| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Articles |
1
Institute of Legal Medicine, Humboldt-University, Hannoversche Strasse 6, D-10115 Berlin, Germany.
2
Department of Psychiatry and Psychotherapy of the Königin-Elisabeth-Hospital, Herzbergstrasse 79, D-10362 Berlin, Germany.
aAuthor for correspondence. Fax 49-30-450-525904; e-mail fritz.pragst{at}charite.de.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Methods: Hair samples from 19 alcoholics in a treatment program, 10 fatalities with verified excessive alcohol consumption, 13 moderate social drinkers who consumed up to 20 g ethanol/day, and 5 strict teetotalers were analyzed in 1–12 segments for four FAEEs (ethyl myristate, ethyl palmitate, ethyl oleate, and ethyl stearate) by external degreasing with n-heptane, extraction with a dimethyl sulfoxide-n-heptane mixture, headspace solid-phase microextraction of the extracts, and gas chromatography-mass spectrometry with deuterated internal standards. The n-heptane washings were analyzed in the same way for FAEEs from the hair surface.
Results: The sum of the four ester concentrations in hair calculated for the proximal 0–6 cm segment was 2.5–13.5 ng/mg (mean, 6.8 ng/mg) for the fatalities, 0.92–11.6 ng/mg (mean, 4.0 ng/mg) for 17 of the alcoholics in treatment, 0.20–0.85 ng/mg (mean, 0.41 ng/mg) for the moderate social drinkers, and 0.06–0.37 ng/mg (mean, 0.16 ng/mg) for the teetotalers. In almost all cases the segmental concentrations increased from proximal to distal. There was no agreement between the self-reported drinking histories of the participants and the FAEE concentrations along the hair length. Ethyl oleate was the dominant ester in all samples.
Conclusions: FAEEs are deposited in hair mainly from sebum. Despite large individual differences, FAEE hair concentrations can be used as markers for excessive alcohol consumption with relatively high accuracy.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-glutamyltransferase (-GT),1
erythrocyte mean cell volume (MCV), or carbohydrate-deficient transferrin, with the disadvantage that increased values can also originate from other pathologic reasons. Direct alcohol markers contain the carbon atoms of ethanol, and for this reason they can originate only from alcohol. Examples, in addition to ethanol itself, are ethyl glucuronide (1), phosphatidylethanol (2)(3), and fatty acid ethyl esters (FAEEs) (4)(5). Unfortunately, these compounds have a comparatively short life span in blood. Therefore, in blood they can be used only for detection of recent alcohol intake. In the past decade hair analyses have been shown to be a very useful tool for retrospective detection of illegal and medical drug abuse (6)(7)(8). In the same way as drugs, alcohol markers could also be deposited and stored in hair. The possibilities of hair analysis for detection of excessive alcohol consumption have been reviewed by Pragst et al. (9). The presence of ethyl glucuronide in hair was described by Skopp et al. (10) and Alt et al. (11). According to their results, the detection of ethyl glucuronide in hair is always associated with alcohol consumption, whereas a negative result does not unambiguously exclude alcohol abuse. The concentrations in positive cases of alcoholism were 0.11–4.0 ng/mg of hair (11).
In a previous report (12), we showed that FAEEs could be more reliable hair markers of chronic alcohol abuse. A very sensitive and specific method for the quantitative analysis of ethyl myristate, ethyl palmitate, ethyl oleate, and ethyl stearate from hair was developed and applied to hair samples from human fatalities with verified excessive alcohol abuse, moderate social drinkers, and teetotalers (12). We also showed that the total concentration of the four esters in hair from the alcoholics (1–29 ng/mg) was clearly higher than in hair from the social drinkers (<0.8 ng/mg). In hair from abstinent children and adult teetotalers, negative results or only traces of the esters were measured.
In principle, distribution of the FAEEs along the hair shaft should provide information about the drinking history of an individual. Therefore, hair samples from 19 individuals who were inpatients in a treatment program in the substance abuse division of a psychiatric department were analyzed in 1–12 segments, and the results were compared with their self-reported data about drinking behavior. For comparison, hair samples from 10 human fatalities with verified excessive previous alcohol consumption, 13 moderate social drinkers, and 5 teetotalers were investigated in the same way.
