Since the majority of patients with gram-negative infection are elderly, reference values should .... serum albumin. To remove inhibitors, the serum samples were.
CLINICAL AND DIAGNOSTIC LABORATORY IMMUNOLOGY, Nov. 1994, p. 684-688
Vol. 1, No. 6
1071-412X/94/$04.00+0 Copyright X 1994, American Society for Microbiology
Endotoxin Levels in Sera of Elderly Individuals T. GOTO,"2t S. EDEN,3 G. NORDENSTAM,1 V. SUNDH,4 C. SVANBORG-EDIN,5 AND I. MATFSBY-BALTZER2* Departments of Clinical Immunology,' Clinical Bacteriology,2 Physiology,3 and Geriatrics and Long-Term-Care Medicine,4 University of Goteborg, 413 46 Goteborg, and Department of Clinical Immunology, University of Lund, Lund,5 Sweden Received 20 October 1993/Returned for modification 25 January 1994/Accepted 20 July 1994
The endotoxin levels in serum of 377 72-year-old individuals were quantitated. The study population was a representative sample of this age group and was participating in a general study of health and disease among the elderly in Goteborg, Sweden. The endotoxin levels in serum were quantified by the chromogenic Limulus amebocyte lysate assay and were correlated with the health status and laboratory findings for each individual. The mean endotoxin levels (± 1 standard deviation) in men and women, when excluding four outliers, were 6.6 -- 3.8 and 6.9 ± 3.8 pg/ml, respectively. All included, 21.5% of individuals had endotoxin levels equal to or above the sensitivity limit of 10 pg/ml. Strong positive correlations were found between endotoxin levels and plasma triglycerides (P > 0.995) and between endotoxin levels and serum protein (P > 0.9875). The endotoxin activity also correlated with mean corpuscular hemoglobin concentration (P < 0.005, negative correlation), body mass index (P > 0.9875), and decreased appetite (P > 0.9875). A high alcohol consumption was associated with increased endotoxin levels (P = 0.995). There are no previous studies which examine endotoxin levels in serum samples from individuals representative of the population. This study showed that elderly individuals had the same mean level of endotoxin as has been found in other age groups. The increased endotoxin levels seen in heavy drinkers may be explained by a decreased ability of the liver to remove endotoxin. The correlations found between endotoxin and triglycerides, protein, mean corpuscular hemoglobin concentration, decreased appetite, and body mass index are discussed.
examinations of the subjects were performed at the Geriatric Vasa Hospital, Goteborg. Questions concerning physical problems, symptoms, and previous diagnoses, investigations, and treatments were included in the questionnaire. The subjects were also asked about daily smoking habits, diet, bowel habits, prostatic disorders, patterns of micturition and other natural functions, sleep, housing, and social conditions. Cancer diagnoses were taken from the Swedish Cancer Registry and confirmed by hospital records when available. A detailed physical examination was performed by physicians and nurses. Serum samples were available from 75% of the participants for endotoxin analysis (Table 1). Overall, 24% of the individuals analyzed for endotoxin were healthy, 46% were adequately treated for illness at the time of the medical examination, 27% showed findings that required treatment or further investigation, and the remaining 3% constituted a combination of the first two groups. Blood sampling and laboratory methods. After the individuals underwent an overnight fast, venous blood was collected in vacuum tubes. The blood was collected in EDTA-containing tubes for hematological analyses, in heparin-containing tubes for plasma analyses, and in tubes without additives for serum analyses (7a, 17). Serum samples were transferred to glass ampoules and sealed. These samples, which were stored at -20°C, were used for endotoxin analyses. In all, 31 blood components were examined, and their levels were compared with the endotoxin levels (17). Of these 31 parameters, 12 are shown in Table 3 (17). Hemoglobin was analyzed by a Coulter S Plus model II analyzer. Mean corpuscular hemoglobin concentration (MCHC) was determined by dividing the hemoglobin concentration by the erythrocyte volume fraction, which also was analyzed (results not shown). The laboratory methods for erythrocyte sedimentation rate, creatinine, aspartate aminotransferase in serum, alanine aminotransferase in serum, alkaline phosphatase in serum, bilirubin, plasma triglycerides (TG), protein in serum, and calcium in serum were applied as described previously (17). The level of C-reactive protein in
The lipid A moiety of lipopolysaccharide (LPS) is the endotoxic principle of gram-negative bacteria (9). Its many biologic activities include the induction of fever, inflammation, and shock. In view of these pronounced activities, measurements of circulating endotoxin levels have been expected to be of value in predicting the risk for shock and death in gramnegative infection (32). Endotoxin can be quantified by the Limulus amoebocyte lysate assay. Several studies have described the endotoxin levels in septic patients, but few have addressed the normal levels in healthy individuals (5, 25, 32). This is essential when introducing endotoxin measurements into the clinical routine. Since the majority of patients with gram-negative infection are elderly, reference values should include this age group. The aim of the present study was to determine the endotoxin levels in serum in a representative sample of 72-year-old individuals with known health status. Any associations between the endotoxin levels and illness, social parameters, or various laboratory parameters of the blood were tested.
