Genetic and Environmental Associations Between C-Reactive Protein and Components of the Metabolic Syndrome

Abstract

Background:

Metabolic syndrome has been suggested to have an association with C-reactive protein (CRP), a biomarker of cardiovascular disease risk. Given that genetic factors influence both metabolic syndrome and CRP, it seems necessary to examine the association with consideration of genetic influence.

Methods:

We conducted a cross-sectional study and a co-twin–control study in 2555 Korean adults composed of twins and their family members. For the co-twin–control study, 113 pairs of monozygotic twins who were discordant in regard to CRP level (>0.5 mg/L) were selected. Cross-trait additive genetic correlation between CRP and metabolic syndrome and the risk for having higher CRP level associated with components of metabolic syndrome were estimated.

Results:

With increasing CRP level, the prevalence of metabolic syndrome increased linearly. Among components of metabolic syndrome, high-density lipoprotein cholesterol (HDL-C) was inversely associated with CRP, whereas other components were positively associated. Most of the components of metabolic syndrome except for HDL-C had a significant genetic correlation with CRP, with the highest correlation for obesity indices. A co-twin–control study that allows full control of genetic influence showed that only obesity was significantly associated with higher CRP levels: Odds ratios (95% confidence intervals) were 1.23 (1.04,1.46) for 1 kg/m² increase in body mass index and 1.12 (1.03,1.22) for 1% increased in total body fat, respectively.

Conclusions:

Although genetic influence played a significant role in the associations between CRP and most metabolic syndrome components, environmental influence that may be modifiable also contributed to the association, especially to the associations between the obesity indices and CRP.

Introduction

C-reactive protein (CRP), one of the inflammatory markers, has been strongly suggested as a biomarker of cardiovascular disease (CVD) risk.^1

–6 The Centers for Disease Control and Prevention (CDC), along with the American Heart Association (AHA) endorsed the use of high-sensitivity CRP (hsCRP) as an adjunct to improve coronary heart disease risk prediction⁷

As many other biological factors do, the level of baseline CRP may vary between the healthy individuals.⁸ Although CRP is supposed to be influenced by age^9,10 and sex,^10,11 genetic factors were also estimated to contribute about 35%–50% of the phenotypic variation of CRP in the twin and family studies.^12
–14 In addition, individual components of metabolic syndrome, such as obesity,^9,11,15 high blood pressure,¹⁶ low high-density lipoprotein cholesterol (HDL-C),¹⁷ hypertriglyceridemia,¹⁷ high fasting glucose,^17,18 fasting hyperinsulinemia,¹⁹ and insulin resistance measured by homeostasis model assessment (HOMA)¹⁶ were also associated with increased CRP level²⁰ in previous studies.^20,21

On the other hand, metabolic syndrome and components of metabolic syndrome are also estimated to be highly heritable.^22,23 Given that both metabolic syndrome and CRP are under strong genetic influence, it seems necessary to look into the role of pleiotropy (genetic effect of a single gene on multiple phenotypic traits, or common genetic factors influencing both phenotypes through shared pathways) in the association between CRP and metabolic syndrome. Efforts for estimating how much of the association between CRP and metabolic syndrome are under common genetic influence may contribute to finding clues for modifying CRP level and, in turn, reducing the risk of CVD. However, up to now, this issue has not been well addressed and findings were somewhat controversial.^24
–26

In this regard, we evaluated the genetic and environmental association between CRP and components of metabolic syndrome in a study of Korean family composed of twins and their family members. In addition, by performing a co-twin–control study, we evaluated the environmental association between CRP and components of metabolic syndrome, with full consideration of the effect of age, sex, and genetic factors.

