Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 7-Day Trial for You or Your Team.

Learn More →

Distribution of Serum C-Reactive Protein and Its Association with Atherosclerotic Risk Factors in a Japanese Population : Jichi Medical School Cohort Study

Distribution of Serum C-Reactive Protein and Its Association with Atherosclerotic Risk Factors in... Abstract The distribution of serum C-reactive protein (CRP) levels and their association with age, sex, and atherosclerotic risk factors were studied in a large Japanese population between 1992 and 1995. The subjects consisted of 2,275 males and 3,832 females aged 30 years and over. CRP was measured by nephelometry. The distribution of CRP was highly skewed toward a lower level than that of previous studies and seemed to be a combination of two separate distribution curves. The increase in CRP with age was statistically significant, and males had higher CRP levels than did females. Males who were current smokers had higher CRP levels than did nonsmokers. Age, systolic blood pressure, diastolic blood pressure, triglycerides, fibrinogen, and body mass index were all positively associated with CRP in both sexes, while total cholesterol and blood glucose were positively related in females only. High density lipoprotein cholesterol was inversely related in both sexes. Multiple logistic regression analysis showed that sex, age, systolic pressure, high density lipoprotein cholesterol, triglycerides, fibrinogen, and body mass index were significant independent variables. In conclusion, the distribution of CRP among the Japanese was quite different from that among Westerners, although CRP levels correlated with other atherosclerotic risk factors, similar to those in Westerners. C-reactive protein, cross-sectional studies, risk factors CRP, C-reactive protein, CV, coefficient of variation Although atherosclerosis is multifactorial, it is now widely recognized that systemic and localized inflammation in the arteries contributes to the initiation or promotion of cardiovascular disease (1). C-reactive protein (CRP) is a protein of the acute-phase reaction of inflammation. It has been shown that elevated levels of serum CRP are related to a poorer outcome in patients with cardiovascular diseases (2, 3). Recently, several studies have suggested that CRP is a predictor of subsequent cardiovascular events in healthy adults (4–7). However, to our knowledge, there have been only two reports on the distribution of serum CRP levels in a large-scale, population-based study that included both males and females and used a hypersensitive assay (8, 9). Moreover, the relation between CRP and atherosclerotic risk factors in females has been reported in only a few studies (6–8, 10). The Jichi Medical School Cohort Study is a prospective, population-based study that aims to explore the risk factors for cardiovascular and cerebrovascular diseases in Japan. The subjects included people of both sexes aged 30 years and over residing in areas throughout Japan. In this report, we show the distribution of serum CRP levels and the relation between serum CRP levels and atherosclerotic risk factors in a large group of healthy people aged 30 years and over. MATERIALS AND METHODS The study design and some descriptive data have been presented previously (11–13). The population for our study included residents of nine rural communities in Japan, including Yamato, Takasu, Wara, Sakuma, Hokutan, Sakugi, Ohkawa, Ainoshima, and Akaike. In Japan, mass screening for cardiocerebrovascular diseases has been conducted since 1982, according to the Health and Medical Service Law for the Aged Act of 1981. Invitations to this mass screening were issued by government offices in each community, and personal invitations were also sent to all the subjects by mail. However, the invitations mentioned that those who were receiving treatment at hospitals or clinics for cardiovascular or cerebrovascular diseases did not need to take the examination. As a result, 2,573 males and 4,186 females aged 30 years and over participated in these examinations. In this study, we excluded those who had had stroke or myocardial infarction as determined by a questionnaire or interview and those from whom some data were incomplete. The final number of study subjects was therefore 2,275 males and 3,832 females. The overall response rate was 52.4 percent; the rates in each community were 51.0 percent in Yamato, 63.0 percent in Takasu, 90.0 percent in Wara, 88.0 percent in Sakuma, 27.0 percent in Hokutan, 38.0 percent in Sakugi, 66.0 percent in Ohkawa, 44.0 percent in Ainoshima, and 26.0 percent in Akaike. To obtain uniform information, we established a central committee composed of the chief medical officers from all participating areas, which developed a detailed manual for data collection. Medical history, habitual food intake, smoking habits, menopausal status, and alcohol consumption were assessed by a questionnaire developed by the committee. Females were considered postmenopausal if their menses had ceased naturally at least 12 months previously. Body height was measured in stocking feet. Body weight was recorded with the subject clothed, and 0.5 kg in summer or 1 kg in other seasons was subtracted from the recorded weight. The body mass index was calculated as weight (kg)/height (m)2. Waist and hip circumferences were collected as optional measurements in Takasu, Wara, Sakuma, Ohkawa, and Ainoshima. There were 921 male and 1,276 female subjects. Waist circumference was measured at the level of the high point of the iliac crest, and hip circumference was measured at the level of maximum extension of the buttocks. The waist-to-hip ratio, calculated as waist circumference divided by hip circumference, was used as an indicator of abdominal visceral fat (14). The systolic and diastolic blood pressures were measured with a fully automated sphygmomanometer (BP203RV-II, Nippon Colin, Komaki, Japan), placed on the right arm of a seated subject who had rested in a sitting position for at least 5 minutes before the measurement. Blood samples were drawn from the antecubital vein of seated subjects with minimal tourniquet use. Specimens were collected in siliconized vacuum glass tubes containing a 1/10 volume of 3.8 percent trisodium citrate for fibrinogen, sodium fluoride for blood glucose, and no additives for lipids, respectively. Tubes were centrifuged at 3,000 g for 15 minutes at room temperature. After separation, the serum samples were stored at 4°C in refrigerated containers if analysis was to be performed within a few days. Otherwise, the samples were frozen until analysis. Plasma samples were frozen as rapidly as possible to −80°C for storage until laboratory examination could be performed. CRP levels were measured by using nephelometry, a latex particle-enhanced immunoassay (NA Latex CRP Kit, Dade Behring, Tokyo, Japan). The value in the calibrator was assigned from the Certified Reference Material 470 (IRMM, Geel, Belgium), an international plasma protein reference material. The material has achieved international standardization in the assay of CRP. The function of the assay was found to be satisfactory (15). Its interassay and intraassay coefficients of variation (CV) were 1.18 and 1.36 percent, respectively. The assay is sensitive enough to detect 0.03 mg/liter of CRP. Undetectable CRP values were recorded as 0.015 mg/liter. Total cholesterol and triglycerides levels were measured by using an enzymatic method (Wako, Osaka, Japan; interassay CV: 1.5 percent for total cholesterol and 1.7 percent for triglycerides). High density lipoprotein cholesterol was measured by using the phosphotungstate precipitation method (Wako, Osaka, Japan; interassay CV: 1.8 percent). Blood glucose was measured by using an enzymatic method (Kanto Chemistry, Tokyo, Japan; interassay CV: 1.9 percent). Lipoprotein(a) levels were measured with an enzyme-linked immunosorbent assay kit (Biopool, Uppsala, Sweden; interassay CV: 3.51 percent). Fibrinogen levels were determined with a one-stage clotting assay kit (Data-Fi, Dade Behring, Miami, Florida; interassay CV: 2.5 percent). Serum insulin levels, which were collected in an optional examination that included 1,222 males and 1,663 females in Takasu, Wara, and Sakuma, were determined with a radioimmunoassay kit (Dainabot, Tokyo, Japan; interassay CV: 4.5 percent). The lower detection limit was 2.5 μU/ml, and insulin levels below this limit were taken as 2.0 μU/ml. Statistical methods Statistical analysis was performed using a Statistical Analysis System 6.12 edition (SAS Institute, Inc., Cary, North Carolina). Descriptive parameters were shown as the mean, standard deviation, and percentiles. For a comparison of the mean values, the unpaired t test, Mann-Whitney U test, or analysis of covariance with an adjustment for age was used. Categorical variables were analyzed by using the chi-square test. Serum CRP levels were divided into two categories, above and below 0.11 mg/liter, because the distribution appeared to be a combination of two separate distributions, with 0.11 mg/liter being the value at which these two distribution curves crossed. To assess the influence of each variable, logistic unconditioned regression models were used. Odds ratios were used to evaluate the association between CRP and other variables. As an additional observation, a multiple linear regression analysis was used in which the dependent variable was natural log-transformed CRP. In these multivariate analyses, we used two models, one that included and one that excluded the waist-to-hip ratio and insulin. Because these variables were measured as optional measurements in this study, they were taken for about one third of all the subjects in our study. The effect of menopausal status on CRP levels was analyzed in females aged 40–59 years, and those who had surgical menopause were excluded. The difference in CRP between premenopausal and postmenopausal women adjusted for age was studied with analysis of covariance. When parametric procedures were used, triglycerides, blood glucose, lipoprotein(a), and insulin were transformed into natural logarithms. A significant difference was defined as p < 0.05. RESULTS Figure 1 shows the distribution of CRP levels among the study subjects. The