MATERIALS AND METHODS Population. The study group consisted of 72-year-old individuals taking part in a general study of aging and health in the elderly in Goteborg, Sweden, and who had been monitored from the age of 70 (23, 26, 31, 32). The number of subjects initially selected for the study made up 23% of the total 70-year-old population (Table 1) (31). The participation rates were 77% at age 70 and 63% at age 72. The study has been described in detail by Eriksson et al. (7a). Interviews and * Corresponding author. Mailing address: Department of Clinical Bacteriology, University of Goteborg, Guldhedsgatan 10, 413 46 Goteborg, Sweden. Phone: 46-31-604728. Fax: 46-31-604975. t Present address: Department of Virology, Faculty of Medicine, Kagoshima-shi 890, Japan.
684
ENDOTOXINS IN SERA OF ELDERLY PERSONS
VOL. 1, 1994
TABLE 1. Number of individuals participating in the study of aging and health in the elderly in Goteborga and the total number of serum samples analyzed Characteristic
Participants at age 70 Participants at age 72 Serum not available Serum excluded Serum tested
No. of men
302 239 11 44 184
No. of women
317 265 12 60 193
685
100
100
90
90
-
80
80 aD
t
70
rd
60
70 3 0
103b
44
50
377
0
Total no.
619 504 23
a See references 7a and 31. bA sequence of serum samples was contaminated.
serum was determined by immunodiffusion (19). Body mass index was defined according to the Quetelet index as weight in kilograms divided by the square of the height in meters (10). Chromogenic Limulus amoebocyte lysate assay, CLAL Sterile pyrogen-free plastic test tubes (12 by 75 mm; Grainer) or glass tubes (12 by 75 mm; Kimble) were used. Glass tubes were rendered pyrogen free by being heated to 180°C for 4 h. The CLAL test kit was obtained from Chromogenix (formerly KABI Diagnostica), Molndal, Sweden. Escherichia coli 0111B4 LPS was included in the test kit as standard LPS (100 pg of LPS corresponding to 1.2 endotoxin units [EU]). The LPS standard concentrations ranged from 1 to 10 pg/ml. The sensitivity of the assay was 1 pg/ml. The standard solutions were diluted with pyrogen-free water containing 0.1% human serum albumin. To remove inhibitors, the serum samples were diluted 1/10 in pyrogen-free water and heat treated for 5 min at 75°C. Accordingly, the sensitivity limit in serum samples was 10 pg/ml. Below this limit, the values were less reliable. The analysis was performed as specified in the manual by the manufacturer. Pyrogen-free microplates were purchased from Sterilin (Hounslow, United Kingdom). In brief, samples (50 ,ul) were added to the wells of a microplate placed on a dry incubator at 37°C (IPS, Diagnostic Pasteur, France). After 3 to 5 min of incubation, the Limulus amoebocyte lysate solution (50 ,ul) was added. Then, 30 min later, the prewarmed substrate solution (100 ,ul) was added to the wells. The color reaction was stopped after 5 min by the addition of 100 ,ul of 20% acetic acid. The plate was read at 405 nm on a microplate reader (Emax; Molecular Devices, Menlo Park, Calif.). Five standard points were included in each run. The day-to-day coefficient of variation was analyzed by running the standard LPS at a concentration of 7.5 pg/ml diluted in 10% pooled normal human serum on 10 different days along with the serum analyses. This sample was also heat treated in the same way as the other serum samples to be tested. The coefficient of variation was found to be 8%. Statistics. The data were statistically analyzed by computer programs specially developed for the Goteborg study (4). Analyses of linear trend and group difference were performed by the nonparametric permutation test (24). This test includes as special cases Fisher's exact test (when both of the variables are dichotomous), the Mann-Whitney test (when one variable is dichotomous and the other is recoded to the rank), and Spearman's rank correlation (when both variables are recoded to the rank). The nonparametric permutation test was used because the variables consisted of discrete, ordinal, or interval scaled measures. In addition, the distribution of many of the interval scaled variables was highly non-Gaussian. When testing for correlation, the rank of endotoxin was used to avoid having significance caused by a few extremely high values. For ties, the mean rank was used. P values lower than 0.0125 or higher than 0.9875 were considered statistically significant.