Subjects and Methods

Study participants

Study participants were twins and their family members enrolled for the Healthy Twin study, a nationwide ongoing multicenter cohort study being conducted as a part of the Korean Genome Epidemiology Study since April, 2005. The Healthy Twin Study has recruited same-sex adult Korean twins (≥30 years of age) and their first-degree adult family members from the general population to investigate the quantitative loci of complex traits, as well as the role of environment in the etiology of complex diseases. Details on methodology of the Healthy Twin Study have been published previously.²⁷ Among the 3079 subjects who have participated in the Healthy Twin study between April, 2005, and March, 2011, and measured hsCRP, a total of 524 were excluded due to the following reasons: Past medical history of cancer, chronic liver disease, rheumatoid arthritis, stroke, or myocardial infarction (n=325); recent use of antiinflammatory medication (n=42), upper respiratory tract medication (n=88), or antibiotics (n=21); extreme (upper 1%) hsCRP level (>19.46) or white blood cell (WBC) count (>11,500) (n=48). Thus, a total of 2555 subjects were included in this study with 425 pairs of monozygotic (MZ) twins, 92 pairs of dizygotic (DZ) twins, and 1521 singleton family members from 652 families.

The zygosity of 67% of the twin pairs was ascertained using 16 short tandem repeat markers (15 autosomal short tandem repeat markers and 1 sex-determining marker). For the remaining 33% of twin pairs, zygosity was determined by a self-administered zygosity questionnaire that was validated to be 94.3% accurate.²⁸ All participants gave written informed consent. The Healthy twin study protocol was approved by the local institutional review board of each participating center.

Study variables

Blood samples were drawn after fasting for 12 hr overnight and shipped to a designated central laboratory institute, accredited by the Korea Association of Quality Assurance for Clinical Laboratory, on the same day. Serum concentrations were measured in fresh sera with commercial kits for hsCRP (turbid immunoassay), glucose (hexokinase enzymatic assay), total cholesterol (enzymatic assay), HDL-C (enzymatic or homogeneous assay), triglyceride (enzymatic assay), and low-density lipoprotein cholesterol (LDL-C) (enzymatic assay) on ADVIA 1650 (Siemens, Germany) or HITACHI 7600-210/HITACHI 7180 (HITACHI, JAPAN) equipment. We calculated non-HDL-C by subtracting HDL-C from total cholesterol. Insulin was measured with radioimmunoassay on an automatic gamma counter (PacKard, USA or Wizard 1470, PerkinElmer Wallac, Turku, Finland). Insulin sensitivity was mathematically estimated using a homeostasis model assessment (HOMA) (fasting plasma insulin×fasting plasma glucose/22.5).²⁹ Interassay coefficient of variation for hsCRP measurement was set below 7%.

Weight (kg) and height (cm) were measured in light clothing using standardized scales and stadiometers. Body mass index (BMI) was calculated as the weight divided by the height squared (kg/m²). Waist circumference (WC) was measured (cm) at the narrowest region between the lower margin of the rib cage and iliac crest as viewed from the front while the participant was standing with arms at the side. Blood pressure was measured manually using a standard mercury sphygmomanometer, while participants were in a sitting position. All the procedures regarding the measurement of body size and blood pressure were standardized between the participating centers through the development of a standard protocol and training of research coordinators and research assistants. All physical measurements were completed twice for each participant, and the average value of two measurements was used for the analysis. Total body fat mass was measured using a dual-energy X-ray absorptiometry (Lunar Radiation, Madison, WI, Delphi W, Hologic, Boston, MA). Then, total body fat percentage (fat%) was calculated as fat mass divided by the sum of fat mass, lean mass, and bone mineral content.

The presence of metabolic syndrome was determined based on the criteria proposed by both National Cholesterol Education Program Adult Treatment Panel III (NCEP ATP III) and International Diabetes Federation (IDF). Metabolic risk factors, such as WC, triglyceride, HDL-C, systolic blood pressure (SBP) and diastolic blood pressure (DBP), and glucose were considered for the diagnosis of metabolic syndrome.

A self-administered questionnaire was used to collect information about weight at 20 years of age and health behaviors (smoking and alcohol consumption). Weight change since 20 years of age was calculated by subtracting the weight at 20 years of age from the current weight. The lifelong amount of cigarette smoking (pack-years) was calculated based on the data of duration for smoking and the mean amount of smoking per day. Weekly consumption of alcohol (grams/week) was calculated based on the data of frequency of drinking per week and the mean amount of consumption at each drinking time.