distribution of serum CRP levels was highly skewed to lower levels and ranged widely. The minimum value was less than 0.03 mg/liter, and the maximum value was 68.2 mg/liter, with 25th, 50th, and 75th percentile values of less than 0.03, 0.12, and 0.30 mg/liter, respectively. The shape of the distribution did not differ by sex, age, or community. In this study, serum CRP levels seemed to have two distributions. One was a lower distribution with a narrow range, and the other was a higher distribution with a very wide range. We considered this to be a combination of two separate distributions and divided CRP levels into two categories: those lower than or equal to 0.11 mg/liter and those higher than 0.11 mg/liter. This was the value at which the two distribution curves were seen to cross. The mean value and standard deviation of the first subpopulation, which had lower CRP levels, were 0.038 and 0.024 mg/liter, respectively. Those of the second population, which had higher CRP levels, were 1.30 and 4.25 mg/liter, respectively. FIGURE 1. View largeDownload slide Distribution of C-reactive protein levels (mg/liter), by age, for 2,275 males and 3,832 females, the Jichi Medical School Cohort Study, Japan, 1992–1995. Because the distribution of C-reactive protein ranged very widely, only those with levels lower than 1.2 mg/liter, which accounts for 92.3 percent of the males and 94.5 percent of the females, are included. FIGURE 1. View largeDownload slide Distribution of C-reactive protein levels (mg/liter), by age, for 2,275 males and 3,832 females, the Jichi Medical School Cohort Study, Japan, 1992–1995. Because the distribution of C-reactive protein ranged very widely, only those with levels lower than 1.2 mg/liter, which accounts for 92.3 percent of the males and 94.5 percent of the females, are included. Table 1 shows the mean CRP levels by age and sex. The mean CRP levels increased with age in both sexes, except for subjects aged under age 39 years. Males had higher CRP levels than did females in all age groups, with mean values of 0.83 and 0.59 mg/liter, respectively. This difference was statistically significant, except for subjects in their fifties. TABLE 1. C-Reactive protein levels (mg/liter) stratified by age and sex, for 2,275 males and 3,832 females, the Jichi Medical School Cohort Study, Japan, 1992–1995 Sex and age (years)  No.  Mean†  (SD)‡  Percentile   90  75  50  25  10  Males                   ≤39  268  1.00  (6.4)  0.64  0.28  0.10  0.04  <0.03   40–49  477  0.63  (2.2)  0.70  0.31  0.14  0.03  <0.03   50–59  467  0.64  (3.5)  0.69  0.31  0.13  <0.03  <0.03   60–69  865  0.89  (3.1)  0.89  0.38  0.18  0.04  <0.03   ≥70  198  1.26  (3.6)  2.56  0.61  0.26  0.08  <0.03   Total  2,275  0.83  (3.6)  0.78  0.35  0.16  0.04  <0.03  Females                   ≤39  402  0.46**  (2.7)  0.35  0.17  0.05  <0.03  <0.03   40–49  739  0.34**  (1.7)  0.44  0.21  0.05  <0.03  <0.03   50–59  964  0.62  (2.9)  0.55  0.25  0.09  <0.03  <0.03   60–69  1,477  0.68*  (3.0)  0.74  0.33  0.14  0.04  <0.03   ≥70  250  0.87*  (3.0)  0.83  0.44  0.22  0.05  <0.03   Total  3,832  0.59**  (2.7)  0.57  0.28  0.09  <0.03  <0.03  Sex and age (years)  No.  Mean†  (SD)‡  Percentile   90  75  50  25  10  Males                   ≤39  268  1.00  (6.4)  0.64  0.28  0.10  0.04  <0.03   40–49  477  0.63  (2.2)  0.70  0.31  0.14  0.03  <0.03   50–59  467  0.64  (3.5)  0.69  0.31  0.13  <0.03  <0.03   60–69  865  0.89  (3.1)  0.89  0.38  0.18  0.04  <0.03   ≥70  198  1.26  (3.6)  2.56  0.61  0.26  0.08  <0.03   Total  2,275  0.83  (3.6)  0.78  0.35  0.16  0.04  <0.03  Females                   ≤39  402  0.46**  (2.7)  0.35  0.17  0.05  <0.03  <0.03   40–49  739  0.34**  (1.7)  0.44  0.21  0.05  <0.03  <0.03   50–59  964  0.62  (2.9)  0.55  0.25  0.09  <0.03  <0.03   60–69  1,477  0.68*  (3.0)  0.74  0.33  0.14  0.04  <0.03   ≥70  250  0.87*  (3.0)  0.83  0.44  0.22  0.05  <0.03   Total  3,832  0.59**  (2.7)  0.57  0.28  0.09  <0.03  <0.03  * p < 0.05; ** p < 0.001, by Mann-Whitney U tests for male versus female. † Arithmetric mean. ‡ SD, standard deviation. View Large Atherosclerotic risk factors in individuals with low CRP and high CRP groups were compared separately in males and females (table 2). Age, systolic pressure, diastolic pressure, triglycerides, fibrinogen, body mass index, insulin, and waist-to-hip ratio were positively associated with CRP in both sexes. High density lipoprotein cholesterol was inversely associated with CRP levels in both sexes. Total cholesterol and blood glucose were correlated among females only, and smoking status was correlated among males only. No significant difference was detected with respect to lipoprotein(a). These associations were also seen after adjustment for age (data not shown). TABLE 2. Mean value of atherosclerotic risk factors stratified by C-reactive protein levels, the Jichi Medical School Cohort Study, Japan, 1992–1995   Male   p value    Low (≤0.11 mg/liter) (n = 992)   High (>0.11 mg/liter) (n = 1,283)     Mean  (SD)*  Mean  (SD)  Age (years)  54.1  (11.9)  56.9  (12.1)  <0.001  Systolic blood pressure (mmHg)  126.9  (20.0)  131.8  (21.8)  <0.001  Diastolic blood pressure (mmHg)  77.3  (12.0)  79.8  (13.0)  <0.001  Total cholesterol (mg/dl)  185.3  (32.4)  185.8  (35.0)  0.730  HDL* cholesterol (mg/dl)  51.8  (13.7)  47.4  (13.5)  <0.001  Fibrinogen (mg/dl)  228.7  (44.8)  257.7  (64.9)  <0.001  BMI* (kg/m2)  22.2  (2.5)  23.2  (3.0)  <0.001  Triglycerides (mg/dl)†  95.2  (55.3–164.0)  103.6  (57.6–186.4)  <0.001  Blood glucose (mg/dl)†  100.6  (82.1–123.2)  100.3  (82.3–122.2)  0.739  Lipoprotein(a) (mg/dl)†  11.8  (4.3–32.4)  11.7  (4.2–32.8)  0.755  Insulin (μU/ml) (n = 1,222)†  3.6  (2.0–6.4)  4.2  (2.1–8.2)  <0.001  Waist-to-hip ratio (n = 921)  0.88  (0.05)  0.89  (0.06)  <0.001  Smoker (%)  47.5     52.1     0.028    Female       Low (≤0.11 mg/liter) (n = 2,019)   High (>0.11 mg/liter) (n = 1,813)       Mean   (SD)   Mean   (SD)     Age (years)  54.0  (11.4)  57.8  (11.0)  <0.001  Systolic blood pressure (mmHg)  123.0  (20.8)  131.2  (22.1)  <0.001  Diastolic blood pressure (mmHg)  74.2  (12.3)  78.2  (12.7)  <0.001  Total cholesterol (mg/dl)  193.0  (34.1)  201.8  (35.4)  <0.001  HDL cholesterol (mg/dl)  54.7  (12.7)  50.8  (12.6)  <0.001  Fibrinogen (mg/dl)  235.3  (44.3)  267.6  (62.2)  <0.001  BMI (kg/m2)  22.3  (4.8)  23.7  (3.3)  <0.001  Triglycerides (mg/dl)†  80.8  (51.5–126.7)  98.8  (57.2–170.7)  <0.001  Blood glucose (mg/dl)†  95.0  (81.2–111.2)  97.4  (81.8–116.0)  <0.001  Lipoprotein(a) (mg/dl)†  13.6  (5.1–35.9)  14.1  (5.6–35.6)  0.209  Insulin (μU/ml) (n = 1,663)†  4.3  (2.5–7.6)  5.1  (2.8–9.3)  <0.001  Waist-to-hip ratio (n = 1,276)  0.81  (0.07)  0.85  (0.07)  <0.001  Smoker (%)  6.1    6.9    0.314    Male   p value    Low (≤0.11 mg/liter) (n = 992)   High (>0.11 mg/liter) (n = 1,283)     Mean  (SD)*  Mean  (SD)  Age (years)  54.1  (11.9)  56.9  (12.1)  <0.001  Systolic blood pressure (mmHg)  126.9  (20.0)  131.8  (21.8)  <0.001  Diastolic blood pressure (mmHg)  77.3  (12.0)  79.8  (13.0)  <0.001  Total cholesterol (mg/dl)  185.3  (32.4)  185.8  (35.0)  0.730  HDL* cholesterol (mg/dl)  51.8  (13.7)  47.4  (13.5)  <0.001  Fibrinogen (mg/dl)  228.7  (44.8)  257.7  (64.9)  <0.001  BMI* (kg/m2)  22.2  (2.5)  23.2  (3.0)  <0.001  Triglycerides (mg/dl)†  95.2  (55.3–164.0)  103.6  (57.6–186.4)  <0.001  Blood glucose (mg/dl)†  100.6  (82.1–123.2)  100.3  (82.3–122.2)  0.739  Lipoprotein(a) (mg/dl)†  11.8  (4.3–32.4)  11.7  (4.2–32.8)  0.755  Insulin (μU/ml) (n = 1,222)†  3.6  (2.0–6.4)  4.2  (2.1–8.2)  <0.001  Waist-to-hip ratio (n = 921)  0.88  (0.05)  0.89  (0.06)  <0.001  Smoker (%)  47.5     52.1     0.028    Female       Low (≤0.11 mg/liter) (n = 2,019)   High (>0.11 mg/liter) (n = 1,813)       Mean   (SD)   Mean   (SD)     Age (years)  54.0  (11.4)  57.8  (11.0)  <0.001  Systolic blood pressure (mmHg)  123.0  (20.8)  131.2  (22.1)  <0.001  Diastolic blood pressure (mmHg)  74.2  (12.3)  78.2  (12.7)  <0.001  Total cholesterol (mg/dl)  193.0  (34.1)  201.8  (35.4)  <0.001  HDL cholesterol (mg/dl)  54.7  (12.7)  50.8  (12.6)  <0.001  Fibrinogen (mg/dl)  235.3  (44.3)  267.6  (62.2)  <0.001  BMI (kg/m2)  22.3  (4.8)  23.7  (3.3)  <0.001  Triglycerides (mg/dl)†  80.8  (51.5–126.7)  98.8  (57.2–170.7)  <0.001  Blood glucose (mg/dl)†  95.0  (81.2–111.2)  97.4  (81.8–116.0)  <0.001  Lipoprotein(a) (mg/dl)†  13.6  (5.1–35.9)  14.1  (5.6–35.6)  0.209  Insulin (μU/ml) (n = 1,663)†  4.3  (2.5–7.6)  5.1  (2.8–9.3)  <0.001  Waist-to-hip ratio (n = 1,276)  0.81  (0.07)  0.85  (0.07)  <0.001  Smoker (%)  6.1    6.9    0.314  * SD, standard deviation; HDL, high density lipoprotein; BMI, body mass index. † Geometric mean (mean (SD)). View Large Logistic regression analysis was performed on variables that were significantly correlated with CRP (table 3). Male sex, age, systolic pressure, triglycerides, fibrinogen, body mass index, and current smoker status were positively correlated with CRP, while high density lipoprotein cholesterol was negatively correlated. When waist-to-hip ratio and insulin were included in the analysis (in a subset of 2,134 subjects), age, systolic pressure, fibrinogen, body mass index, current smoker status, and waist-to-hip ratio were positively associated with CRP. However, the associations with sex, high density lipoprotein cholesterol, and triglycerides disappeared. There was no significant association between insulin and CRP. As an additional observation, a multiple linear regression analysis, of which the dependent variable was natural log-transformed CRP, was conducted. There were 1,630 samples with CRP levels less than the lowest detection value (0.03 mg/liter); such samples were considered to be half of the lowest detection value (0.015 mg/liter) in this study. As shown in table 4, the associations of CRP and sex, age, systolic pressure, triglycerides, fibrinogen, body mass index, smoking status, and high density lipoprotein cholesterol were not basically different from the results from the logistic regression analysis, but a negative association between CRP and total cholesterol appeared. When waist-to-hip ratio and insulin were included in the analysis, age, systolic pressure, fibrinogen, body mass index, smoking status, and waist-to-hip ratio were positively associated with