Because of their lipophilic character, the FAEEs could be excreted by the sebaceous glands and be distributed from there onto the hair surface. For this reason, FAEEs from the hair surface [external FAEEs (e-FAEEs)] and FAEEs deposited in the hair matrix [internal FAEEs (i-FAEEs)] were analyzed separately for all hair segments. From the results, information about the mechanism of the incorporation of FAEEs into the hair matrix could be obtained.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
. The study was approved by the local ethics commission of the University Hospital Charité. The members of groups A, C, and D gave written informed consent. The postmortem samples (group B) were investigated in agreement with the assignment of the public prosecutor.
|
In a questionnaire, each member of groups A and C provided a drinking history (approximate daily amount and kind of beverages and changes in drinking behavior over the last year, overall duration of excessive alcohol consumption, periods of abstinence) as well as the frequency and date of the last hair washing, dyeing, bleaching, use of hair lotion, or other kinds of hair care. In addition, body weight and size and selected clinical laboratory results [-GT, glutamate pyruvate transaminase (GPT), and MCV] were noted.
sample preparation
Depending on the hair length and the reported drinking history, the hair samples were cut into 1–12 segments. Only three samples were investigated in full length without division into segments. The length of the segments was 1–2.5 cm in the proximal region and up to 12 cm in the distal region. Each segment (20–50 mg) was weighed, cut to pieces 
1 mm in length, and washed twice with 1 mL of n-heptane for 1 min at 25 °C in a thermomixer (type "comfort"; Eppendorf) for separation of the external lipids from the hair surface. After centrifugation, both n-heptane washings were separated and combined in a 10-mL headspace vial. A solution containing 40 ng of each of the four d5-FAEEs (internal standards; concentration of each, 2 mg/L) in 20 µL of chloroform was added, the solvent was evaporated by a nitrogen stream at 40 °C, and the residue was analyzed by headspace solid-phase microextraction (HS-SPME) and gas chromatography-mass spectrometry (GC-MS) for the e-FAEE concentrations.
To each washed hair sample, 0.5 mL of dimethyl sulfoxide, 2 mL of n-heptane, and 20 µL of the d5-FAEE solution were added, and the mixture was shaken in the thermomixer for 15 h at 25 °C. After centrifugation, the n-heptane phase was separated and transferred into a headspace vial. The solvent was evaporated, and the i-FAEE concentrations in the residue were measured by HS-SPME and GC-MS.
hs-spme and gc-ms
The instruments, reagents, experimental conditions, and method validation for the HS-SPME and GC-MS measurements as well as the mass spectra of the four FAEEs analyzed (ethyl myristate, ethyl palmitate, ethyl oleate, and ethyl stearate) and the corresponding deuterated d5-FAEE internal standards have been described in detail in a previous report (12).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The large amounts of data that were produced by the determination of these eight FAEE concentrations for all segments of all samples cannot be fully included in this report. Rather, Tables 2–5
show the patterns of total FAEE concentrations in the investigated hair segments as a reduced-scale graphic for each individual. Furthermore, to have comparable data, we calculated the mean sums of the concentrations of the four esters in the proximal 6 cm of the sample (
i-FAEE and e-FAEE) from the segmental concentrations for all cases with hair lengths 
6 cm. For shorter hair, the sums of the concentrations in the total hair lengths are given. In the same way, the concentration ratios of the four esters (myristate:palmitate:oleate:stearate) in hair and in sebum were calculated for the proximal 6 cm of the samples or for the total hair lengths in the case of shorter samples.
|
|
|
|
group a: alcoholics in withdrawal treatment
In Table 2
the self-reported data concerning alcohol consumption for group A are compared with the results of the segmental hair analysis. With the exception of cases HV07 and HV08, the concentration sum i-FAEE was 0.92–11.6 ng/mg (mean, 4.0 ng/mg), which might be a typical range for alcoholics. There was no correlation between the reported daily alcohol consumption and the hair concentration. The low concentration in case HV08 could be explained by 2 months of abstinence before sampling and by the relatively low alcohol consumption (60–80 g ethanol/day) before that time. However, there was no similar explanation for case HV07 because that individual consumed large amounts of alcohol until 2 weeks before sampling and had strong withdrawal symptoms. It was unusual in this case that this individual had particularly bristly hair.