-
Q) 4
Z
1
60
-
305o 4
40
40
30
30
20 10
0
'I I
0 -
a) 20
P.4
a)
10o
0
20
Endotoxin (pg/ml) FIG. 1. Distribution of endotoxin levels in sera from 377 individuals at the age of 72 years. Hatched bars represent men, and open bars represent women. The accumulated frequency in the percentage of the total number of individuals is also shown (solid line). The 95th percentile is indicated for the total group.
RESULTS The distribution of endotoxin levels in serum in the test population is shown in Fig. 1 and Table 2. The endotoxin levels did not follow a Gaussian distribution. Four individuals, one man and three women, had endotoxin levels of 654, 150, 225, and 773 pg/ml. A total of 81 individuals (21.5%) had endotoxin levels equal to or above the sensitivity limit of 10 pg/ml (Table 2). The mean endotoxin levels in men and women, excluding the four outliers, were similar (6.6 ± 3.8 pg/ml and 6.9 ± 3.8 pg/ml, respectively) (Table 3). The median values in both men and women were 6 pg/ml, the 97.5th percentile being 17 pg/ml. The endotoxin levels in serum in all individuals were further analyzed in relation to parameters of health and disease in the population (Table 4). There was no correlation with the social parameters analyzed. Among the morbidity parameters, the most significant correlation was seen between endotoxin levels and a decreased appetite in women (rank correlation, r = 0.1694) or both men and women taken together (r = 0.1388) (Table 4). A weak correlation was observed for the whole group when endotoxin levels were analyzed against the presence of dyspnea (r = 0.1209). Among the laboratory parameters, the levels of TG (women, r = 0.2683; men, r = 0.2401) and protein (women, r = 0.1860; men, r = 0.2077) were strongly correlated with the endotoxin levels in women and men and in both women and men taken together (r = 0.2559 for TG; r = 0.1908 for protein). Body mass index (women, r = 0.2211), subscapular skin depth (women, r = 0.2377), and MCHC TABLE 2. Distribution of endotoxin levels in men and women at 72 years of age Endotoxin concn
No. (%) of men
No. (%) of women
Total no. (%)
0-9
149 (81.0) 29 (15.8) 3 (1.6) 1 (0.5) 2 (1.1)
147 (76.2) 32 (16.6) 9 (4.6) 2 (1.0) 3 (1.6)
296 (78.5) 61 (16.2) 12 (3.2) 3 (0.8) 5 (1.3)
(pg/ml) 10-14 15-19 20-24
.25
686
CLIN. DIAGN. LAB. IMMUNOL.
GOTO ET AL.
TABLE 3. Endotoxin levels in men and women at 72 years of age Concn of endotoxin (pg/mi)
Subject
teNstoed 1 SD
Median (+ 97.5th percentile)
Valueb in
Mean ± 1 SD
men
Parameterf
excluding outliers
Men Women
184 193
10 ± 47 13 ± 58
6.0 (17) 6.0 (17)
6.6 ± 3.8b 6.9 ± 3.8c
Total
377
11 ± 53
6.0 (17)
6.8 ± 3.8
a SD, standard deviation. b Excluding one value. c Excluding three values.