CRP, while total cholesterol and blood glucose were negatively associated. The associations with sex, high density lipoprotein cholesterol, and triglycerides disappeared. There was no significant association between insulin and CRP. TABLE 3. Multiple logistic regression analysis with C-reactive protein* as the dependent variable, the Jichi Medical School Cohort Study, Japan, 1992–1995 Variables  OR 1†,‡ (n = 5,903)  95% CI†  OR 2‡ (n = 2,134)  95% CI  Female (vs. male)  0.75  0.65, 0.86  0.89  0.69, 1.14  Age (years)  1.16  1.09, 1.23  1.17  1.05, 1.30  Systolic blood pressure (mmHg)  1.22  1.15, 1.30  1.14  1.03, 1.26  Total cholesterol (mg/dl)  1.03  0.96, 1.10  1.02  0.91, 1.15  HDL† cholesterol (mg/dl)  0.84  0.79, 0.90  0.94  0.83, 1.05  Triglycerides (mg/dl)  1.15  1.07, 1.23  1.08  0.95, 1.24  Blood glucose (mg/dl)  0.95  0.90, 1.01  0.92  0.80, 1.05  Fibrinogen (mg/dl)  1.83  1.71, 1.96  2.34  2.06, 2.66  BMI† (kg/m2)  1.52  1.40, 1.64  1.44  1.27, 1.64  Current smoker (vs. nonsmoker)  1.36  1.16, 1.60  1.41  1.08, 1.85  Insulin (μU/ml)      1.06  0.94, 1.19  Waist-to-hip ratio      1.29  1.13, 1.46  Variables  OR 1†,‡ (n = 5,903)  95% CI†  OR 2‡ (n = 2,134)  95% CI  Female (vs. male)  0.75  0.65, 0.86  0.89  0.69, 1.14  Age (years)  1.16  1.09, 1.23  1.17  1.05, 1.30  Systolic blood pressure (mmHg)  1.22  1.15, 1.30  1.14  1.03, 1.26  Total cholesterol (mg/dl)  1.03  0.96, 1.10  1.02  0.91, 1.15  HDL† cholesterol (mg/dl)  0.84  0.79, 0.90  0.94  0.83, 1.05  Triglycerides (mg/dl)  1.15  1.07, 1.23  1.08  0.95, 1.24  Blood glucose (mg/dl)  0.95  0.90, 1.01  0.92  0.80, 1.05  Fibrinogen (mg/dl)  1.83  1.71, 1.96  2.34  2.06, 2.66  BMI† (kg/m2)  1.52  1.40, 1.64  1.44  1.27, 1.64  Current smoker (vs. nonsmoker)  1.36  1.16, 1.60  1.41  1.08, 1.85  Insulin (μU/ml)      1.06  0.94, 1.19  Waist-to-hip ratio      1.29  1.13, 1.46  * A dichotomous datum C-reactive protein (CRP ≤ 0.11/CRP > 0.11) was the dependent variable. † OR, odds ratio; CI, confidence interval; HDL, high density lipoprotein; BMI, body mass index. ‡ Odds ratio 1 included all of the subjects in the study. Odds ratio 2 included some subjects in the study with data on waist-to-hip ratio and insulin. Odds ratio for 1 standard deviation increase, except for sex and smoking status. View Large TABLE 4. Multiple linear regression analysis with C-reactive protein* (log-transformed) as the dependent variable, the Jichi Medical School Cohort Study, Japan, 1992–1995 Variables  Coefficient 1† (n = 5,903)  95% CI‡  Coefficient 2§ (n = 2,134)  95% CI  Female (vs. male)  −0.237  −0.328, −0.146  −0.041  −0.194, 0.112  Age (years)  0.006  0.003, 0.010  0.006  0.001, 0.012  Systolic blood pressure (mmHg)  0.007  0.005, 0.009  0.004  0.001, 0.007  Total cholesterol (mg/dl)  −0.002  −0.003, −0.001  −0.003  −0.005, −0.001  HDL‡ cholesterol (mg/dl)  −0.009  −0.012, −0.006  −0.001  −0.007, 0.004  Triglycerides (mg/dl)  0.230  0.147, 0.313  0.047  −0.102, 0.196  Blood glucose (mg/dl)  −0.118  −0.331, 0.096  −0.432  −0.863, −0.001  Fibrinogen (mg/dl)  0.011  0.011, 0.012  0.013  0.012, 0.015  BMI‡ (kg/m2)  0.048  0.038, 0.059  0.071  0.046, 0.096  Current smoker (vs. nonsmoker)  0.151  0.045, 0.257  0.247  0.084, 0.410  Insulin (μU/ml)      0.091  −0.021, 0.204  Waist-to-hip ratio      2.623  1.583, 3,663  Variables  Coefficient 1† (n = 5,903)  95% CI‡  Coefficient 2§ (n = 2,134)  95% CI  Female (vs. male)  −0.237  −0.328, −0.146  −0.041  −0.194, 0.112  Age (years)  0.006  0.003, 0.010  0.006  0.001, 0.012  Systolic blood pressure (mmHg)  0.007  0.005, 0.009  0.004  0.001, 0.007  Total cholesterol (mg/dl)  −0.002  −0.003, −0.001  −0.003  −0.005, −0.001  HDL‡ cholesterol (mg/dl)  −0.009  −0.012, −0.006  −0.001  −0.007, 0.004  Triglycerides (mg/dl)  0.230  0.147, 0.313  0.047  −0.102, 0.196  Blood glucose (mg/dl)  −0.118  −0.331, 0.096  −0.432  −0.863, −0.001  Fibrinogen (mg/dl)  0.011  0.011, 0.012  0.013  0.012, 0.015  BMI‡ (kg/m2)  0.048  0.038, 0.059  0.071  0.046, 0.096  Current smoker (vs. nonsmoker)  0.151  0.045, 0.257  0.247  0.084, 0.410  Insulin (μU/ml)      0.091  −0.021, 0.204  Waist-to-hip ratio      2.623  1.583, 3,663  * A numerical datum of lognormal C-reactive protein (lnCRP) as the dependent variable. A CRP level of less than the lowest detection value (0.03 mg/liter) was taken to be the half value (0.015 mg/liter). † Coefficient 1 included all of the subjects in the study. ‡ CI, confidence interval; HDL, high density lipoprotein; BMI, body mass index. § Coefficient 2 included some subjects in the study with data on waist-to-hip ratio and insulin. View Large CRP levels were compared with menopausal status. Premenopausal women tended to have lower CRP levels than did postmenopausal women. The age-adjusted geometric means were 0.08 and 0.10 mg/liter, respectively (p = 0.10). DISCUSSION To the best of our knowledge, this is the first study to provide cross-sectional data in relation to atherosclerotic risk factors and serum CRP levels in a large-scale, population-based study of the Japanese. An additional strength of this study was the quality of the sample collection and the precision of the CRP measurements, in which CV levels were lower than those reported in previous reports (5–7). In previous studies, the distribution of serum CRP levels was skewed, with a log-normal distribution for males (16–18). Koenig et al. (17) reported that the distribution of CRP levels was 55–80 percent for CRP values of less than 2 mg/liter, which is well below the range seen in routine CRP measurements used to monitor active inflammatory, infective, or tissue-damaging disorders (2, 4, 5, 7). In addition, in other population-based studies that included both sexes, the distribution also showed log-normal distribution (8–10). In our study, however, the distribution of CRP was highly skewed to lower levels than in previous studies (table 5), with a median value of 0.12 mg/liter, and 94 percent with less than 2 mg/liter, and appeared to have a combination of different distributions. TABLE 5. Comparison of the C-reactive protein levels in the Jichi Medical School Cohort Study with previous studies Study  Age (years)  Sex  No. of subjects  Median (mg/liter)  Mean (mg/liter)  Our data  ≥30  Male  2,275  0.16  0.83      Female  3,832  0.09  0.59  Kuller et al. (4)  35–57  Male  256    2.9  Ridker et al. (5)  40–84  Male  543  1.13  1.1*  Ridker et al. (6)†    Female  244  3.75    Tracy et al. (7)  ≥65  Male  89    2.32      Female  57    1.73  Tracy et al. (10)‡  ≥65  Male  400    2.67      Female      2.66  Danesb et al. (8)  35–64  Both  704 (225 females)  1.6    Mendall et al. (16)  50–69  Male  303  1.72    Koenig et al. (17)  45–64  Male  936  1.584  1.623*  Rohde et al. (18)  40–84  Male  1,172  1.3  2.0  Ford (9)  ≥20  Male  7,325  2.1  3.65      Female  8,244  2.1  4.59  Study  Age (years)  Sex  No. of subjects  Median (mg/liter)  Mean (mg/liter)  Our data  ≥30  Male  2,275  0.16  0.83      Female  3,832  0.09  0.59  Kuller et al. (4)  35–57  Male  256    2.9  Ridker et al. (5)  40–84  Male  543  1.13  1.1*  Ridker et al. (6)†    Female  244  3.75    Tracy et al. (7)  ≥65  Male  89    2.32      Female  57    1.73  Tracy et al. (10)‡  ≥65  Male  400    2.67      Female      2.66  Danesb et al. (8)  35–64  Both  704 (225 females)  1.6    Mendall et al. (16)  50–69  Male  303  1.72    Koenig et al. (17)  45–64  Male  936  1.584  1.623*  Rohde et al. (18)  40–84  Male  1,172  1.3  2.0  Ford (9)  ≥20  Male  7,325  2.1  3.65      Female  8,244  2.1  4.59  * Geometric mean. † Age not given in the paper. ‡ Numbers of subjects not given in the paper. The blank column not described in the paper. View Large In previous studies (8, 10, 16, 18), the relations among serum CRP and atherosclerotic risk factors were inconsistent. Mendall et al. (16) and other investigators in the United States (10, 18) showed that CRP was related to several atherosclerotic risk factors, while Danesh et al. (8) found no association except for smoking status and body mass index. Our data showed results similar to those in the study by Mendall et al. and in other US studies, except for total cholesterol and blood glucose in males. Although the reasons for the discrepancies between our results and the report by Danesh et al. are unclear, some factors that may be involved are differences in CRP assays and ethnic differences on the clustering of risk factors. However, our results suggest that CRP may be a risk marker for atherosclerotic diseases in the Japanese, as well as in Westerners. In females, there was no difference between smokers and nonsmokers. Because there were few female smokers in this study, however, the effect of smoking among females could not be ascertained. Our results showed positive associations among CRP, body mass index, and waist-to-hip ratio, but not with insulin. Hak et al. (19) showed that CRP was associated with body mass index and waist and hip circumference, but not with waist-to-hip ratio after adjustment for body mass index in 186 healthy, middle-aged women. In our study, the association between CRP and waist circumference showed results similar to those with the waist-to-hip ratio (data not shown). All previous studies have shown a positive association between CRP and body mass index, and recently, an association among CRP, insulin, and obesity has been suggested (9, 19–22). The mechanisms for the reason CRP had a positive association with body mass index or obesity may be explained as follows. The main modulators of CRP are interleukin 1 and 6 and tumor necrosis factor alpha (23). Adipocytes from obese humans have been shown to overproduce tumor necrosis factor alpha messenger RNA (24). Tumor necrosis factor alpha is a potent inducer of interleukin 6 production in various cells, which may explain the positive association between CRP and body mass index (16). In vitro, human abdominal visceral adipose tissue releases more interleukin 6 compared with subcutaneous adipose tissue (25); our data may support this. Although our data are cross-sectional and cannot address causality or exclude confounders that can explain the results, they suggest that CRP is strongly associated with fat distribution. The sex difference in CRP is controversial. In our study, males had higher CRP levels than did females. This result was