Fig. 1
shows the segmental concentrations in the 31-cm long hair sample from a 48-year-old male alcoholic (case HV01) who consumed 200–300 g ethanol/day (beer or wine) for a long time and stopped drinking 8 days before sampling. A strong increase in the concentrations from proximal to distal was found for the first 12 cm, although he did not change his drinking behavior during the last months. In the more distal segments, the concentrations slowly decreased. The dominant ester was ethyl oleate, followed by ethyl palmitate, with lower concentrations of ethyl myristate and ethyl stearate. For the e-FAEEs, a similar distribution was found, with the maximum concentrations more distal in the 21–24 cm segment and with an even higher portion of ethyl oleate. A similar distribution of segmental concentrations was found in all other cases with consistent drinking behavior over a long time (majority of the individuals in Table 2
).
|
Surprisingly, this segmental concentration distribution was also found in cases who reported a relapse of 2–5 weeks after a longer period of abstinence (cases HV05, HV06, HV10, and HV17). As an example, Fig. 2
shows the i-FAEE and e-FAEE concentrations in the segments of the 51-year-old male alcoholic HV05, who had a relapse after 14 months of abstinence and reported a daily consumption of 400 g of ethanol (brandy) for a period of 2 weeks. The sample was collected 3 days after he had again stopped drinking. Instead of finding high concentrations only in the proximal 1 cm segment, as expected, we found the esters over the full length of the hair, which indicates that, in general, a previous longer period of abstinence is not indicated by low i-FAEE concentrations in the corresponding segments if it is followed by an intense drinking period before collection of the hair sample. In general, agreement between the distribution of the FAEEs in the hair segments and the history of drinking behavior was not found. The e-FAEE concentration profile for case HV05 was similar to the profile for i-FAEEs, again with a higher portion of ethyl oleate, particularly in the proximal segments.
|
group b: fatalities with previous excessive alcohol consumption
The results of the segmental hair analysis for group B are given in Table 3
. The excessive alcohol consumption by these individuals was known from police reports and was confirmed by the corresponding autopsy findings. However, there were no details about the daily ethanol consumption or the drinking history. Only one individual (436/00) died from acute alcohol poisoning. In five cases, the individuals were sober at the time of death, and for two of these cases, death from withdrawal syndrome was stated because no other reason could be found. One case died from bleeding of the esophagus (063/01), and another case died of an injury that occurred while the individual was intoxicated (441/00). In the other cases, the cause of death was not related to, or only indirectly related to, the alcohol abuse.
The concentration sum i-FAEE (2.5–13.5 ng/mg; mean, 6.8 ng/mg) was higher in this group than in the withdrawal patients (Table 2
). Presumably these individuals had even higher alcohol consumption. Furthermore, a higher proportion of group B came from an antisocial environment and had neglected hair care. This may have favored the deposition from external lipids as can be seen from the very high sebum content (up to 44 ng/mg of hair). Characteristic was a high excess of ethyl oleate in the e-FAEE profile (up to 93%). In most cases, the segmental concentrations increased from proximal to distal. Only in cases 004/01 and 063/01 did the concentration distribution deviate from this pattern.
group c: moderate social drinkers
Hair samples were collected from 13 moderate social drinkers. Data concerning drinking behavior were collected by questionnaire and are given in Table 4
together with the results of the hair analysis. The reported drinking amounts were only roughly estimated by the members of this group. Daily records of the alcoholic beverages consumed by two volunteers (VA03 and VA04) for 2 months showed that a realistic estimation in cases of occasional drinking is rather difficult and may lead to underreported drinking amounts. The concentration sum i-FAEE in this group (0.20–0.85 ng/mg; mean, 0.41 ng/mg) was clearly below the range for the alcoholics. This is also obvious from Fig. 3
, in which the i-FAEE concentrations in the hair segments of three moderate social drinkers with a medium alcohol consumption of 6 or 12 g ethanol/day are compared in the same scale with those of two alcoholics in withdrawal treatment and two teetotalers. A relationship between the reported drinking amount and the i-FAEE concentration was not found.
