Environ. Sci. Technol. 2010, 44, 483–490
Pyrethroid and Organophosphorus Pesticides in Composite Diet Samples from Atlanta, USA Adults A N N E M . R I E D E R E R , * ,† RONALD E. HUNTER JR.,‡ S T E V E N W . H A Y D E N , † A N D P . B A R R Y R Y A N †,‡ Department of Environmental and Occupational Health, Rollins School of Public Health, Emory University, 1518 Clifton Road NE, Atlanta, Georgia 30322, and Department of Chemistry, Emory University, Atlanta, Georgia 30322
Received August 13, 2009. Revised manuscript received November 6, 2009. Accepted November 19, 2009.
Four pyrethroid (permethrin, cyfluthrin, cypermethrin, deltamethrin) and 3 organophosphorus (chlorpyrifos, diazinon, malathion) pesticides were measured in 4 days of 24 h duplicate diet samples collected from 12 Atlanta adults over two cycles (2005-2006). Samples were composited into 9 categories, by food type, to evaluate their contribution to daily intakes. The resulting 437 samples were analyzed using a multiresidue method using liquid-liquid and solid-phase extraction followed by quantification via gas chromatograph with electroncapture detection. Total daily intakes (mg/kg-d) were calculated by summing the mass of a pesticide in all composites collected that day and dividing by body weight. Chlorpyrifos, diazinon, and cypermethrin in were detected in a range of composite types at frequencies g30%, whereas other pesticides were detected at lower frequencies. Concentrations ranged from the detection limits (0.38-0.88 ng/g) to several hundred ng/ g, exceeding U.S. tolerances in a few cases. We also detected pesticides in some foods labeled organic. Total daily intakes were below the U.S. Environmental Protection Agency’s oral reference doses, except in 6% of cases when the organophosphorus concentrations were summed. Results show frequent dietary exposure of our participants to the target pesticides from a range of food types.
Introduction Dietary intake is an important potential source of nonoccupational pesticide exposure among U.S. adults, particularly for pyrethroid and organophosphorus (OP) pesticides (1-6). These are registered for a range of agricultural and livestock applications in the United States and elsewhere. Data from the U.S. Department of Agriculture’s 2005-2007 Pesticide Data Program (PDP) showed detectable pyrethroid and OP residues in 35 and 44 respectively, commodities including fruits, vegetables, nuts, dairy, grains, and meats (7-9). Commodities with the highest pyrethroid or OP detection frequencies were wheat, almonds, honey, spinach, celery, and cherries. Malathion, for example, was detected in 65% * Corresponding author phone: (404) 712-8458; fax (404) 727-8744; e-mail:
[email protected]. † Department of Environmental and Occupational Health (institution where work was performed). ‡ Department of Chemistry. 10.1021/es902479h
2010 American Chemical Society
Published on Web 12/08/2009
of wheat samples while the permethrin was detected in 56% of spinach samples. Certain commodities (blueberries, cherries, grapes, green beans, collard greens, kale, lettuce, peaches) contained >10 different pyrethroids/OPs during these PDP years. Whereas the PDP reports residues in raw commodities, the U.S. Food and Drug Administration’s total diet study (TDS) examines levels in ready-to-be-eaten foods (10). Among the most commonly detected pesticides in the 2004-2006 TDS were the malathion (18% of samples), chlorpyrifosmethyl (16%), chlorpyrifos (8%), and permethrin (6%) (11-13). In our own study of OPs in 379 duplicate 4-day solid food samples from 75 Maryland adults in 1995-1996, we detected malathion and chlorpyrifos in 75% and 38%, respectively (14). A 1990s study of Arizona residents found chlorpyrifos and diazinon in 22% and 8% respectively of 24 h duplicate solid diet samples (15). The objective of the present study was to measure pyrethroid and OP pesticides in 24 h duplicate diet samples collected from 12 adult volunteers in Atlanta, Georgia, USA. Volunteers were recruited from the Emory University community. We collected 8 days of samples from each participant, in two cycles of 4 consecutive days each to evaluate potential seasonal differences (Cycle 1: July-October 2005, Cycle 2: January-April 2006). We focused on 4 pyrethroids (permethrin, cyfluthrin, cypermethrin, deltamethrin) and 3 OPs (chlorpyrifos, diazinon, malathion) to represent those commonly used in U.S. agriculture. These OPs were among the top 10 most commonly used across all market sectors in 1999 and 2001 (16). Although there is a lack of analogous data on the pyrethroids, the 4 we included are those we observed during 2005-2008 to be the most commonly sold for residential use in Atlanta stores.
