
88 22.07 19.15 17.08 0.1 74.6
88 0.65 0.47 0.665 0 4.73
88 0 0 0 0 0.03
DDT6
EDDT 6
%DDT
88 0.04 0.02 0.05 0.0005 0 0.36 0.36
88 0.69 0.49 0.68 0.07 0 4.8 4.8
88 5.59 3.96 5.94 0.63 0 24 24
Whole milk:^g I"1. Milk fat: mg k g " 1 .
6
Three factors were examined to determine their influence on the levels of organochlorines in breast milk. These were parity, the age of the infant, and the age of the mother. Percentage milk fat was not correlated with the determined levels. The parity groups (Fig. 2) were compared with respect to EDDT in a one-way analysis of variance by multiple comparisons to find that the mean EDDT level (24.82 mg kg" ) of the primiparous mothers was significantly higher than mothers breast feeding their second or subsequent child. The coefficient of determination for the second-order polynomial regression (which gave the best fit) for the means of the parities was 95.6% [EDDT = 32.7 - 8.34(Parity) + 7.89(Parity)2], but was only 25.1% when all raw data were used. The slopes of the separate multiplicative regression analysis for DDE and DDT (graphs not shown) separately were different. For DDE it was -0.71 (r .534; p < .001) with the expression In DDE = 2.43 - 0.71
DETERMINANTS OF DDT LEVELS IN BREAST MILK
58
99
L. MULTIPLICATIVE
r
Corr Coerf. = 6.8846
46
f
s E 30
h
20
lL_
a
E
p < O.OuOl
CoeTf Deter.
= 78.26X
LnY=-e.64-1.09LnX
h
I
Journal of Toxicology and Environmental Health 1990.31:93-115.
D
h
16
IS DDT
26
(milk f a t : mg/1)
FIGURE 1. Relationship between levels of DDT and DDE in milk fat of breast milk of the exposed group. Relevant statistics are displayed for a multiplicative model (V - aXb).
L
2S [_-
| S4.B2 "f *
r-
t L
13.94'