|
The i-FAEE concentrations also increased from proximal to distal for this group. Exceptions were volunteers SH04 and SH09. However, for e-FAEEs, a decrease from proximal to distal was found more frequently. In some of the volunteers (SH02, SH06, SH09, and SH015), the FAEE concentration in sebum was much higher than in hair. The samples from these volunteers were collected in the first days of the year 2001, and these individuals reported consuming more drinks than usual during the Christmas holidays and on New Year’s Eve. It can be concluded that a single drinking event will primarily lead to an increased e-FAEE concentration and will not substantially affect the concentration in the hair matrix (i-FAEEs). Finally, it was observed that in this group, ethyl oleate in the hair as well as in the sebum was not as dominant as in both alcoholic groups A and B described above.
group d: teetotalers
Low FAEE concentrations (i-FAEE = 0.06–0.37 ng/mg; mean, 0.16 ng/mg) were also measured in hair from the teetotalers; one member of this group (VA01) was a female Moslem who declared never having consumed alcoholic beverages in her life. The results from group D are given in Table 5
, and the segmental i-FAEE concentrations of two cases with the typical increase from proximal to distal are shown in Fig. 3
. The highest values (i-FAEE = 0.37 ng/mg; e-FAEE = 1.4 ng/mg) were measured in the sample from case SH01. This individual was a hairdresser, which will be discussed later.
Comparisons of the ranges and mean values for i-FAEE, e-FAEE, and the concentration ratios of the four esters for groups A–D are shown in Table 6
. The FAEE concentrations increased in the series from teetotalers (group D) to moderate social drinkers (group C), patients in withdrawal treatment (group A), and fatalities with alcohol history (group B), with a clear difference between the alcoholic groups A and B and the nonalcoholic groups C and D.
|
Furthermore, a difference in the FAEE ratio was apparent between i-FAEEs and e-FAEEs. The e-FAEEs generally contained a higher portion of ethyl oleate at the expense of the three saturated esters. In the same way, ethyl oleate was clearly dominant in the i-FAEEs and e-FAEEs of alcoholic groups A and B compared with the nonalcoholic groups C and D.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
sebum as the main route of faee incorporation in hair
It follows from the results that the FAEEs are deposited mainly from sebum into hair. This deposition route can best explain the presence of high FAEE concentrations in hair segments that grew during periods of abstinence (e.g., see Fig. 2
) and the lack of agreement between drinking history and segmental concentrations generally found in these investigations.
The concentrations of the external lipids (e-FAEEs) given in the Tables 2–6
and in Figs. 1
and 2
always relate to the hair weight. The weight of the lipids determined after evaporation of the two n-heptane washings was <0.1–12% of the hair weight. That means that, as a rule, the FAEE concentration related to the extract weight was more than two orders of magnitude higher than that in hair (i-FAEE). This concentration gradient favors diffusion into the hair matrix.
The increase in FAEE concentrations from proximal to distal found in most cases could be explained by the sebum deposition route. The hair is continuously bathed by sebum, and this leads to an accumulation of the concentrations with increasing age, i.e., distance of the hair from the skin. Furthermore, long hair is usually shampooed more intensively near the skin. This may decrease the deposition rate from sebum in the proximal segments. Another reason for higher distal FAEE concentrations could be that aging hair allows better diffusion of the external lipids into the hair matrix.