consistent with one nested case-control study of 89 males and 57 females aged 65 years and over (7). One cross-sectional study of 400 subjects, which included both sexes aged 65 years and over, showed that males and females had similar values (10), while other studies showed that females had higher CRP levels than did males (6, 9). The reasons for these discrepancies are unclear, but it is important to note that the females in this analysis had lower body mass index and lower rates of smoking, which are associated with decreased CRP, than those in previous studies. Although the sex difference in our analysis may reflect the fact that males were more likely than females to be current smokers, the difference remained even after adjustment for smoking status: This may reflect sex-specific effects. In our study, premenopausal women tended to have lower CRP levels than did postmenopausal women. Hak et al. (19) also showed that CRP levels were slightly higher in postmenopausal compared with premenopausal women, although this was not statistically significant. Although experimental data suggest an inhibitory effect of estrogens on interleukin 6 gene expression (19, 26), recent studies suggested an increase in CRP with hormone replacement therapy (27–31). However, in rural areas of Japan, hormone replacement therapy is not yet widely used, so we believe that the effect of this therapy was small in our analysis. We think, therefore, that sex hormonal effect may be one reason why females had lower CRP levels than did males. In multivariate analyses, which included the waist-to-hip ratio and insulin, the results were different from those excluding these parameters. Although this may indicate that the waist-to-hip ratio has a stronger association with CRP than other variables, it is important to note that the data on waist-to-hip ratio and insulin included only about one third of all the subjects in our study. Therefore, we believe that the results including these data need to be considered carefully and that further studies are needed. The much lower levels of CRP compared with previous studies may be due to the fact that our study population included females, who had very low smoking rates, and a lower body mass index. However, even taking these factors into consideration, the difference was considerable. This finding suggests, therefore, that the Japanese have lower CRP levels than do Westerners, which may be reflected in the lower coronary event rate among the Japanese. In this study, CRP levels seemed to show a combination of two separate distributions. This finding cannot be explained, but it may indicate that the higher CRP group underwent subclinical atherosclerotic changes, possibly resulting in future atherosclerotic events. In conclusion, serum CRP levels correlated with atherosclerotic risk factors in the Japanese as well as in Westerners. However, the distribution of CRP in this study was quite different from that in previous studies, suggesting that the Japanese have much lower CRP levels than do those of other developed countries and that this could reflect the low coronary event rate among the Japanese. Reprint requests to Dr. Seishi Yamada, Department of Internal Medicine, Wara National Health Insurance Hospital, 882 Sawa Wara, Gujyo, Gifu 501–4595, Japan (e-mail: [email protected]). Supported in part by grants from the Foundation for the Department of the Community, Tochigi, Japan. REFERENCES 1. Ross R. The pathogenesis of atherosclerosis: a perspective for the 1990s. Nature  1993; 362: 802–9. Google Scholar 2. Haverkate F, Thompson SG, Pyke SD, et al. Production of C-reactive protein and risk of coronary events in stable and unstable angina. European Concerted Action on Thrombosis and Disabilities Angina Pectoris Study Group. Lancet  1997; 349: 462–6. Google Scholar 3. Liuzzo G, Biasucci LM, Gallimore JR, et al. The prognostic value of C-reactive protein and serum amyloid A protein in severe unstable angina. N Engl J Med  1994; 331: 417–24. Google Scholar 4. Kuller LH, Tracy RP, Shaten J, et al. Relation of C-reactive protein and coronary heart disease in the MRFIT nested case-control study. Multiple Risk Factor Intervention Trial. Am J Epidemiol  1996; 144: 537–47. Google Scholar 5. Ridker PM, Cushman M, Stampfer MJ, et al. Inflammation, aspirin, and the risk of cardiovascular disease in apparently healthy men. N Engl J Med  1997; 336: 973–9. Google Scholar 6. Ridker PM, Buring JE, Shih J, et al. Prospective study of C-reactive protein and the risk of future cardiovascular events among apparently healthy women. Circulation  1998; 98: 731–3. Google Scholar 7. Tracy RP, Lemaitre RN, Psaty BM, et al. Relationship of C-reactive protein to risk of cardiovascular disease in the elderly. Results from the Cardiovascular Health Study and the Rural Health Promotion Project. Arterioscler Thromb Vasc Biol  1997; 17: 1121–7. Google Scholar 8. Danesh J, Muir J, Wong YK, et al. Risk factors for coronary heart disease and acute-phase proteins. A population-based study. Eur Heart J  1999; 20: 954–9. Google Scholar 9. Ford ES. Body mass index, diabetes, and C-reactive protein among U.S. adults. Diabetes Care  1999; 22: 1971–7. Google Scholar 10. Tracy RP, Psaty BM, Macy E, et al. Lifetime smoking exposure affects the association of C-reactive protein with cardiovascular disease risk factors and subclinical disease in healthy elderly subjects. Arterioscler Thromb Vasc Biol  1997; 17: 2167–76. Google Scholar 11. Nago N, Kayaba K, Hiraoka J, et al. Lipoprotein(a) levels in the Japanese population: influence of age and sex, and relation to atherosclerotic risk factors: the Jichi Medical School Cohort Study. Am J Epidemiol  1995; 141: 815–21. Google Scholar 12. Gotoh T, Kuroda T, Yamasawa M, et al. Correlation between lipoprotein(a) and aortic valve sclerosis assessed by echocardiography (the JMS Cardiac Echo and Cohort Study). Am J Cardiol  1995; 76: 928–32. Google Scholar 13. Kario K, Nago N, Kayaba K, et al. Characteristics of the insulin resistance syndrome in a Japanese population: The Jichi Medical School Cohort Study. Arterioscler Thromb Vasc Biol  1996; 16: 269–74. Google Scholar 14. Schreiner PJ, Terry JG, Evans GW, et al. Sex-specific associations of magnetic resonance imaging-derived intra-abdominal and subcutaneous fat areas with conventional indices. The Atherosclerosis Risk in Communities Study. Am J Epidemiol  1996; 144: 335–45. Google Scholar 15. Itoh Y, Ichihara K, Kanno T, et al. Serum protein standardization project in Japan: evaluation of an IFCC reference material (RPPHS/CRM470) and establishment of reference intervals. J Clin Lab Anal  1997; 11: 39–44. Google Scholar 16. Mendall MA, Patel P, Ballam L, et al. C reactive protein and its relation to cardiovascular risk factors: a population based cross sectional study. BMJ  1996; 312: 1061–5. Google Scholar 17. Koenig W, Sund M, Frohlich M, et al. C-Reactive protein, a sensitive marker of inflammation, predicts future risk of coronary heart disease in initially healthy middle-aged men: results from the MONICA (Monitoring Trends and Determinants in Cardiovascular Disease) Augsburg Cohort Study, 1984 to 1992. Circulation  1999; 99: 237–42. Google Scholar 18. Rohde LEP, Hennekens CH, Ridker PM. Survey of C-reactive protein and cardiovascular risk factors in apparently healthy men. Am J Cardiol  1999; 84: 1018–22. Google Scholar 19. Hak AE, Stehouwer CD, Bots ML, et al. Associations of C-reactive protein with measures of obesity, insulin resistance, and subclinical atherosclerosis in healthy, middle-aged women. Arterioscler Thromb Vasc Biol  1999; 19: 1986–91. Google Scholar 20. Strandberg TE, Tilvis RS. C-reactive protein, cardiovascular risk factors, and prospective study in the elderly. Arterioscler Thromb Vasc Biol  2000; 20: 1057–60. Google Scholar 21. Yudkin JS, Stehouwer CD, Emeis JJ, et al. C-reactive protein in healthy subjects: associations with insulin resistance, and endothelial dysfunction: a potential cytokines originating from adipose tissue? Arterioscler Thromb Vasc Biol  1999; 19: 972–8. Google Scholar 22. Visser M, Bouter LM, McQuillan GM, et al. Elevated C-reactive protein levels in overweight and obese. JAMA  1999; 282: 2131–5. Google Scholar 23. Danesh J, Collins R, Appleby P, et al. Association of fibrinogen, C-reactive protein, albumin, or leukocyte count with coronary heart disease: meta-analyses of prospective studies. JAMA  1998; 279: 1477–82. Google Scholar 24. Hotamisligil GS, Arner P, Caro JF, et al. Increased adipose tissue expression of tumor necrosis factor-alpha in human obesity and insulin resistance. J Clin Invest  1995; 95: 2409–15. Google Scholar 25. Fried SK, Bunkin DA, Greenberg AS. Omental and subcutaneous adipose tissues of obese subjects release interleukin-6: depot difference and regulation by glucocorticoid. J Clin Endocrinol Metab  1998; 83: 847–50. Google Scholar 26. Ray P, Ghosh SK, Zhang DH, et al. Repression of interleukin-6 gene expression by 17 inhibition of the DNA-binding activity of the transcription NF-IL6 and NF-kappa B by the estrogen receptor. FEBS Lett  1997; 409: 79–85. Google Scholar 27. Walsh BW, Paul S, Wild RA, et al. The effects of hormone replacement therapy and raloxifene on C-reactive protein and homocysteine in healthy postmenopausal women: a randomized, controlled trial. J Clin Endocrinol Metab  2000; 85: 214–18. Google Scholar 28. van Baal WM, Kenemans P, van der Mooren MJ, et al. Increased C-reactive protein levels during short-term hormone replacement therapy in healthy postmenopausal women. Thromb Haemost  1999; 81: 925–8. Google Scholar 29. Ridker PM, Hennekens CH, Rifai N, et al. Hormone replacement therapy and increased plasma concentration of C- reactive protein. Circulation  1999; 100: 713–16. Google Scholar 30. Cushman M, Meilahn EN, Psaty BM, et al. Hormone replacement therapy, inflammation, and hemostasis in women. Arterioscler Thromb Vasc Biol  1999; 19: 893–9. Google Scholar 31. Cushman M, Legault C, Barrett-Connor E, et al. Effect of postmenopausal hormones on inflammation-sensitive proteins: the Postmenopausal Estrogen/Progestin Interventions (PEPI) Study. Circulation  1999; 100: 717–22. Google Scholar http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png American Journal of Epidemiology Oxford University Press