Statistical analyses

Becuase hsCRP did not normally distribute in this study, log transformation of CRP data was done. Associations of CRP with the components of metabolic syndrome were evaluated using a linear mixed model in which the correlation structures from family relationships was considered by adjusting for household effect as well as twin effects as random effects. Other covariates (age, sex, smoking, and alcohol consumption) were adjusted as fixed effects. Alcohol intake^18,30 and cigarette smoking^9
–11 have been reported to be associated with CRP level in previous studies.

To ascertain evidence of common genetic regulation between CRP and the components of metabolic syndrome, we conducted bivariate analysis, using SOLAR (Sequential Oligogenic Linkage Analysis Routines) ver. 4.2.0,³¹ which allowed partitioning of the phenotypic correlations into genetic (ρ_G) and environmental correlations (ρ_E). The presence of a significant genetic correlation between two characteristics could be considered evidence of pleiotropy.

To evaluate environmental associations between CRP and components of metabolic syndrome with full consideration of the effect of age, sex, and genetic factors, we conducted a co-twin–control study in 113 pairs of mozygotic twins. If a twin pair were discordant by more than 0.5 mg/L regarding CRP levels, they were selected for the co-twin–control study. The CRP level for identifying discordant twin pairs was set on the basis of a clinical intervention study evaluating the effect of lipid-lowering agent on CRP level.³² We conducted a within-pair comparison of each component of metabolic syndrome using a paired t-test. We estimated the risk for having higher CRP levels associated with the components of metabolic syndrome by conditional logistic regression analysis with an adjustment for smoking and alcohol consumption, in which a person who had a higher CRP level than his/her co-twin was compared with the co-twin regarding the presumed risk factors. All of the hypotheses were tested bidirectionally with an alpha level set at 0.05.

Results

Table 1 shows metabolic and behavioral characteristics according to the level of CRP. Subjects with higher CRP levels tended to be older than those with lower CRP levels. All of the obesity-related indices, including BMI, total body fat, WC, and weight change since 20 years of age, increased gradually with the increase of CRP level. Glucose, insulin, and HOMA also increased gradually, according to the increase in CRP level. SBP and DBP, total cholesterol, non-HDL-C, triglycerides, and LDL-C were more likely to be higher in participants who have higher CRP levels. HDL-C levels decreased with increasing CRP levels. Prevalence of metabolic syndrome increased progressively with increasing CRP level whether we applied the diagnostic criteria recommended by NCEP ATP III or by IDF. When we repeated this analysis with restriction to males or females, these findings did not change except for alcohol consumption (p for trend=0.43 in males) and smoking (p for trend=0.35 in females).

Table 1.

Metabolic and Behavioral Characteristics According to the Level of C-Reactive Protein