A sebaceous gland exits into the pilary canal of each hair follicle, and the hair is already bathed in sebum before it reaches the skin surface. Sebum is produced by a holocrine mechanism (15)(16). The sebaceous cells differentiate from a small size in contact with the basement membrane through a maturation zone, where they accumulate lipids, to the zone of necrosis, where they increase 100- to 150-fold in volume as a result of lipid accumulation. The cells then undergo lysis and release their contents into the sebaceous duct. The main components of sebum are triglycerides, wax esters, cholesterol esters, and squalene (17). The fatty acid composition of triglycerides and wax esters is unique for sebum and also contains branched-chain acids, acids with an odd number of carbon atoms, and acids with an unusual placing of the double bond at position 
6. The proportions of branched-chain fatty acids in sebaceous wax esters are individually different and possibly genetically controlled (18). Synthesis of the acids and esterification occur within the cells. Free fatty acids in sebum are believed to be formed mainly outside of the glands by the action of bacterial lipases. There is a large individual variation in sebaceous lipid composition.
From this synthesis mechanism of the sebum constituents, it can be concluded that the fatty ethyl esters are also synthesized from ethanol within the sebaceous cells. Therefore, not the complete esters, but only ethanol must be transported from the blood vessels to the glands.
The cell transition time, i.e., the time between the germinative cell division and cell disintegration, has been determined as 9–25 days (16)(17). The average time between synthesis of sebum and its excretion was estimated at 8 days, leading to an overall transit time from germinative cell division to sebum of 13–14 days. From this it follows that the FAEEs should appear in sebum 8 days after alcohol consumption. However, because ester synthesis may proceed until cell disintegration, they could appear some days before.
Sebum production is controlled by hormones (16)(18); it therefore depends on age (19)(20)(21) and sex (22) and is subject to individual differences. It also depends to a certain degree on the season (23). In addition to the differences in hair treatment, this may be one more reason for the missing correlation between drinking amounts and FAEE concentrations in hair.
alternative incorporation mechanisms
It cannot be excluded that a smaller portion of the FAEEs is also incorporated from systemic blood circulation into the hair root or is formed from ethanol and fatty acids in the basal cells of the hair root (mechanisms a and b). The highly lipophilic esters should effectively penetrate the cell membranes. It can be seen from Tables 2–6
that the ratio between the four esters in the external lipids is characteristically different from that in the hair extracts. In all samples, the portion of ethyl oleate was clearly higher in the e-FAEEs than in the i-FAEEs. In the e-FAEEs of some alcoholics, oleate was by far the dominant ester (e.g., 100% in case HV07; Table 2
). In addition, the ratio between the remaining three esters is in most cases different in e-FAEEs than in i-FAEEs. There are three possible explanations for this: (a) the saturated esters are additionally incorporated by competitive mechanisms in the hair root to a higher extent; (b) the esters are deposited with a different efficiency from sebum with a particularly lower rate for ethyl oleate; or (c) the esters are to a different extent eliminated from hair by hydrolysis or oxidation. It can be concluded from the missing agreement between periods of drinking or abstinence and the segmental concentrations for all four esters that the last two explanations are more probable.
Nevertheless, hair also contains lipids that do not originate from sebum but from the membranes of the hair cells and that form a multilayered material after keratinization, the cell membrane complex. An essential part of these integral lipids is covalently bound to the cell surfaces and can be extracted only after hair hydrolysis (24)(25). This cell membrane complex could also contain lower amounts of FAEE already incorporated in the hair root and should be the host for the deposition of FAEE from sebum into the completed hair.
faee concentrations in hair of teetotalers
The origin of the low FAEE concentrations in these cases is not yet clear. Possible explanations include endogenous ethanol or formation from ethanol-containing hair care products. In a separate experiment with three volunteers, we found that after 2 months of daily treatment with an alcohol-containing hair lotion (62.5% ethanol by volume), FAEE concentrations were similar to those found in the alcoholics (Auwärter H, Sporkert F, Pragst F, unpublished results). Therefore, the values measured in the sample from the teetotaling hairdresser SH01 (0.37 ng/mg i-FAEEs and 1.4 ng/mg e-FAEEs) might have been caused by her increased use of hair cosmetics.
i-FAEE concentrations in hair segments as markers of alcohol consumption
In Table 6
, the ranges and means of the total concentrations of the four esters for the 6 cm proximal hair segments are compared among groups A–D. When we used a cutoff of 1.0 ng/mg, strong alcoholics could be distinguished from teetotalers and moderate social drinkers with relatively high accuracy. However, there were no drinkers with a problematic daily dose of, e.g., 30–60 g of ethanol among the investigated individuals. It can be expected that such borderline cases would be more difficult to recognize.