-0.1685) correlated with endotoxin levels in women, although, taken together, both men and women also showed a significant correlation (r = 0.1420, 0.1698, and -0.1612, respectively) (Table 4). A negative correlation was observed between MCHC and endotoxin. In addition, calcium levels correlated positively with endotoxin levels for men and women taken together (r = 0.1373) (Table 4). A weaker correlation was seen for creatinine or bilirubin and only when the whole population was analyzed (r = 0.1156 and -0.1277, respectively). All individuals were further divided into two groups according to whether the endotoxin levels were equal to, above, or below the sensitivity limit of 10 pg/ml (Table 4, right-hand column). With this subdivision, analysis of the correlation between endotoxin and the various parameters showed a correlation pattern partly similar to that when the endotoxin values were ranked (Table 4). Significant correlations were seen with TG, protein, MCHC, body mass index, and decreased appetite. The mean values for MCHC, TG, protein, and body mass index are shown for the two groups in Fig. 2. In an interview with 73 men and 68 women among the 377 participants, 5 men reported a daily or almost daily intake of .12 cl of alcoholic beverage. Four of these five men had endotoxin levels above 10 pg/ml, while only 16 of the 68 men drinking no alcohol or drinking no more than once a week had levels above 10 pg/ml (Table 4). No women reported daily alcohol consumption. The 20 individuals with high endotoxin levels, outside the 95th percentile (>14 pg/ml), were further analyzed (Table 5). Weight and TG were higher in these individuals than in those whose endotoxin levels were within the 95th percentile. Among the females, weight was significantly higher and MCHC was significantly lower. The four highest endotoxin levels (>150 pg/ml) correlated only with MCHC when compared with the others (P = 0.0098). (women,
TABLE 4. Correlation of endotoxin levels in serum in 184 men and 193 women with some social and medical parameters
r =
(n = 184)
Valueb in women (n = 193)
n
P
n
P
Social parameters Living in institution Coffee drinking Alcohol consumption ADL, retarded
7 169 73e 72
_d ND -
6 186 68e 95
ND -
Morbidity Myocardial infarction Cancer (all) Diabetes Hypertension Dementia Bacteriuria, screening Dyspnea Decreased appetite
20 27 15 41 23 9 29 6
-
10 26 14 54 16 32 21 10
166 157 184 184 177
-
176 161 191 190 183
Laboratory parameters Body mass index Subscapular skin depth Hemoglobin MCHC ESR CRP Creatinine ASAT ALAT ALP Bilirubin TG
Protein Calcium Mortality Died within 2 years
-
182 183 183 177 176 180 146 170
** -
192 190 190 182 181 187 162 179
14
-
15
***
Total
Total P for
pb
P > 0.9875; **, 0.005 > P > 0.995; ***, 0.0005 > P> 0.9995; ND, not determined. P values are two tailed. c Individuals with endotoxin levels of .10 pg/ml were compared with those with endotoxin levels of
0
0.5
332 E0.0
ENDOTOXIN
ENDOTOXIN
71
30
x Q
70
28
z
z 69 o 0.
68
vt
26
>.
24
0
m
67 66
< 10 pg/ml
> 10 pg/ml
22
20
< 10
ENDOTOXIN
pg/ml
> 10 pg/ml
ENDOTOXIN
FIG. 2. Comparison of MCHC (a), TG (b), protein level in serum (c), and body mass index (d) in elderly individuals having an endotoxin level in serum of 20 pg/ml, in patients with chronic renal failure and in those on hemodialysis (20). However, the endotoxin levels could not be linked to symptoms characteristic of endotoxemia. The increased endotoxin levels in this group of patients may partly be due to a decreased clearance capacity of the reticuloendothelial system. Blood monocytes or granulocytes of uremic patients express a decreased phagocytic capacity (27). Thus, the
TABLE 5. Analysis of elderly individuals with endotoxin levels outside the 95th percentilea compared with those within the 95th percentile P value' for: Parameter
Wt TG MCHC
Men
Women
Total
NSC NS NS
0.9935 NS 0.0046
0.9894 0.9934 NS
a Levels outside the 95th percentile are >14 pg, which was found in 6 men and 14 women. b p values are two tailed. c NS, not significant.