Experimental Section Sample Collection and Processing. Samples were handled with only glass or metal equipment that was trace-cleaned prior to use. Trace-cleaning steps are detailed in the Supporting Information. We trained participants and provided written instructions on collecting duplicate diet samples. We asked them to separate food items into nine composite types: above-ground vegetables, beans/nuts/ legumes/miscellaneous, below- ground vegetables, dairy, fats/oils, fruit/fruit juices, grains, meat/fish/eggs, and nondairy beverages. We classified vegetables as above- (e.g., lettuce) or below-ground (e.g., carrots) depending on where the bulk of the edible portion grows. We defined fruit juices as those containing g10% juice; participants were asked to put fruit juices/drinks with 30% of grains samples, cypermethrin in >30% of beans/nuts/legumes/ miscellaneous and below-ground vegetables samples, and cyfluthrin in >30% of below-ground vegetables and dairy samples. With respect to total diet, chlorpyrifos and cyfluthrin were the most frequently detected analytes in Cycles 1 and 2, respectively. These results provide evidence of frequent dietary exposure of the participants to one or more of the target pesticides. Pesticide Concentrations by Composite Type. Figure 1 presents box plots of pesticide concentrations (ng/g) by composite type and sampling cycle for samples >LOD. Median permethrin concentrations >LOD ranged from 21 ng/g (n ) 3, Cycle 2) in fruit/fruit juices to 460 ng/g in a single Cycle 2 beans/nuts/legumes/miscellaneous sample comprised of a veggie burrito, peanut butter, and raspberry jelly. Permethrin is currently registered in the United States for use on corn, livestock, and variety of fruits and vegetables;
there are no current U.S. tolerances for beans, peanut butter, or raspberries (21). Median cyfluthrin concentrations >LOD ranged from 9.1 ng/g in fruit/fruit juices (n ) 10, Cycle 2) to 264.9 ng/g in meat/fish/eggs (n ) 3, Cycle 1). The highest level (397.0 ng/ g) was found in a Cycle 1 beans/nuts/legumes/miscellaneous sample of hummus. Current U.S. registrations for cyfluthrin include a wide range of items; the current U.S. tolerance for dried/shelled peas and beans is 150 ng/g (22). Median cypermethrin concentrations >LOD ranged from 13 ng/g in fruit/fruit juices (n ) 7, Cycle 2) to 170 ng/g in grains (n ) 4, Cycle 2). The highest level (468 ng/g) was in a Cycle 2 beans/nuts/legumes/miscellaneous sample comprised of peanut butter bars, chutney, and orzo. Registered uses for cypermethrin are similar to those for permethrin and cyfluthrin (23). Median deltamethrin concentrations >LOD ranged from 6.9 ng/g in a single Cycle 1 dairy sample to 130.7 ng/g in grains (n ) 2, Cycle 2). The highest level (388.7 ng/g) was found in a Cycle 2 grains sample consisting of bread, bagel, chocolate muffin, and pineapple cake. Deltamethrin is registered for similar uses as the other pyrethroids, although for a more restricted set of fruits and vegetables (24). The current U.S. tolerances for cereal grains and wheat bran are 1000 ng/g and 5000 ng/g, respectively (24). Median chlorpyrifos concentrations >LOD ranged from 2.4 ng/g for meat/fish/eggs (n ) 4, Cycle 2) to 193.7 ng/g for beans/nuts/legumes/miscellaneous (n ) 6, Cycle 1). The highest level (435.8 