Distribution of Serum C-Reactive Protein and Its Association with Atherosclerotic Risk Factors in a Japanese Population : Jichi Medical School Cohort Study

Download PDF
8 pages

Loading next page...
 
/lp/oxford-university-press/distribution-of-serum-c-reactive-protein-and-its-association-with-6Dq0KUBgU6

References (30)

Publisher
Oxford University Press
ISSN
0002-9262
eISSN
1476-6256
DOI
10.1093/aje/153.12.1183
Publisher site
See Article on Publisher Site

Abstract

Abstract The distribution of serum C-reactive protein (CRP) levels and their association with age, sex, and atherosclerotic risk factors were studied in a large Japanese population between 1992 and 1995. The subjects consisted of 2,275 males and 3,832 females aged 30 years and over. CRP was measured by nephelometry. The distribution of CRP was highly skewed toward a lower level than that of previous studies and seemed to be a combination of two separate distribution curves. The increase in CRP with age was statistically significant, and males had higher CRP levels than did females. Males who were current smokers had higher CRP levels than did nonsmokers. Age, systolic blood pressure, diastolic blood pressure, triglycerides, fibrinogen, and body mass index were all positively associated with CRP in both sexes, while total cholesterol and blood glucose were positively related in females only. High density lipoprotein cholesterol was inversely related in both sexes. Multiple logistic regression analysis showed that sex, age, systolic pressure, high density lipoprotein cholesterol, triglycerides, fibrinogen, and body mass index were significant independent variables. In conclusion, the distribution of CRP among the Japanese was quite different from that among Westerners, although CRP levels correlated with other atherosclerotic risk factors, similar to those in Westerners. C-reactive protein, cross-sectional studies, risk factors CRP, C-reactive protein, CV, coefficient of variation Although atherosclerosis is multifactorial, it is now widely recognized that systemic and localized inflammation in the arteries contributes to the initiation or promotion of cardiovascular disease (1). C-reactive protein (CRP) is a protein of the acute-phase reaction of inflammation. It has been shown that elevated levels of serum CRP are related to a poorer outcome in patients with cardiovascular diseases (2, 3). Recently, several studies have suggested that CRP is a predictor of subsequent cardiovascular events in healthy adults (4–7). However, to our knowledge, there have been only two reports on the distribution of serum CRP levels in a large-scale, population-based study that included both males and females and used a hypersensitive assay (8, 9). Moreover, the relation between CRP and atherosclerotic risk factors in females has been reported in only a few studies (6–8, 10). The Jichi Medical School Cohort Study is a prospective, population-based study that aims to explore the risk factors for cardiovascular and cerebrovascular diseases in Japan. The subjects included people of both sexes aged 30 years and over residing in areas throughout Japan. In this report, we show the distribution of serum CRP levels and the relation between serum CRP levels and atherosclerotic risk factors in a large group of healthy people aged 30 years and over. MATERIALS AND METHODS The study design and some descriptive data have been presented previously (11–13). The population for our study included residents of nine rural communities in Japan, including Yamato, Takasu, Wara, Sakuma, Hokutan, Sakugi, Ohkawa, Ainoshima, and Akaike. In Japan, mass screening for cardiocerebrovascular diseases has been conducted since 1982, according to the Health and Medical Service Law for the Aged Act of 1981. Invitations to this mass screening were issued by government offices in each community, and personal invitations were also sent to all the subjects by mail. However, the invitations mentioned that those who were receiving treatment at hospitals or clinics for cardiovascular or cerebrovascular diseases did not need to take the examination. As a result, 2,573 males and 4,186 females aged 30 years and over participated in these examinations. In this study, we excluded those who had had stroke or myocardial infarction as determined by a questionnaire or interview and those from whom some data were incomplete. The final number of study subjects was therefore 2,275 males and 3,832 females. The overall response rate was 52.4 percent; the rates in each community were 51.0 percent in Yamato, 63.0 percent in Takasu, 90.0 percent in Wara, 88.0 percent in Sakuma, 27.0 percent in Hokutan, 38.0 percent in Sakugi, 66.0 percent in Ohkawa, 44.0 percent in Ainoshima, and 26.0 percent in Akaike. To obtain uniform information, we established a central committee composed of the chief medical officers from all participating areas, which developed a detailed manual for data collection. Medical history, habitual food intake, smoking habits, menopausal status, and alcohol consumption were assessed by a questionnaire developed by the committee. Females were considered postmenopausal if their menses had ceased naturally at least 12 months previously. Body height was measured in stocking feet. Body weight was recorded with the subject clothed, and 0.5 kg in summer or 1 kg in other seasons was subtracted from the recorded weight. The body mass index was calculated as weight (kg)/height (m)2. Waist and hip circumferences were collected as optional measurements in Takasu, Wara, Sakuma, Ohkawa, and Ainoshima. There were 921 male and 1,276 female subjects. Waist circumference was measured at the level of the high point of the iliac crest, and hip circumference was measured at the level of maximum extension of the buttocks. The waist-to-hip ratio, calculated as waist circumference divided by hip circumference, was used as an indicator of abdominal visceral fat (14). The systolic and diastolic blood pressures were measured with a fully automated sphygmomanometer (BP203RV-II, Nippon Colin, Komaki, Japan), placed on the right arm of a seated subject who had rested in a sitting position for at least 5 minutes before the measurement. Blood samples were drawn from the antecubital vein of seated subjects with minimal tourniquet use. Specimens were collected in siliconized vacuum glass tubes containing a 1/10 volume of 3.8 percent trisodium citrate for fibrinogen, sodium fluoride for blood glucose, and no additives for lipids, respectively. Tubes were centrifuged at 3,000 g for 15 minutes at room temperature. After separation, the serum samples were stored at 4°C in refrigerated containers if analysis was to be performed within a few days. Otherwise, the samples were frozen until analysis. Plasma samples were frozen as rapidly as possible to −80°C for storage until laboratory examination could be performed. CRP levels were measured by using nephelometry, a latex particle-enhanced immunoassay (NA Latex CRP Kit, Dade Behring, Tokyo, Japan). The value in the calibrator was assigned from the Certified Reference Material 470 (IRMM, Geel, Belgium), an international plasma protein reference material. The material has achieved international standardization in the assay of CRP. The function of the assay was found to be satisfactory (15). Its interassay and intraassay coefficients of variation (CV) were 1.18 and 1.36 percent, respectively. The assay is sensitive enough to detect 0.03 mg/liter of CRP. Undetectable CRP values were recorded as 0.015 mg/liter. Total cholesterol and triglycerides levels were measured by using an enzymatic method (Wako, Osaka, Japan; interassay CV: 1.5 percent for total cholesterol and 1.7 percent for triglycerides). High density lipoprotein cholesterol was measured by using the phosphotungstate precipitation method (Wako, Osaka, Japan; interassay CV: 1.8 percent). Blood glucose was measured by using an enzymatic method (Kanto Chemistry, Tokyo, Japan; interassay CV: 1.9 percent). Lipoprotein(a) levels were measured with an enzyme-linked immunosorbent assay kit (Biopool, Uppsala, Sweden; interassay CV: 3.51 percent). Fibrinogen levels were determined with a one-stage clotting assay kit (Data-Fi, Dade Behring, Miami, Florida; interassay CV: 2.5 percent). Serum insulin levels, which were collected in an optional examination that included 1,222 males and 1,663 females in Takasu, Wara, and Sakuma, were determined with a radioimmunoassay kit (Dainabot, Tokyo, Japan; interassay CV: 4.5 percent). The lower detection limit was 2.5 μU/ml, and insulin levels below this limit were taken as 2.0 μU/ml. Statistical methods Statistical analysis was performed using a Statistical Analysis System 6.12 edition (SAS Institute, Inc., Cary, North Carolina). Descriptive parameters were shown as the mean, standard deviation, and percentiles. For a comparison of the mean values, the unpaired t test, Mann-Whitney U test, or analysis of covariance with an adjustment for age was used. Categorical variables were analyzed by using the chi-square test. Serum CRP levels were divided into two categories, above and below 0.11 mg/liter, because the distribution appeared to be a combination of two separate distributions, with 0.11 mg/liter being the value at which these two distribution curves crossed. To assess the influence of each variable, logistic unconditioned regression models were used. Odds ratios were used to evaluate the association between CRP and other variables. As an additional observation, a multiple linear regression analysis was used in which the dependent variable was natural log-transformed CRP. In these multivariate analyses, we used two models, one that included and one that excluded the waist-to-hip ratio and insulin. Because these variables were measured as optional measurements in this study, they were taken for about one third of all the subjects in our study. The effect of menopausal status on CRP levels was analyzed in females aged 40–59 years, and those who had surgical menopause were excluded. The difference in CRP between premenopausal and postmenopausal women adjusted for age was studied with analysis of covariance. When parametric procedures were used, triglycerides, blood glucose, lipoprotein(a), and insulin were transformed into natural logarithms. A significant difference was defined as p < 