		Levels of C-reactive protein
Characteristics, mean±SD or %	Number of persons ^a	<0.5	0.5 to <1	1–3	>3	P_trend ^b
Number of persons in each C-reactive protein group		1214	678	487	176	2555
C-reactive protein (mg/L)	2555	0.27±0.11	0.67±0.14	1.61±0.55	6.16±3.58	<0.01
Male	2555	31.1%	45.7%	48.1%	50.0%	<0.01
Age (years)	2555	39.8±11.3	45.0±13.1	45.6±13.4	46.9±13.6	<0.01
Body mass index (kg/m²)	2555	22.4±2.7	24.0±2.8	25.0±3.4	25.7±4.5	<0.01
Total body fat (%)	2547	27.1±7.7	28.4±7.9	29.5±8.2	29.5±9.2	<0.01
Waist circumference (cm)	2554	76.4±7.9	81.9±8.2	84.7±8.8	86.2±10.8	<0.01
Weight change since 20-year old (kg)	2101	4.5±6.4	8.0±7.4	10.1±8.2	10.8±11.3	<0.01
Systolic blood pressure (mmHg)	2547	112.2±15.1	118.2±16.5	121.6±17.4	121.4±19.9	<0.01
Diastolic blood pressure (mmHg)	2547	71.7±10.2	74.1±10.5	76.9±11.6	75.8±11.4	<0.01
Total cholesterol (mmol/L)	2555	4.74±0.88	5.03±0.93	5.04±0.96	4.98±0.84	<0.01
Non-high-density lipoprotein cholesterol (mmol/L)	2555	3.40±0.86	3.74±0.91	3.83±0.96	3.78±0.87	<0.01
Low-density lipoprotein cholesterol (mmol/L)	2555	2.70±0.74	2.99±0.82	2.98±0.83	2.98±0.79	<0.01
High-density lipoprotein cholesterol (mmol/L)	2555	1.34±0.32	1.29±0.31	1.21±0.31	1.20±0.33	<0.01
Triglyceride (mmol/L)	2555	1.05±0.80	1.44±0.92	1.67±1.28	1.56±1.14	<0.01
Glucose (mmol/L)	2555	4.96±0.62	5.25±0.98	5.37±1.19	5.52±1.36	<0.01
Insulin (pmol/L)	2555	44.2±22.7	48.2±33.0	52.6±33.1	56.5±42.8	<0.01
Homeostasis model assessment^c	2555	1.37±0.76	1.59±1.20	1.85±1.46	2.00±1.87	<0.01
Metabolic syndrome by ATP III criteria	2547	4.04%	13.27%	23.82%	32.95%	<0.01^d
Metabolic syndrome by IDF criteria	2547	5.11%	16.37%	28.95%	31.25%	<0.01^d

In some participants, data were missing for some variables.

Linear trend for each component (continuous form) according to the level of C-reactive protein (continuous form) was examined by linear regression analysis.

(Glucose×insulin)/22.5.

Trend for each component (two categories) according to the level of C-reactive protein (4 categories) was examined by the Mantel–Haenszel chi-squared test.

SD, standard deviation; ATP, National Cholesterol Education Program Adult Treatment Panel; IDF, International Diabetes Federation.

Table 2 shows the association and the cross-trait correlation between CRP and components of metabolic syndrome. CRP was inversely associated with HDL-C. All other components of metabolic syndrome were positively associated with CRP. Obesity-related indices had stronger phenotypic correlations with CRP than other components, such as blood pressure, lipid profiles, glucose, insulin, and HOMA. When the phenotypic correlations were further partitioned into genetic (ρ_G) and environmental correlations, genetic correlations between obesity indices and CRP were significantly stronger than environmental correlations. HDL-C had no genetic correlation with CRP, whereas there was a weak but significant inverse environmental correlation. LDL-C, DBP, and glucose had weak positive genetic correlations with CRP, whereas they had no environmental correlation with CRP. SBP, non-HDL-C, triglyceride, insulin, and HOMA had weak but significant positive genetic and environmental correlations with CRP.

Table 2.

Associations and Cross-Trait Correlations (95% Confidence Intervals) Between C-Reactive Protein and Components of Metabolic Syndrome