ng/g) was found in the Cycle 2 grains sample described above with the highest deltamethrin concentration. Chlorpyrifos is registered for use on a range of commodities; the current U.S. tolerance for wheat grain is 500 ng/g (25). Chlorpyrifos was detected in LOD ranged from 21.2 ng/g in a single Cycle 1 sample of below ground vegetables to 248.5 ng/g in grains samples (n ) 6, Cycle 1). The highest level (380.8 ng/g) was found in a grains sample VOL. 44, NO. 1, 2010 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
485
FIGURE 1. Box plots of target analyte concentrations (ng/g, samples > LOD only) by composite type; AGV ) above ground vegetables, BGV ) below ground vegetables, BNL ) beans/nuts/legumes/ miscellaneous, DAI ) dairy, FRU ) fruit/fruit juice, GRA ) grains, MEE ) meat/fish/eggs; lefthand box ) Cycle 1 data; righthand box ) Cycle 2 data; lower box boundary ) 25th percentile, line in box ) median, upper box boundary ) 75th percentile, whiskers ) 10th and 90th percentiles, • ) observation 90th percentile; stand-alone horizontal line indicates only one sample > LOD for that type/cycle; missing box/line indicates no samples > LOD for that type/cycle. of whole wheat sourdough bread. Diazinon is registered for use on a range of crops, but not wheat (26). Diazinon was detected in only 1 of 1361 of wheat samples analyzed in the 2005 and 2006 PDP, at 5 ng/g (8, 9). Median malathion concentrations >LOD ranged from 11.6 ng/g in beans/nuts/legumes/miscellaneous (n ) 3, Cycle 2) to 348.8 ng/g in grains (n ) 2, Cycle 1). The highest level (377.9 ng/g) was found in a Cycle 1 beans/nuts/legumes/ miscellaneous sample consisting of beans, a sandwich, energy bars, and peanut butter. Malathion is currently registered 486
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 44, NO. 1, 2010
for use on a range of crops; the current U.S. tolerance for beans, soybeans, and peanuts is 8 ng/g (27). Detection Differences Between Sampling Cycles. The number of samples analyzed in each composite group was similar in Cycles 1 and 2. Table 1 illustrates differences in detection frequencies by cycle for certain composite types, whereas Figure 1 shows no clear pattern of difference in median concentrations by cycle. The between-cycle detection frequency differences are not likely due to field or laboratory contamination because no analytes were detected in any
TABLE 2. Total Daily Intakes (mg/kg Body Weight-d) Versus Oral Reference Dose (RfD) by Pesticide ( RfD
median
max
4.0 × 10-6 5.8 × 10-5 6.0 × 10-5 6.2 × 10-6 1.7 × 10-5 7.5 × 10-5
1.5 × 10-3 1.4 × 10-3 1.7 × 10-3 2.7 × 10-3 3.3 × 10-3 3.0 × 10-3
5.0 × 10-6 3.2 × 10-6
3.8 × 10-3 2.1 × 10-4
0 0 0 1 10 1 0 6
a Source of oral reference doses, unless otherwise noted: www.epa.gov/iris [accessed 2 Feb 2009]. b Agency for Toxic Substances and Disease Registry Maximum Residue Level (MRL) for chronic duration (g365 days) oral exposure (Available: www.atsdr.cdc.gov/toxprofiles/tp86.html, accessed 2 Feb 2009]. c MRL for intermediate duration (15-364 days) oral exposure. d Chlorpyrifos, diazinon and malathion concentrations each converted to methamidophos equivalents following U.S. EPA’s Organophosphorus Cumulative Risk Assessment guidelines (www.epa.gov/pesticides/cumulative/2006-op/ op_cra_main.pdf) and summed; RfD shown is for methamidophos.