0.05. RESULTS Figure 1 shows the distribution of CRP levels among the study subjects. The distribution of serum CRP levels was highly skewed to lower levels and ranged widely. The minimum value was less than 0.03 mg/liter, and the maximum value was 68.2 mg/liter, with 25th, 50th, and 75th percentile values of less than 0.03, 0.12, and 0.30 mg/liter, respectively. The shape of the distribution did not differ by sex, age, or community. In this study, serum CRP levels seemed to have two distributions. One was a lower distribution with a narrow range, and the other was a higher distribution with a very wide range. We considered this to be a combination of two separate distributions and divided CRP levels into two categories: those lower than or equal to 0.11 mg/liter and those higher than 0.11 mg/liter. This was the value at which the two distribution curves were seen to cross. The mean value and standard deviation of the first subpopulation, which had lower CRP levels, were 0.038 and 0.024 mg/liter, respectively. Those of the second population, which had higher CRP levels, were 1.30 and 4.25 mg/liter, respectively. FIGURE 1. View largeDownload slide Distribution of C-reactive protein levels (mg/liter), by age, for 2,275 males and 3,832 females, the Jichi Medical School Cohort Study, Japan, 1992–1995. Because the distribution of C-reactive protein ranged very widely, only those with levels lower than 1.2 mg/liter, which accounts for 92.3 percent of the males and 94.5 percent of the females, are included. FIGURE 1. View largeDownload slide Distribution of C-reactive protein levels (mg/liter), by age, for 2,275 males and 3,832 females, the Jichi Medical School Cohort Study, Japan, 1992–1995. Because the distribution of C-reactive protein ranged very widely, only those with levels lower than 1.2 mg/liter, which accounts for 92.3 percent of the males and 94.5 percent of the females, are included. Table 1 shows the mean CRP levels by age and sex. The mean CRP levels increased with age in both sexes, except for subjects aged under age 39 years. Males had higher CRP levels than did females in all age groups, with mean values of 0.83 and 0.59 mg/liter, respectively. This difference was statistically significant, except for subjects in their fifties. TABLE 1. C-Reactive protein levels (mg/liter) stratified by age and sex, for 2,275 males and 3,832 females, the Jichi Medical School Cohort Study, Japan, 1992–1995 Sex and age (years)  No.  Mean†  (SD)‡  Percentile   90  75  50  25  10  Males                   ≤39  268  1.00  (6.4)  0.64  0.28  0.10  0.04  <0.03   40–49  477  0.63  (2.2)  0.70  0.31  0.14  0.03  <0.03   50–59  467  0.64  (3.5)  0.69  0.31  0.13  <0.03  <0.03   60–69  865  0.89  (3.1)  0.89  0.38  0.18  0.04  <0.03   ≥70  198  1.26  (3.6)  2.56  0.61  0.26  0.08  <0.03   Total  2,275  0.83  (3.6)  0.78  0.35  0.16  0.04  <0.03  Females                   ≤39  402  0.46**  (2.7)  0.35  0.17  0.05  <0.03  <0.03   40–49  739  0.34**  (1.7)  0.44  0.21  0.05  <0.03  <0.03   50–59  964  0.62  (2.9)  0.55  0.25  0.09  <0.03  <0.03   60–69  1,477  0.68*  (3.0)  0.74  0.33  0.14  0.04  <0.03   ≥70  250  0.87*  (3.0)  0.83  0.44  0.22  0.05  <0.03   Total  3,832  0.59**  (2.7)  0.57  0.28  0.09  <0.03  <0.03  Sex and age (years)  No.  Mean†  (SD)‡  Percentile   90  75  50  25  10  Males                   ≤39  268  1.00  (6.4)  0.64  0.28  0.10  0.04  <0.03   40–49  477  0.63  (2.2)  0.70  0.31  0.14  0.03  <0.03   50–59  467  0.64  (3.5)  0.69  0.31  0.13  <0.03  <0.03   60–69  865  0.89  (3.1)  0.89  0.38  0.18  0.04  <0.03   ≥70  198  1.26  (3.6)  2.56  0.61  0.26  0.08  <0.03   Total  2,275  0.83  (3.6)  0.78  0.35  0.16  0.04  <0.03  Females                   ≤39  402  0.46**  (2.7)  0.35  0.17  0.05  <0.03  <0.03   40–49  739  0.34**  (1.7)  0.44  0.21  0.05  <0.03  <0.03   50–59  964  0.62  (2.9)  0.55  0.25  0.09  <0.03  <0.03   60–69  1,477  0.68*  (3.0)  0.74  0.33  0.14  0.04  <0.03   ≥70  250  0.87*  (3.0)  0.83  0.44  0.22  0.05  <0.03   Total  3,832  0.59**  (2.7)  0.57  0.28  0.09  <0.03  <0.03  * p < 0.05; ** p < 0.001, by Mann-Whitney U tests for male versus female. † Arithmetric mean. ‡ SD, standard deviation. View Large Atherosclerotic risk factors in individuals with low CRP and high CRP groups were compared separately in males and females (table 2). Age, systolic pressure, diastolic pressure, triglycerides, fibrinogen, body mass index, insulin, and waist-to-hip ratio were positively associated with CRP in both sexes. High density lipoprotein cholesterol was inversely associated with CRP levels in both sexes. Total cholesterol and blood glucose were correlated among females only, and smoking status was correlated among males only. No significant difference was detected with respect to lipoprotein(a). These associations were also seen after adjustment for age (data not shown). TABLE 2. Mean value of atherosclerotic risk factors stratified by C-reactive protein levels, the Jichi Medical School Cohort Study, Japan, 1992–1995   Male   p value    Low (≤0.11 mg/liter) (n = 992)   High (>0.11 mg/liter) (n = 1,283)     Mean  (SD)*  Mean  (SD)  Age (years)  54.1  (11.9)  56.9  (12.1)  <0.001  Systolic blood pressure (mmHg)  126.9  (20.0)  131.8  (21.8)  <0.001  Diastolic blood pressure (mmHg)  77.3  (12.0)  79.8  (13.0)  <0.001  Total cholesterol (mg/dl)  185.3  (32.4)  185.8  (35.0)  0.730  HDL* cholesterol (mg/dl)  51.8  (13.7)  47.4  (13.5)  <0.001  Fibrinogen (mg/dl)  228.7  (44.8)  257.7  (64.9)  <0.001  BMI* (kg/m2)  22.2  (2.5)  23.2  (3.0)  <0.001  Triglycerides (mg/dl)†  95.2  (55.3–164.0)  103.6  (57.6–186.4)  <0.001  Blood glucose (mg/dl)†  100.6  (82.1–123.2)  100.3  (82.3–122.2)  0.739  Lipoprotein(a) (mg/dl)†  11.8  (4.3–32.4)  11.7  (4.2–32.8)  0.755  Insulin (μU/ml) (n = 1,222)†  3.6  (2.0–6.4)  4.2  (2.1–8.2)  <0.001  Waist-to-hip ratio (n = 921)  0.88  (0.05)  0.89  (0.06)  <0.001  Smoker (%)  47.5     52.1     0.028    Female       Low (≤0.11 mg/liter) (n = 2,019)   High (>0.11 mg/liter) (n = 1,813)       Mean   (SD)   Mean   (SD)     Age (years)  54.0  (11.4)  57.8  (11.0)  <0.001  Systolic blood pressure (mmHg)  123.0  (20.8)  131.2  (22.1)  <0.001  Diastolic blood pressure (mmHg)  74.2  (12.3)  78.2  (12.7)  <0.001  Total cholesterol (mg/dl)  193.0  (34.1)  201.8  (35.4)  <0.001  HDL cholesterol (mg/dl)  54.7  (12.7)  50.8  (12.6)  <0.001  Fibrinogen (mg/dl)  235.3  (44.3)  267.6  (62.2)  <0.001  BMI (kg/m2)  22.3  (4.8)  23.7  (3.3)  <0.001  Triglycerides (mg/dl)†  80.8  (51.5–126.7)  98.8  (57.2–170.7)  <0.001  Blood glucose (mg/dl)†  95.0  (81.2–111.2)  97.4  (81.8–116.0)  <0.001  Lipoprotein(a) (mg/dl)†  13.6  (5.1–35.9)  14.1  (5.6–35.6)  0.209  Insulin (μU/ml) (n = 1,663)†  4.3  (2.5–7.6)  5.1  (2.8–9.3)  <0.001  Waist-to-hip ratio (n = 1,276)  0.81  (0.07)  0.85  (0.07)  <0.001  Smoker (%)  6.1    6.9    0.314    Male   p value    Low (≤0.11 mg/liter) (n = 992)   High (>0.11 mg/liter) (n = 1,283)     Mean  (SD)*  Mean  (SD)  Age (years)  54.1  (11.9)  56.9  (12.1)  <0.001  Systolic blood pressure (mmHg)  126.9  (20.0)  131.8  (21.8)  <0.001  Diastolic blood pressure (mmHg)  77.3  (12.0)  79.8  (13.0)  <0.001  Total cholesterol (mg/dl)  185.3  (32.4)  185.8  (35.0)  0.730  HDL* cholesterol (mg/dl)  51.8  (13.7)  47.4  (13.5)  <0.001  Fibrinogen (mg/dl)  228.7  (44.8)  257.7  (64.9)  <0.001  BMI* (kg/m2)  22.2  (2.5)  23.2  (3.0)  <0.001  Triglycerides (mg/dl)†  95.2  (55.3–164.0)  103.6  (57.6–186.4)  <0.001  Blood glucose (mg/dl)†  100.6  (82.1–123.2)  100.3  (82.3–122.2)  0.739  Lipoprotein(a) (mg/dl)†  11.8  (4.3–32.4)  11.7  (4.2–32.8)  0.755  Insulin (μU/ml) (n = 1,222)†  3.6  (2.0–6.4)  4.2  (2.1–8.2)  <0.001  Waist-to-hip ratio (n = 921)  0.88  (0.05)  0.89  (0.06)  <0.001  Smoker (%)  47.5     52.1     0.028    Female       Low (≤0.11 mg/liter) (n = 2,019)   High (>0.11 mg/liter) (n = 1,813)       Mean   (SD)   Mean   (SD)     Age (years)  54.0  (11.4)  57.8  (11.0)  <0.001  Systolic blood pressure (mmHg)  123.0  (20.8)  131.2  (22.1)  <0.001  Diastolic blood pressure (mmHg)  74.2  (12.3)  78.2  (12.7)  <0.001  Total cholesterol (mg/dl)  193.0  (34.1)  201.8  (35.4)  <0.001  HDL cholesterol (mg/dl)  54.7  (12.7)  50.8  (12.6)  <0.001  Fibrinogen (mg/dl)  235.3  (44.3)  267.6  (62.2)  <0.001  BMI (kg/m2)  22.3  (4.8)  23.7  (3.3)  <0.001  Triglycerides (mg/dl)†  80.8  (51.5–126.7)  98.8  (57.2–170.7)  <0.001  Blood glucose (mg/dl)†  95.0  (81.2–111.2)  97.4  (81.8–116.0)  <0.001  Lipoprotein(a) (mg/dl)†  13.6  (5.1–35.9)  14.1  (5.6–35.6)  0.209  Insulin (μU/ml) (n = 1,663)†  4.3  (2.5–7.6)  5.1  (2.8–9.3)  <0.001  Waist-to-hip ratio (n = 1,276)  0.81  (0.07)  0.85  (0.07)  <0.001  Smoker (%)  6.1    6.9    0.314  * SD, standard deviation; HDL, high density lipoprotein; BMI, body mass index. † Geometric mean (mean (SD)). View Large Logistic regression analysis was performed on variables that were significantly correlated with CRP (table 3). Male sex, age, systolic pressure, triglycerides, fibrinogen, body mass index, and current smoker status were positively correlated with CRP, while high density lipoprotein cholesterol was negatively correlated. When waist-to-hip ratio and insulin were included in the analysis (in a subset of 2,134 subjects), age, systolic pressure, fibrinogen, body mass index, current smoker status, and waist-to-hip ratio were positively associated with CRP. However, the associations with sex, high density lipoprotein cholesterol, and triglycerides disappeared. There was no significant association between insulin and CRP. As an additional observation, a multiple linear regression analysis, of which the dependent variable was natural log-transformed CRP, was conducted. There were 1,630 samples with CRP levels less than the lowest detection value (0.03 mg/liter); such samples were considered to be half of the lowest detection value (0.015 mg/liter) in this study. As shown in table 4, the associations of CRP and sex, age, systolic pressure, triglycerides, fibrinogen, body mass index, smoking status, and high density lipoprotein cholesterol were not basically different from the results from the logistic regression analysis, but a negative association between CRP and total cholesterol appeared. When waist-to-hip ratio and insulin were included in the analysis, age, systolic pressure, fibrinogen, body mass index, smoking status, and