		Cross-trait correlations
Components of metabolic syndrome	Beta coefficients ^a	Phenotypic ^b	Additive genetic ^c	Environmental ^c
Body mass index (kg/m²)	0.043 (0.037,0.049)	0.315 (0.280,0.350)	0.473 (0.393,0.545)	0.205 (0.125,0.282)
Total body fat (%)	0.019 (0.016,0.022)	0.323 (0.288,0.358)	0.417 (0.333,0.494)	0.214 (0.143,0.283)
Waist circumference (cm)	0.017 (0.015,0.019)	0.349 (0.314,0.382)	0.526 (0.453,0.593)	0.188 (0.110,0.264)
Weight change since 20-year old (kg)	0.014 (0.011,0.016)	0.310 (0.270,0.348)	0.512 (0.400,0.609)	0.171 (0.089,0.251)
Systolic blood pressure (mmHg)	0.028 (0.015,0.042)	0.139 (0.101,0.177)	0.177 (0.050,0.297)	0.090 (0.011,0.167)
Diastolic blood pressure (mmHg)	0.033 (0.013,0.051)	0.089 (0.050,0.127)	0.153 (0.038,0.263)	0.052 (–0.027,0.130)
Total cholesterol (mmol/L)	0.036 (0.014,0.058)	0.122 (0.084,0.160)	0.103 (−0.014,0.218)	0.098 (0.021,0.173)
Non-high-density lipoprotein cholesterol (mmol/L)	0.060 (0.038,0.083)	0.159 (0.121,0.196)	0.131 (0.022,0.236)	0.169 (0.094,0.242)
Low-density lipoprotein cholesterol (mmol/L)	0.052 (0.026,0.078)	0.137 (0.099,0.175)	0.202 (0.087,0.312)	0.058 (–0.021,0.136)
High-density lipoprotein cholesterol (mmol/L)	–0.161 (–0.210,–0.110)	–0.110 (−0.148,-0.072)	–0.072(–0.179,0.036)	–0.149 (–0.224,–0.072)
Triglyceride (mmol/L)	0.075 (0.054,0.095)	0.256 (0.219,0.292)	0.284 (0.153,0.405)	0.136 (0.066,0.204)
Glucose (mmol/L)	0.079 (0.049,0.110)	0.140 (0.102,0.178)	0.333 (0.215,0.442)	0.008 (–0.070,0.085)
Insulin (pmol/L)	0.022 (0.015,0.028)	0.128 (0.089,0.166)	0.170 (0.024,0.309)	0.105 (0.033,0.175)
Homeostasis model assessment^d	0.071 (0.052,0.092)	0.150 (0.112,0.188)	0.267 (0.142,0.383)	0.085 (0.007,0.162)

Beta coefficient for log-transformed level of C-reactive protein assessed by linear mixed model in which random effect (household, twin pair) and fixed effect (age, sex, smoking habit, and alcohol consumption) were adjusted.

Correlation coefficients (standard deviation) were assessed by Spearman correlation analysis with an adjustment for age, sex, smoking habit, and alcohol consumption.

Estimates (standard error) were assessed by bivariate analysis with an adjustment for age, sex, smoking habit, and alcohol consumption.

(Glucose×insulin)/22.5.

Table 3 shows the findings from co-twin–control analysis in 113 pairs of MZ twins whose within-pair difference in CRP level was greater than 0.5. Compared with co-twins, twin members who had higher CRP levels had significantly higher BMI, total body fat, WC, and weight change since 20 years of age. However, other components of metabolic syndrome such as blood pressure, lipid profiles, glucose, insulin, and HOMA did not differ between the twin pairs having discordant CRP levels. When we examined the risk for having higher CRP levels associated with components of metabolic syndrome by conditional logistic regression analysis with an adjustment for smoking and alcohol consumption, increases in BMI and total body fat were shown to be significantly associated with the risk for having higher CRP levels.

Table 3.

Risk Estimation for Having Higher Level of C-Reactive Protein: Co-twin–Control Analysis in 113 Pairs of Monozygotic Twins Whose C-Reactive Protein Levels Differ Each Other by More than 0.5 mg/L