687
clinical significance of endotoxemia may vary depending on the patient group. The endotoxin levels correlated positively with parameters such as body mass index, skin depth, and TG, which are all increased in obese persons. It is well known, however, that TG also increases in humans or animals during infection or when exposed to endotoxin, the underlying factor being tumor necrosis factor-induced reduction of lipoprotein lipase activity (14, 30). TG and TG-rich lipoproteins have also been associated with endotoxin-neutralizing activity, as analyzed by the Limulus amoebocyte lysate assay and the ability to prevent death from sepsis in rats (2, 12, 13). Anorexia, weight loss, and anemia are all readily induced by endotoxin, as well as increased TG concentrations. A decreased appetite was found to be associated with increasing endotoxin levels, especially in women, in our study. Weight loss, however, was not monitored in this study. Although no parameters of anemia (hemoglobin level, hematocrit, erythrocyte count) were significantly correlated with increasing endotoxin levels, a descriptive parameter of erythrocytes, the MCHC, correlated negatively with endotoxin levels. Reduced MCHCs have been reported in anemic patients with chronic infections (6). Endotoxin also correlated with increasing protein levels in serum. This observation may well be consistent with the role of endotoxin as an inducer of acute-phase proteins via mediators such as tumor necrosis factor, interleukin-1, and interleukin-6, which also augment the synthesis of other proteins (16). Increased protein levels in serum have also been found in patients with chronic inflammation. The correlations between endotoxin and protein levels, TG, MCHC, and a decreased appetite were also found when comparing individuals having endotoxin levels equal to or above the sensitivity limit of 10 pg/ml with those having levels below that limit. Taken together, these correlations could be indications of subacute infections. Asymptomatic infections such as urinary tract infections or pulmonary infections are not uncommon in elderly people. No correlation, however, was established between urinary tract infections found at screening and endotoxin levels. An association of Chlamydia pneumoniae infection with chronic coronary heart disease and acute myocardial infarction has been suggested (28). Immune complexes containing chlamydial LPS were found in 57% of the patients with acute myocardial infarction compared with 12% of the controls (18). Thus, the connection between high TG levels and chronic coronary heart disease may have chronic infection as one explanation, which could result in increased endotoxin levels in serum. Since the elderly individuals in our study participated in a prospective health study, the relevance of our findings for the future health status may be evaluated. The endotoxin levels of clinical significance with regard to gram-negative bacteremia or septicemia have varied in different studies (7, 25, 33). In previous studies, 5, 10, and 100 pg/ml have all been used as limits for a positive test in evaluations of the Limulus amoebocyte lysate assay as a help in the diagnosis of patients with gram-negative bacteremia or at risk of developing septicemia. In our study, four individuals (1%) showed endotoxin levels of 150 pg/ml or more. Obviously, the endotoxin levels in these elderly people were not high enough to induce symptoms of endotoxemia, since the probands were not acutely ill at the time of blood sampling. Two women, one of whom suffered from obesity, had been surgically treated for gallstones 8 years earlier. A third woman suffered from hypertonia and a wheeze in her chest. One man suffered from back pain. Hyporesponsiveness (tolerance) to endotoxin is a wellknown phenomenon. The "early" tolerance state occurs upon rechallenge with endotoxin after an adequate interval or during continuous exposure over a longer period in both
688
GOTO ET AL.
humans and animals (11). Tolerance induced in the early phase of endotoxin exposure is not due to antibodies against the endotoxin. The cells of the reticuloendothelial system are held responsible for the refractory phase of the host, possibly because of decreased tumor necrosis factor production, which has been seen during induction of endotoxin tolerance in rats (29). Tumor necrosis factor is one of the major mediators of endotoxin shock. A hyporesponsive state of the host could explain the tolerance to the high endotoxin levels in the four individuals. Endotoxin is rapidly cleared from the blood by the reticuloendothelial system, especially the liver. An increased level of endotoxin in blood may arise from an abnormal influx of endotoxin from the gut, overwhelming the elimination system, or from a defective ability of the liver to take up and degrade endotoxin. Patients with liver cirrhosis or chronic inflammatory bowel disease have increased levels of endotoxins in blood (3, 34). No liver diseases, however, were found in our study group. Endotoxemia has also been associated with alcoholic liver injury (22) and obstructive jaundice (1). Heavy alcohol intake in the absence of liver disease caused transient endotoxemia. Individuals with a daily intake of alcohol had increased endotoxin levels in this study (3). A high alcohol consumption is also associated with increased TG levels. In conclusion, endotoxin levels of 10 pg/ml or more in the elderly may be connected with high alcohol intake, obesity, or possibly asymptomatic infections. Whether these correlations are also present in younger individuals is not known.