waist-to-hip ratio were positively associated with CRP, while total cholesterol and blood glucose were negatively associated. The associations with sex, high density lipoprotein cholesterol, and triglycerides disappeared. There was no significant association between insulin and CRP. TABLE 3. Multiple logistic regression analysis with C-reactive protein* as the dependent variable, the Jichi Medical School Cohort Study, Japan, 1992–1995 Variables  OR 1†,‡ (n = 5,903)  95% CI†  OR 2‡ (n = 2,134)  95% CI  Female (vs. male)  0.75  0.65, 0.86  0.89  0.69, 1.14  Age (years)  1.16  1.09, 1.23  1.17  1.05, 1.30  Systolic blood pressure (mmHg)  1.22  1.15, 1.30  1.14  1.03, 1.26  Total cholesterol (mg/dl)  1.03  0.96, 1.10  1.02  0.91, 1.15  HDL† cholesterol (mg/dl)  0.84  0.79, 0.90  0.94  0.83, 1.05  Triglycerides (mg/dl)  1.15  1.07, 1.23  1.08  0.95, 1.24  Blood glucose (mg/dl)  0.95  0.90, 1.01  0.92  0.80, 1.05  Fibrinogen (mg/dl)  1.83  1.71, 1.96  2.34  2.06, 2.66  BMI† (kg/m2)  1.52  1.40, 1.64  1.44  1.27, 1.64  Current smoker (vs. nonsmoker)  1.36  1.16, 1.60  1.41  1.08, 1.85  Insulin (μU/ml)      1.06  0.94, 1.19  Waist-to-hip ratio      1.29  1.13, 1.46  Variables  OR 1†,‡ (n = 5,903)  95% CI†  OR 2‡ (n = 2,134)  95% CI  Female (vs. male)  0.75  0.65, 0.86  0.89  0.69, 1.14  Age (years)  1.16  1.09, 1.23  1.17  1.05, 1.30  Systolic blood pressure (mmHg)  1.22  1.15, 1.30  1.14  1.03, 1.26  Total cholesterol (mg/dl)  1.03  0.96, 1.10  1.02  0.91, 1.15  HDL† cholesterol (mg/dl)  0.84  0.79, 0.90  0.94  0.83, 1.05  Triglycerides (mg/dl)  1.15  1.07, 1.23  1.08  0.95, 1.24  Blood glucose (mg/dl)  0.95  0.90, 1.01  0.92  0.80, 1.05  Fibrinogen (mg/dl)  1.83  1.71, 1.96  2.34  2.06, 2.66  BMI† (kg/m2)  1.52  1.40, 1.64  1.44  1.27, 1.64  Current smoker (vs. nonsmoker)  1.36  1.16, 1.60  1.41  1.08, 1.85  Insulin (μU/ml)      1.06  0.94, 1.19  Waist-to-hip ratio      1.29  1.13, 1.46  * A dichotomous datum C-reactive protein (CRP ≤ 0.11/CRP > 0.11) was the dependent variable. † OR, odds ratio; CI, confidence interval; HDL, high density lipoprotein; BMI, body mass index. ‡ Odds ratio 1 included all of the subjects in the study. Odds ratio 2 included some subjects in the study with data on waist-to-hip ratio and insulin. Odds ratio for 1 standard deviation increase, except for sex and smoking status. View Large TABLE 4. Multiple linear regression analysis with C-reactive protein* (log-transformed) as the dependent variable, the Jichi Medical School Cohort Study, Japan, 1992–1995 Variables  Coefficient 1† (n = 5,903)  95% CI‡  Coefficient 2§ (n = 2,134)  95% CI  Female (vs. male)  −0.237  −0.328, −0.146  −0.041  −0.194, 0.112  Age (years)  0.006  0.003, 0.010  0.006  0.001, 0.012  Systolic blood pressure (mmHg)  0.007  0.005, 0.009  0.004  0.001, 0.007  Total cholesterol (mg/dl)  −0.002  −0.003, −0.001  −0.003  −0.005, −0.001  HDL‡ cholesterol (mg/dl)  −0.009  −0.012, −0.006  −0.001  −0.007, 0.004  Triglycerides (mg/dl)  0.230  0.147, 0.313  0.047  −0.102, 0.196  Blood glucose (mg/dl)  −0.118  −0.331, 0.096  −0.432  −0.863, −0.001  Fibrinogen (mg/dl)  0.011  0.011, 0.012  0.013  0.012, 0.015  BMI‡ (kg/m2)  0.048  0.038, 0.059  0.071  0.046, 0.096  Current smoker (vs. nonsmoker)  0.151  0.045, 0.257  0.247  0.084, 0.410  Insulin (μU/ml)      0.091  −0.021, 0.204  Waist-to-hip ratio      2.623  1.583, 3,663  Variables  Coefficient 1† (n = 5,903)  95% CI‡  Coefficient 2§ (n = 2,134)  95% CI  Female (vs. male)  −0.237  −0.328, −0.146  −0.041  −0.194, 0.112  Age (years)  0.006  0.003, 0.010  0.006  0.001, 0.012  Systolic blood pressure (mmHg)  0.007  0.005, 0.009  0.004  0.001, 0.007  Total cholesterol (mg/dl)  −0.002  −0.003, −0.001  −0.003  −0.005, −0.001  HDL‡ cholesterol (mg/dl)  −0.009  −0.012, −0.006  −0.001  −0.007, 0.004  Triglycerides (mg/dl)  0.230  0.147, 0.313  0.047  −0.102, 0.196  Blood glucose (mg/dl)  −0.118  −0.331, 0.096  −0.432  −0.863, −0.001  Fibrinogen (mg/dl)  0.011  0.011, 0.012  0.013  0.012, 0.015  BMI‡ (kg/m2)  0.048  0.038, 0.059  0.071  0.046, 0.096  Current smoker (vs. nonsmoker)  0.151  0.045, 0.257  0.247  0.084, 0.410  Insulin (μU/ml)      0.091  −0.021, 0.204  Waist-to-hip ratio      2.623  1.583, 3,663  * A numerical datum of lognormal C-reactive protein (lnCRP) as the dependent variable. A CRP level of less than the lowest detection value (0.03 mg/liter) was taken to be the half value (0.015 mg/liter). † Coefficient 1 included all of the subjects in the study. ‡ CI, confidence interval; HDL, high density lipoprotein; BMI, body mass index. § Coefficient 2 included some subjects in the study with data on waist-to-hip ratio and insulin. View Large CRP levels were compared with menopausal status. Premenopausal women tended to have lower CRP levels than did postmenopausal women. The age-adjusted geometric means were 0.08 and 0.10 mg/liter, respectively (p = 0.10). DISCUSSION To the best of our knowledge, this is the first study to provide cross-sectional data in relation to atherosclerotic risk factors and serum CRP levels in a large-scale, population-based study of the Japanese. An additional strength of this study was the quality of the sample collection and the precision of the CRP measurements, in which CV levels were lower than those reported in previous reports (5–7). In previous studies, the distribution of serum CRP levels was skewed, with a log-normal distribution for males (16–18). Koenig et al. (17) reported that the distribution of CRP levels was 55–80 percent for CRP values of less than 2 mg/liter, which is well below the range seen in routine CRP measurements used to monitor active inflammatory, infective, or tissue-damaging disorders (2, 4, 5, 7). In addition, in other population-based studies that included both sexes, the distribution also showed log-normal distribution (8–10). In our study, however, the distribution of CRP was highly skewed to lower levels than in previous studies (table 5), with a median value of 0.12 mg/liter, and 94 percent with less than 2 mg/liter, and appeared to have a combination of different distributions. TABLE 5. Comparison of the C-reactive protein levels in the Jichi Medical School Cohort Study with previous studies Study  Age (years)  Sex  No. of subjects  Median (mg/liter)  Mean (mg/liter)  Our data  ≥30  Male  2,275  0.16  0.83      Female  3,832  0.09  0.59  Kuller et al. (4)  35–57  Male  256    2.9  Ridker et al. (5)  40–84  Male  543  1.13  1.1*  Ridker et al. (6)†    Female  244  3.75    Tracy et al. (7)  ≥65  Male  89    2.32      Female  57    1.73  Tracy et al. (10)‡  ≥65  Male  400    2.67      Female      2.66  Danesb et al. (8)  35–64  Both  704 (225 females)  1.6    Mendall et al. (16)  50–69  Male  303  1.72    Koenig et al. (17)  45–64  Male  936  1.584  1.623*  Rohde et al. (18)  40–84  Male  1,172  1.3  2.0  Ford (9)  ≥20  Male  7,325  2.1  3.65      Female  8,244  2.1  4.59  Study  Age (years)  Sex  No. of subjects  Median (mg/liter)  Mean (mg/liter)  Our data  ≥30  Male  2,275  0.16  0.83      Female  3,832  0.09  0.59  Kuller et al. (4)  35–57  Male  256    2.9  Ridker et al. (5)  40–84  Male  543  1.13  1.1*  Ridker et al. (6)†    Female  244  3.75    Tracy et al. (7)  ≥65  Male  89    2.32      Female  57    1.73  Tracy et al. (10)‡  ≥65  Male  400    2.67      Female      2.66  Danesb et al. (8)  35–64  Both  704 (225 females)  1.6    Mendall et al. (16)  50–69  Male  303  1.72    Koenig et al. (17)  45–64  Male  936  1.584  1.623*  Rohde et al. (18)  40–84  Male  1,172  1.3  2.0  Ford (9)  ≥20  Male  7,325  2.1  3.65      Female  8,244  2.1  4.59  * Geometric mean. † Age not given in the paper. ‡ Numbers of subjects not given in the paper. The blank column not described in the paper. View Large In previous studies (8, 10, 16, 18), the relations among serum CRP and atherosclerotic risk factors were inconsistent. Mendall et al. (16) and other investigators in the United States (10, 18) showed that CRP was related to several atherosclerotic risk factors, while Danesh et al. (8) found no association except for smoking status and body mass index. Our data showed results similar to those in the study by Mendall et al. and in other US studies, except for total cholesterol and blood glucose in males. Although the reasons for the discrepancies between our results and the report by Danesh et al. are unclear, some factors that may be involved are differences in CRP assays and ethnic differences on the clustering of risk factors. However, our results suggest that CRP may be a risk marker for atherosclerotic diseases in the Japanese, as well as in Westerners. In females, there was no difference between smokers and nonsmokers. Because there were few female smokers in this study, however, the effect of smoking among females could not be ascertained. Our results showed positive associations among CRP, body mass index, and waist-to-hip ratio, but not with insulin. Hak et al. (19) showed that CRP was associated with body mass index and waist and hip circumference, but not with waist-to-hip ratio after adjustment for body mass index in 186 healthy, middle-aged women. In our study, the association between CRP and waist circumference showed results similar to those with the waist-to-hip ratio (data not shown). All previous studies have shown a positive association between CRP and body mass index, and recently, an association among CRP, insulin, and obesity has been suggested (9, 19–22). The mechanisms for the reason CRP had a positive association with body mass index or obesity may be explained as follows. The main modulators of CRP are interleukin 1 and 6 and tumor necrosis factor alpha (23). Adipocytes from obese humans have been shown to overproduce tumor necrosis factor alpha messenger RNA (24). Tumor necrosis factor alpha is a potent inducer of interleukin 6 production in various cells, which may explain the positive association between CRP and body mass index (16). In vitro, human abdominal visceral adipose tissue releases more interleukin 6 compared with subcutaneous adipose tissue (25); our data may support this. Although our data are cross-sectional and cannot address causality or exclude confounders that can explain the results, they suggest that CRP is strongly associated with fat distribution. The sex difference in CRP is controversial. In our study, males had higher CRP levels than did females. This