		Level of C-reactive protein
Components of metabolic syndrome	Number of twin pairs	High	Low	P_difference	Odds ratio (95% CI) ^a
Body mass index (kg/m²), mean (SD)	130	23.9 (3.3)	23.4 (3.1)	0.007	1.23 (1.04,1.46)
Total body fat (%), mean (SD)	128	29.4 (8.4)	28.2 (8.0)	0.003	1.12 (1.03,1.22)
Waist circumference (cm), mean (SD)	130	81.1 (8.1)	79.9 (8.0)	0.032	1.06 (0.99,1.12)
Weight change since 20 years old (kg), mean (SD)	115	8.60 (7.72)	7.65 (7.46)	<0.001	1.01 (0.96,1.08)
Systolic blood pressure (mmHg), mean (SD)	129	114.5 (14.5)	114.4 (16.0)	0.989	0.99 (0.97,1.03)
Diastolic blood pressure (mmHg), mean (SD)	129	73.5 (10.7)	74.0 (11.6)	0.442	0.98 (0.94,1.03)
Total cholesterol (mmol/L), mean (SD)	130	4.85 (0.94)	4.79 (0.82)	0.445	1.06 (0.97,1.16)
Non-high density lipoprotein cholesterol (mmol/L), mean (SD)	130	3.56 (0.94)	3.50 (0.79)	0.383	1.20 (0.72,2.00)
Low density lipoprotein cholesterol (mmol/L)	130	2.86 (0.83)	2.84 (0.71)	0.763	1.06 (0.97,1.16)
High density lipoprotein cholesterol (mmol/L)	130	1.29 (0.34)	1.29 (0.32)	1.00	0.96 (0.30,3.12)
Triglycerides (mmol/L)	130	1.19 (0.73)	1.14 (0.71)	0.515	1.09 (0.69,1.72)
Glucose (mmol/L)	130	5.01 (0.62)	5.05 (0.97)	0.644	0.90 (0.55,1.49)
Insulin (mmol/L)	130	45.4 (26.8)	45.1(20.4)	0.880	1.00 (0.99,1.01)
Homeostasis model assessment^b	130	1.44 (0.94)	1.42 (0.70)	0.800	1.02 (0.69,1.50)

Estimates for having higher level of C-reactive protein than co-twin were assessed by Conditional logistic regression analysis with an adjustment for smoking habit and alcohol consumption.

(Glucose×insulin)/22.5.

95% CI, 95% confidence intervals; SD, standard deviation.

Discussion

The important aspect of CRP in its association with metabolic syndrome^17,20 is its role as a biomarker of increased risk for CVD. In this study of Korean twins and their family members, CRP was significantly associated with metabolic syndrome per se and each component of metabolic syndrome, which was very consistent with the findings from other studies.^{6,9,15

–18} The strength of the genetic and environmental correlations with CRP varied across the components of metabolic syndrome. Among the components of metabolic syndrome, obesity indices have the highest correlation with CRP in our study, as well as in other previous studies.²⁴

Several mechanisms could be assumed to explain the strong association between CRP and obesity. First, adipose tissue has been suggested to be an important source of proinflammatory cytokines, such as interleukin-6 (IL-6) and tumor necrosis factor-α (TNF-α), that in turn regulate chronic elevation of CRP level.^25,33 Thus, CRP may be responsible for the metabolic consequences of obesity through the action of those proinflammatory cytokines. However, genes encoding proteins that are involved in the antiinflammatory and immune response have been reported to be expressed differentially, according to the presence of metabolic syndrome, in the visceral adipose tissue of obese men.³⁴ In addition, the positive relation between CRP and CVD risk was found to be valid, even within morbidly obese persons.³³ Thus, there could be some other factors other than the secondary effect of obesity. Second, it may be possible that both obesity and CRP are under the regulation of a shared genetic effect, which may explain the association between CRP and obesity to some extent. Greenfield et al.²⁴ investigated the role of environmental effect in 94 pairs of monozygotic twins by conducting a co-twin–control study.³⁴ In the study, within-pair differences in CRP were associated with within-pair differences in obesity measures, and these investigators concluded that obesity is an important determinant of CRP independent of genetic influences. However, co-twin–control design cannot give information about genetic effect because genetic influence is fully controlled. Keenan et al.²⁶ conducted a variance component analysis in 38 extended families, in which CRP level was found to have significant environmental correlation (ρ_E=0.43±0.08, p<0.0001) but no genetic correlation (ρ_G=0.015±0.25, p=0.95) with BMI. However, a Chinese study including 590 pairs of monozygotic or dizygotic adolescent twins showed that 86%–89% of the phenotypic correlations between adiposity measures and CRP were attributed to shared genetic factors and 11%–14% to common unique environmental factors in both sexes.²⁵

One of the strengths of our study is that we could partition the correlation between CRP and each component of metabolic syndrome into genetic and environmental components, using the advantage of twin and family data. In our study, we found that strong genetic correlations exist between obesity indices and CRP, and the genetic correlations were greater than those of environmental correlations for all obesity indices. The consistency in the findings between our study and Chinese study strongly suggest that both obesity and CRP are under the regulation of shared genetic effect to some extent.