blanks from either cycle. Further, after we extracted samples from each cycle, we analyzed them together, thus our detection limits did not change. The differences are not likely due to changes in regulatory status because most registered uses did not change within the study time frame. The differences between cycles may be due to changes in the types of foods in the composite samples or seasonal changes in residue levels. These findings underscore the need for seasonal sampling of pesticides in duplicate diet samples to obtain a more complete picture of intake. Comparison to Other Duplicate Diet Studies. For most target analytes, our results were similar to those of other U.S. duplicate diet studies published to date, although the small sample sizes of all of our studies limit comparison. MacIntosh et al. (28) detected chlorpyrifos in 38% of duplicate solid food samples (median 0.4 ng/g) collected from 75 Maryland adults in 1995-96. They detected malathion more frequently (e.g., 75% of samples) than we did, however their detection limit (0.05 ng/g) was lower than ours. Diazinon was not measured. Fenske et al. (29) measured chlorpyrifos in two fresh produce composites (12-350 ng/g) and malathion in four processed food samples (4-21 ng/g) collected from 7 Washington children aged 2-5 in 1998. Diazinon was not detected. Bradman et al. (30) measured chlorpyrifos and malathion in 10% and 30% respectively of duplicate diet samples collected from 10 California children aged 6-12 months in 2002 at concentrations ranging from 1-8 ng/g. This was the only published study we could locate to measure pyrethroids in duplicate diet samples; none were detected. In Cycle 2, we detected malathion less frequently than these studies. Further, we detected it few grains samples in either cycle. We did not expect this, particularly because malathion was detected in 66% and 63% of wheat samples in the 2005 and 2006 PDP respectively at concentrations ranging from 5-2500 ng/g (7, 8). Because we achieved suitable recoveries of malathion in spiked food samples in our method development work (18), we do not believe our low number of malathion detects is due to analytical error. Pesticides in Foods Labeled “Organic”. Of the 999 individual foods recorded in the food logs (excluding fats/ oils), 18% had an “organic” label. We detected one or more target analytes (except malathion) in half of the 47 composite samples comprised only of organic foods. This is consistent with other U.S. studies that show OP and pyrethroid residues in organic foods, although generally at lower frequencies and concentrations than in conventional foods (31). We measured 457 ng/g of cypermethrin in a sample comprised of raw, domestic (California), precut, “organic” carrots; this
is five times the U.S. tolerance for carrots (100 ng/g) (23). For comparison, the 2006 PDP reported no cypermethrin detects in 744 carrot samples at detection limits of 30-60 ng/g (9). The participant who submitted this sample reported never using pesticides in her rented apartment since moving there in 2004, so it is unlikely that the cypermethrin measured represents household contamination. Four other all-“organic” samples contained pesticide concentrations >100 ng/g. An “organic” 14-grain bread sample from the same participant described above contained 330 ng/g of diazinon. A sample of carrots, yellow onions, potato, and sweet potato, all “organic,” cooked with olive oil, pepper, and rosemary contained 180 ng/g of cypermethrin and 267 ng/g of permethrin; this participant reported using no pesticides in her apartment the month before samples were collected. A sample of “organic” sprouted grain tortilla from this participant contained 133 ng/g of diazinon. A sample of “organic” homemade chicken soup (chicken, carrots, potatoes) collected from another participant contained 109 ng/g cyfluthrin; this participant reported using no pesticides in her apartment the month before samples were collected, although she reported using a cyfluthrin crackand-crevice treatment during the previous cycle. We detected 40 ng/g of deltamethrin (and 38 ng/g of cyfluthrin) in an “organic” lettuce sample of U.S. origin. Deltamethrin is not currently registered for use on U.S. lettuce (24) and this participant reported using no pesticides in her rented townhome the month before sampling. Total Daily Pesticide Intakes. Table 2 presents descriptive statistics of participants’ total daily intakes by pesticide and compares them to the oral RfDs. There was no significant difference in mean daily intakes by sampling cycle for most analytes except cyfluthrin (p < 0.05), thus Table 2 combines data for both cycles. Permethrin, cypermethrin, cyfluthrin, and malathion intakes were all below the RfDs. One intake exceeded the chlorpyrifos oral RfD. On that day, the participant’s chlorpyrifos intake came from a beans/nuts/ legumes/miscellaneous sample comprised of cereal, soy milk, cookies, peanut butter, and a granola bar. Another participant’s intake exceeded the intermediate duration MRL for diazinon; this intake came from a beans/nuts/legumes/ miscellaneous sample (214 ng/g diazinon) comprised of an energy bar, and a fruit/fruit juices sample (362 ng/g diazinon) comprised of grapes and a smoothie. Six percent of methamidophos-equivalent OP intakes exceeded the methamidophos RfD. This illustrates how combining OP residues using the U.S. EPA’s cumulative approach can produce an exceedance of the health guidelines VOL. 44, NO. 1, 2010 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
487
FIGURE 2. Boxplots of the percent contribution of each composite type to total daily intake (n ) 84) of the target pesticides (