result was consistent with one nested case-control study of 89 males and 57 females aged 65 years and over (7). One cross-sectional study of 400 subjects, which included both sexes aged 65 years and over, showed that males and females had similar values (10), while other studies showed that females had higher CRP levels than did males (6, 9). The reasons for these discrepancies are unclear, but it is important to note that the females in this analysis had lower body mass index and lower rates of smoking, which are associated with decreased CRP, than those in previous studies. Although the sex difference in our analysis may reflect the fact that males were more likely than females to be current smokers, the difference remained even after adjustment for smoking status: This may reflect sex-specific effects. In our study, premenopausal women tended to have lower CRP levels than did postmenopausal women. Hak et al. (19) also showed that CRP levels were slightly higher in postmenopausal compared with premenopausal women, although this was not statistically significant. Although experimental data suggest an inhibitory effect of estrogens on interleukin 6 gene expression (19, 26), recent studies suggested an increase in CRP with hormone replacement therapy (27–31). However, in rural areas of Japan, hormone replacement therapy is not yet widely used, so we believe that the effect of this therapy was small in our analysis. We think, therefore, that sex hormonal effect may be one reason why females had lower CRP levels than did males. In multivariate analyses, which included the waist-to-hip ratio and insulin, the results were different from those excluding these parameters. Although this may indicate that the waist-to-hip ratio has a stronger association with CRP than other variables, it is important to note that the data on waist-to-hip ratio and insulin included only about one third of all the subjects in our study. Therefore, we believe that the results including these data need to be considered carefully and that further studies are needed. The much lower levels of CRP compared with previous studies may be due to the fact that our study population included females, who had very low smoking rates, and a lower body mass index. However, even taking these factors into consideration, the difference was considerable. This finding suggests, therefore, that the Japanese have lower CRP levels than do Westerners, which may be reflected in the lower coronary event rate among the Japanese. In this study, CRP levels seemed to show a combination of two separate distributions. This finding cannot be explained, but it may indicate that the higher CRP group underwent subclinical atherosclerotic changes, possibly resulting in future atherosclerotic events. In conclusion, serum CRP levels correlated with atherosclerotic risk factors in the Japanese as well as in Westerners. However, the distribution of CRP in this study was quite different from that in previous studies, suggesting that the Japanese have much lower CRP levels than do those of other developed countries and that this could reflect the low coronary event rate among the Japanese. Reprint requests to Dr. Seishi Yamada, Department of Internal Medicine, Wara National Health Insurance Hospital, 882 Sawa Wara, Gujyo, Gifu 501–4595, Japan (e-mail: [email protected]). Supported in part by grants from the Foundation for the Department of the Community, Tochigi, Japan. REFERENCES 1. Ross R. The pathogenesis of atherosclerosis: a perspective for the 1990s. Nature  1993; 362: 802–9. Google Scholar 2. Haverkate F, Thompson SG, Pyke SD, et al. Production of C-reactive protein and risk of coronary events in stable and unstable angina. European Concerted Action on Thrombosis and Disabilities Angina Pectoris Study Group. Lancet  1997; 349: 462–6. Google Scholar 3. Liuzzo G, Biasucci LM, Gallimore JR, et al. The prognostic value of C-reactive protein and serum amyloid A protein in severe unstable angina. N Engl J Med  1994; 331: 417–24. Google Scholar 4. Kuller LH, Tracy RP, Shaten J, et al. Relation of C-reactive protein and coronary heart disease in the MRFIT nested case-control study. Multiple Risk Factor Intervention Trial. Am J Epidemiol  1996; 144: 537–47. Google Scholar 5. Ridker PM, Cushman M, Stampfer MJ, et al. Inflammation, aspirin, and the risk of cardiovascular disease in apparently healthy men. N Engl J Med  1997; 336: 973–9. Google Scholar 6. Ridker PM, Buring JE, Shih J, et al. Prospective study of C-reactive protein and the risk of future cardiovascular events among apparently healthy women. Circulation  1998; 98: 731–3. Google Scholar 7. Tracy RP, Lemaitre RN, Psaty BM, et al. Relationship of C-reactive protein to risk of cardiovascular disease in the elderly. Results from the Cardiovascular Health Study and the Rural Health Promotion Project. Arterioscler Thromb Vasc Biol  1997; 17: 1121–7. Google Scholar 8. Danesh J, Muir J, Wong YK, et al. Risk factors for coronary heart disease and acute-phase proteins. A population-based study. Eur Heart J  1999; 20: 954–9. Google Scholar 9. Ford ES. Body mass index, diabetes, and C-reactive protein among U.S. adults. Diabetes Care  1999; 22: 1971–7. Google Scholar 10. Tracy RP, Psaty BM, Macy E, et al. Lifetime smoking exposure affects the association of C-reactive protein with cardiovascular disease risk factors and subclinical disease in healthy elderly subjects. Arterioscler Thromb Vasc Biol  1997; 17: 2167–76. Google Scholar 11. Nago N, Kayaba K, Hiraoka J, et al. Lipoprotein(a) levels in the Japanese population: influence of age and sex, and relation to atherosclerotic risk factors: the Jichi Medical School Cohort Study. Am J Epidemiol  1995; 141: 815–21. Google Scholar 12. Gotoh T, Kuroda T, Yamasawa M, et al. Correlation between lipoprotein(a) and aortic valve sclerosis assessed by echocardiography (the JMS Cardiac Echo and Cohort Study). Am J Cardiol  1995; 76: 928–32. Google Scholar 13. Kario K, Nago N, Kayaba K, et al. Characteristics of the insulin resistance syndrome in a Japanese population: The Jichi Medical School Cohort Study. Arterioscler Thromb Vasc Biol  1996; 16: 269–74. Google Scholar 14. Schreiner PJ, Terry JG, Evans GW, et al. Sex-specific associations of magnetic resonance imaging-derived intra-abdominal and subcutaneous fat areas with conventional indices. The Atherosclerosis Risk in Communities Study. Am J Epidemiol  1996; 144: 335–45. Google Scholar 15. Itoh Y, Ichihara K, Kanno T, et al. Serum protein standardization project in Japan: evaluation of an IFCC reference material (RPPHS/CRM470) and establishment of reference intervals. J Clin Lab Anal  1997; 11: 39–44. Google Scholar 16. Mendall MA, Patel P, Ballam L, et al. C reactive protein and its relation to cardiovascular risk factors: a population based cross sectional study. BMJ  1996; 312: 1061–5. Google Scholar 17. Koenig W, Sund M, Frohlich M, et al. C-Reactive protein, a sensitive marker of inflammation, predicts future risk of coronary heart disease in initially healthy middle-aged men: results from the MONICA (Monitoring Trends and Determinants in Cardiovascular Disease) Augsburg Cohort Study, 1984 to 1992. Circulation  1999; 99: 237–42. Google Scholar 18. Rohde LEP, Hennekens CH, Ridker PM. Survey of C-reactive protein and cardiovascular risk factors in apparently healthy men. Am J Cardiol  1999; 84: 1018–22. Google Scholar 19. Hak AE, Stehouwer CD, Bots ML, et al. Associations of C-reactive protein with measures of obesity, insulin resistance, and subclinical atherosclerosis in healthy, middle-aged women. Arterioscler Thromb Vasc Biol  1999; 19: 1986–91. Google Scholar 20. Strandberg TE, Tilvis RS. C-reactive protein, cardiovascular risk factors, and prospective study in the elderly. Arterioscler Thromb Vasc Biol  2000; 20: 1057–60. Google Scholar 21. Yudkin JS, Stehouwer CD, Emeis JJ, et al. C-reactive protein in healthy subjects: associations with insulin resistance, and endothelial dysfunction: a potential cytokines originating from adipose tissue? Arterioscler Thromb Vasc Biol  1999; 19: 972–8. Google Scholar 22. Visser M, Bouter LM, McQuillan GM, et al. Elevated C-reactive protein levels in overweight and obese. JAMA  1999; 282: 2131–5. Google Scholar 23. Danesh J, Collins R, Appleby P, et al. Association of fibrinogen, C-reactive protein, albumin, or leukocyte count with coronary heart disease: meta-analyses of prospective studies. JAMA  1998; 279: 1477–82. Google Scholar 24. Hotamisligil GS, Arner P, Caro JF, et al. Increased adipose tissue expression of tumor necrosis factor-alpha in human obesity and insulin resistance. J Clin Invest  1995; 95: 2409–15. Google Scholar 25. Fried SK, Bunkin DA, Greenberg AS. Omental and subcutaneous adipose tissues of obese subjects release interleukin-6: depot difference and regulation by glucocorticoid. J Clin Endocrinol Metab  1998; 83: 847–50. Google Scholar 26. Ray P, Ghosh SK, Zhang DH, et al. Repression of interleukin-6 gene expression by 17 inhibition of the DNA-binding activity of the transcription NF-IL6 and NF-kappa B by the estrogen receptor. FEBS Lett  1997; 409: 79–85. Google Scholar 27. Walsh BW, Paul S, Wild RA, et al. The effects of hormone replacement therapy and raloxifene on C-reactive protein and homocysteine in healthy postmenopausal women: a randomized, controlled trial. J Clin Endocrinol Metab  2000; 85: 214–18. Google Scholar 28. van Baal WM, Kenemans P, van der Mooren MJ, et al. Increased C-reactive protein levels during short-term hormone replacement therapy in healthy postmenopausal women. Thromb Haemost  1999; 81: 925–8. Google Scholar 29. Ridker PM, Hennekens CH, Rifai N, et al. Hormone replacement therapy and increased plasma concentration of C- reactive protein. Circulation  1999; 100: 713–16. Google Scholar 30. Cushman M, Meilahn EN, Psaty BM, et al. Hormone replacement therapy, inflammation, and hemostasis in women. Arterioscler Thromb Vasc Biol  1999; 19: 893–9. Google Scholar 31. Cushman M, Legault C, Barrett-Connor E, et al. Effect of postmenopausal hormones on inflammation-sensitive proteins: the Postmenopausal Estrogen/Progestin Interventions (PEPI) Study. Circulation  1999; 100: 717–22. Google Scholar

Journal

American Journal of Epidemiology – Oxford University Press

Published: Jun 15, 2001

There are no references for this article.