There was a genome-wide association study supporting the findings of our study, which was conducted with 336,108 SNPs in 6345 healthy women participating in the Women's Genome Health Study to find genes outside the CRP locus involving the regulation of CRP level. Several genes that are directly involved in metabolic syndrome, insulin resistance, and weight homeostatis were revealed to be associated with CRP level, such as leptin receptor (LEPR), interleukin 6 receptor (IL^r), glucokinase (hexokinase 4) regulator (GCKR), and the hepatic transcription-factor gene HNF1 homeobox A (HNF1A). In accordance with the findings from the genome-wide association study, there were significant additive genetic associations between CRP and obesity indices in our study. To further investigate genetic pathways that may link CRP level with obesity and metabolic syndrome in the Korean population, future studies for identifying candidate genes focused on the CRP gene itself as well as the genes outside the CRP locus are warranted.

Using a co-twin–control study design, we demonstrated that obesity played the most significant role in determining the level of CRP among the components of the metabolic syndrome. The co-twin–control study allowed us to examine the unique environmental influence of the components of the metabolic syndrome on CRP, after the full consideration of sex, age, genetic factors, and even shared environmental factors. Findings from the co-twin–control study that only greater BMI and total body fat are associated with the risk of having higher CRP suggest that environmental factors play a significant role in determining the association between obesity and the CRP. However, our study has the limitation of suggesting a plausible pathway through which environmental factors influence obesity and CRP level. Therfore, further study is needed on specific environmental factors that were reported to have some influence on the level of CRP, such as dietary intake of macronutrients or micronutrients.²¹

There were some limitations in our study. First, we used hsCRP values obtained from a single measurement for this study. Given that CRP is a clinical acute-phase reactant, there might be some measurement bias. However, studies with repeated measurement of CRP in the same individuals showed CRP concentration was relatively stable over the course of several months.^35
–37 Thus, significant bias from the single-measurement of CRP seems less likely. Second, due to the limitation of the cross-sectional design, our study could not tell exactly which obesity control would result in a reduction of CRP levels, as suggested by the findings of significant association between obesity and CRP from the co-twin–control study. Third, the mean age of participants in this study was 42.8 [standard deviation (SD) 12.7] years, and 94% were younger than 65 years. Therefore, findings from our study are not generalizable to an old population. In conclusion, the significant association between CRP and metabolic syndrome components was largely determined by genetic influence. However, an environmental influence was also found to play a significant role in the association between CRP and some components of metabolic syndrome, especially on the association between CRP and obesity indices.

Footnotes

Acknowledgments

This study was supported by the National Genome Research Institute, Korea, National Institute of Health research contract 2011E7101100 and by the Basic Science Research Program through the National Research Foundation of Korea funded by the Ministry of Education, Science and Technology (NRF 2012-0004255).

Author Disclosure Statement

No competing financial interests exist.

References

Boekholdt

, Hack

, Sandhu

et al. C-reactive protein levels and coronary artery disease incidence and mortality in apparently healthy men and women: The EPIC-Norfolk prospective population study 1993–2003. Atherosclerosis, 2006; 187:415–422.

Ridker

, Rifai

, Cook

et al. Non-HDL cholesterol, apolipoproteins A-I and B100, standard lipid measures, lipid ratios, and CRP as risk factors for cardiovascular disease in women. JAMA, 2005; 294:326–333.

Sattar

, Murray

, McConnachie

et al.