CLIN. DIAGN. LAB. IMMUNOL. 10. Garrow, J. S., S. E. Blaza, P. M. Warwick, and M. A. Ashwall. 1980. Predisposition to obesity. Lancet i:1103-1104. 11. Greisman, S. E. 1983. Induction of endotoxin tolerance, p. 149179. In A. Nonotny (ed.), Beneficial effects of endotoxins. Plenum Publishing Corp., New York. 12. Harris, H. W., E. B. Eichbaum, J. P. Kane, and J. H. Rapp. 1991.
13.
14.
15. 16. 17.
18.
19.
20. 21.
ACKNOWLEDGMENTS This study was supported by the Swedish Medical Research Council (grant 16X-09488) and by Chromogenix, Molndal, Sweden. This study is part of the Gerontological and Geriatric Population Study in Goteborg, Sweden (project group: B. Steen [project leader], S. Berg, H. Djurfalt, S. Eden, S. Landahl, D. Mellstr6m), and supported by grants from the Delegation for Social Research within the Ministry of Health and Social Affairs, the Goteborg Medical Services Administration, the Medical Research Council, the Goteborg Medical Society, and Stiftelsen Gamla Tjanarinnor. REFERENCES 1. Bailey, M. E. 1976. Endotoxin, bile salts and renal function in obstructive jaundice. Br. J. Surg. 73:774-778. 2. Berger, D., S. Schleich, and M. Seidelmann. 1990. Correlation between endotoxin-neutralizing capacity of human plasma as tested by the Limulus-amoebocyte-lysate-test and plasma protein levels. FEBS Lett. 277:33-36. 3. Bode, C., et al. 1987. Endotoxemia in patients with alcoholic and non-alcoholic cirrhosis and in subjects with no evidence of chronic liver disease following acute alcoholic excess. J. Hepatol. 4:8-14. 4. Bradley, J. V. 1968. Distribution-free statistical tests, p. 68-86. Prentice-Hall, London. 5. Brandtzaeg, P., P. Kierulf, P. Gaustad, A. Skulberg, J. Bruun, S. Halvorsen, and E. Sorensen. 1989. Plasma endotoxin as a predictor of multiple organ failure and death in systemic meningococcal disease. J. Infect. Dis. 159:195-204. 6. Cheruow, B., and S. F. Wollner. 1978. Is the anemia of chronic disorders normocytic-normochromic? Milit. Med. 143:345-346. 7. Dolan, S. A., L. Riegle, R. Berzofsky, and M. Cooperstoclk 1987. Clinical evaluation of the plasma chromogenic Limulus assay. Prog. Clin. Biol. Res. 231:405-416. I- . G., D. MelistrOm, and A. Svanborg. 1987. Medical-social 7a.Er, intervention in a 70-year-old Swedish population. A general presentation of methodological experience. Compr. Gerontol. C. 1:49-56. 8. Fribewr, P. 1985. The design of a reliable endotoxin test. Prog. Clin. Biol. Res. 189:139-149. 9. Galanos, C., 0. Liuderitz, E. T. Rietschel, and 0. Westphal. 1977. Newer aspects of the chemistry and biology of bacterial lipopolysaccharide, with special reference to their lipid A component. Int. Rev. Biochem. 14:288-289.
22.
23. 24. 25.
26.
27.
28.
29. 30. 31. 32.
33. 34.