For the members of group A, the comparison of the i-FAEE concentration sum with -GT, GPT, and MCV produced a relatively high sensitivity of i-FAEEs for the detection of alcohol abuse. On the basis of a cutoff value of 1.0 ng/mg, 15 of the 19 cases (80%) were positive and 2 additional cases (HV06 and HV15) were just below this limit. No false-positive result was obtained for the nonalcoholic volunteers shown in Tables 4
and 5
. Only 59% of the -GT values, 47% of the GPT values, and 27% of the MCV values were above the upper limits of the reference intervals for the laboratory where the serum or blood markers were determined. In three cases (HV05, HV17, and HV18), two of whom had relapses after a longer period of abstinence, the excessive alcohol consumption was not detected by all three markers but was clearly seen from the high FAEE hair concentrations. However, cases HV07 and HV08, who had clearly negative hair results, also had laboratory values within the reference intervals. Carbohydrate-deficient transferrin was not measured in these cases and might have given better agreement.
Because teetotalers can also have low FAEE concentrations in hair, it is not unambiguously possible to confirm or refute the claim of strict abstinence. From the examples in Table 5
and other cases not involved in this study (12), it follows that a i-FAEE up to 0.4 ng/mg is not in contradiction to strict abstinence. In the case of higher values, an external reason should be investigated, such as regular use of ethanol-containing hair cosmetics if alcohol consumption can be excluded.
Despite these drawbacks, the quantitative analysis of FAEEs in hair can be an efficient tool to support the clinical diagnosis of alcohol abuse, particularly in combination with, or as supplement to the classic alcohol markers, which very often fail. In postmortem investigations and particularly in putrefied cases, the detection of alcohol abuse by blood or serum markers is very difficult or even impossible. In such cases, the measurement of the FAEE concentrations in hair can be a reliable method for evaluating alcohol abuse before death.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
-GT, -glutamyltransferase; MCV, erythrocyte mean cell volume; FAEE, fatty acid ethyl ester; e-FAEE, FAEE from the hair surface; i-FAEE, FAEE from internal hair structures; GPT, glutamate pyruvate transaminase; HS-SPME, headspace solid-phase microextraction; GC-MS, gas chromatography-mass spectrometry; and FAEE, sum of the concentrations of ethyl myristate, ethyl palmitate, ethyl oleate, and ethyl stearate in the proximal 0–6 cm hair sections. 
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The following articles in journals at HighWire Press have cited this article:
|
|
 |
|
  V. KULAGA, D. CAPRARA, U. IQBAL, B. KAPUR, J. KLEIN, J. REYNOLDS, J. BRIEN, and G. KOREN FATTY ACID ETHYL ESTERS (FAEE); COMPARATIVE ACCUMULATION IN HUMAN AND GUINEA PIG HAIR AS A BIOMARKER FOR PRENATAL ALCOHOL EXPOSURE Alcohol Alcohol., September 1, 2006; 41(5): 534 - 539. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. M. Wurst, S. Alexson, M. Wolfersdorf, G. Bechtel, S. Forster, C. Alling, S. Aradottir, K. Jachau, P. Huber, J. P. Allen, et al. CONCENTRATION OF FATTY ACID ETHYL ESTERS IN HAIR OF ALCOHOLICS: COMPARISON TO OTHER BIOLOGICAL STATE MARKERS AND SELF REPORTED-ETHANOL INTAKE Alcohol Alcohol., January 1, 2004; 39(1): 33 - 38. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Hartwig, V. Auwarter, and F. Pragst FATTY ACID ETHYL ESTERS IN SCALP, PUBIC, AXILLARY, BEARD AND BODY HAIR AS MARKERS FOR ALCOHOL MISUSE Alcohol Alcohol., March 1, 2003; 38(2): 163 - 167. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Klein, D. Chan, and G. Koren Neonatal Hair Analysis as a Biomarker for in Utero Alcohol Exposure N. Engl. J. Med., December 19, 2002; 347(25): 2086 - 2086. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