Detection of endotoxin in triglyceride-rich lipoproteins in vitro. J. Lab. Clin. Med. 118:186-193. Harris, H. W., C. Grunfeld, K. R. Feingold, T. E. Read, J. P. Kane, A. L. Jones, E. B. Eichbaum, G. F. Bland, and J. H. Rapp. 1993. Chylomicrons alter the fate of endotoxin, decreasing tumor necrosis factor release and preventing death. Clin. Invest. 91:1028-1034. Kawakami, M., and A. Cerami. 1981. Studies of endotoxin-induced decrease in lipoprotein lipase activity. J. Exp. Med. 154:631-639. Kinsey, S. E., and S. J. Machin. 1988. Endotoxaemia in neutropenic patients. Lancet ii:345. Kuschner, I. 1982. The phenomenon of the acute phase response. Ann. N. Y. Acad. Sci. 389:39-48. Landahl, S., R. Jagenburg, and A. Svanborg. 1981. Blood components in a 70-year-old population. Clin. Chim. Acta 112:301-304. Leinonen, M., E. Linnanmiiki, K. Mattila, M. Nieminen, V. Valtonen, M. Leirisalo-Repo, and P. Saikku. 1990. Circulating immune complexes containing chlamydial lipopolysaccharide in acute myocardial infarction. Microb. Pathog. 9:67-73. Nilsson, L. A. 1968. Comparative testing of precipitation method for quantification of C-reactive protein in blood serum. Acta Pathol. Microbiol. Scand. 73:129. Nisbeth, U., R. Hallgren, 0. Eriksson, and B. G. Danielson. 1987. Endotoxemia in chronic renal failure. Nephron 45:93-97. Nitsche, D., M. Kriewitz, A. Rossberg, and H. Hamelmann. 1987. The quantitative determination of endotoxin in plasma samples of septic patients with peritonitis using the chromogenic substrate and its correlation with the clinical course of peritonitis. Prog. Clin. Biol. Res. 231:417-429. Nolan, J. P. 1981. Endotoxin, reticuloendothelial function and liver injury. Hepatology 1:458-465. Nordenstam, G., V. Sundh, K. Lincoln, A. Svanborg, and C. Svanborg-Eden. 1989. Bacteriuria in representative samples of individuals 72 to 79 years of age. Am. J. Epidemiol. 130:1176-1186. Oden, A., and H. Wedel. 1975. Arguments for Fisher's permutation test. Ann. Stat. 3:518-520. Pearson, F. C., J. Dubczak, M. Weary, G. Bruszer, and G. Donohue. 1985. Detection of endotoxin in the plasma of patients with gram-negative bacterial sepsis by the Limulus amoebocyte lysate assay. J. Clin. Microbiol. 21:865-868. Rinder, A., S. Roupe, B. Steen, and A. Svanborg. 1975. Seventyyear-old people in Gothenburg. A population study in an industrialized swedish city. General presentation of the study. Acta Med. Scand. 198:397-407. Ringvir, S., L. Van Looy, P. Van de Heyning, and G. Leroux-Reels. 1975. Impairment of phagocytic activity of macrophages as studied by the skin-window test in patients on regular hemodialysis treatment. Clin. Nephrol. 4:234. Saildk, P., ML Leinonen, and K Mattila. 1988. Serological evidence of an association of a novel Chlamydia TWAR, with chronic coronary heart disease and acute myocardial infarction. Lancet ii:983-985. Sanchez-Cantu, L., H. N. Rode, and N. V. Christon. 1989. Endotoxin tolerance is associated with reduced secretion of tumor necrosis factor. Arch. Surg. 124:1432-1436. Semb, H., J. Peterson, J. Tavernier, and T. Olivecrona. 1987. Multiple effects of tumor necrosis factor on lipoprotein lipase in vivo. J. Biol. Chem. 262:8390-8394. Steen, B., and H. Djurfeldt. 1993. The gerontological and geriatric population studies in Gothenburg, Sweden. Z. Gerontol. 26:163-169. Svanborg, A. 1977. Seventy-year-old people in Gothenburg. A population study in an industrialized Swedish city. II. General presentation of social and medical conditions. Acta Med. Scand. Suppl. 611:5-37. Van Deventer, S. J. H., H. R. Buller, J. T. Ten Cate, A. Sturk, and W. Pauw. 1988. Endotoxaemia: an early predictor of septicaemia in febrile patients. Lancet i:605-608. Wellman, W., P. C. Fink, F. Benner, and F. W. Schmidt. 1986. Endotoxemia in active Crohn's disease. Treatment with whole gut irrigation and 5-aminosalicylic acid. Gut 27:814-820.