C-reactive protein and prediction of coronary heart disease and global vascular events in the Prospective Study of Pravastatin in the Elderly at Risk (PROSPER)

Circulation, 2007; 115:981–989.

Cao

, Arnold

, Manolio

et al. Association of carotid artery intima-media thickness, plaques, and C-reactive protein with future cardiovascular disease and all-cause mortality: The Cardiovascular Health Study. Circulation, 2007; 116:32–38.

Koenig

, Lowel

, Baumert

et al. C-reactive protein modulates risk prediction based on the Framingham Score: implications for future risk assessment: results from a large cohort study in southern Germany. Circulation, 2004; 109:1349–1353.

Danesh

, Wheeler

, Hirschfield

et al. C-reactive protein and other circulating markers of inflammation in the prediction of coronary heart disease. N Engl J Med, 2004; 350:1387–1397.

Pearson

, Mensah

, Alexander

et al. Markers of inflammation and cardiovascular disease: application to clinical and public health practice: A statement for healthcare professionals from the Centers for Disease Control and Prevention and the American Heart Association. Circulation, 2003; 107:499–511.

Ridker

. High-sensitivity C-reactive protein: potential adjunct for global risk assessment in the primary prevention of cardiovascular disease. Circulation, 2001; 103:1813–1818.

Mendall

, Patel

, Ballam

et al. C reactive protein and its relation to cardiovascular risk factors: A population based cross sectional study. BMJ, 1996; 312:1061–1065.

10.

Slade

, Offenbacher

, Beck

et al. Acute-phase inflammatory response to periodontal disease in the US population. J Dent Res, 2000; 79:49–57.

11.

Danesh

, Muir

, Wong

et al. Risk factors for coronary heart disease and acute-phase proteins. A population-based study. Eur Heart J, 1999; 20:954–959.

12.

MacGregor

, Gallimore

, Spector

et al. Genetic effects on baseline values of C-reactive protein and serum amyloid a protein: A comparison of monozygotic and dizygotic twins. Clin Chem, 2004; 50:130–134.

13.

Pankow

, Folsom

, Cushman

et al. Familial and genetic determinants of systemic markers of inflammation: The NHLBI family heart study. Atherosclerosis, 2001; 154:681–689.

14.

Crawford

, Sanders

, Qin

et al. Genetic variation is associated with C-reactive protein levels in the Third National Health and Nutrition Examination Survey. Circulation, 2006; 114:2458–2465.

15.

Visser

, Bouter

, McQuillan

et al. Elevated C-reactive protein levels in overweight and obese adults. JAMA, 1999; 282:2131–2135.

16.

Festa

, D'Agostino

Jr. , Howard

et al.

Chronic subclinical inflammation as part of the insulin resistance syndrome: The Insulin Resistance Atherosclerosis Study (IRAS)

Circulation, 2000; 102:42–47.

17.

Ridker

, Buring

, Cook

et al. C-reactive protein, the metabolic syndrome, and risk of incident cardiovascular events: An 8-year follow-up of 14 719 initially healthy American women. Circulation, 2003; 107:391–397.

18.

Koenig

, Sund

, Frohlich

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; 99:237–242.

19.

Reiner

, Barber

, Guan

et al. Polymorphisms of the HNF1A gene encoding hepatocyte nuclear factor-1 alpha are associated with C-reactive protein. Am J Hum Genet, 2008; 82:1193–1201.

20.

Haffner

. The metabolic syndrome: Inflammation, diabetes mellitus, and cardiovascular disease. Am J Cardiol, 2006; 97:3A–11A.

21.

Shen

, Ordovas

. Impact of genetic and environmental factors on hsCRP concentrations and response to therapeutic agents. Clin Chem, 2009; 55:256–264.

22.

Sung

, Lee

, Song

. Heritabilities of the metabolic syndrome phenotypes and related factors in Korean twins. J Clin Endocrinol Metab, 2009; 94:4946–4952.

23.

Kraja

, Hunt

, Pankow

et al. Quantitative trait loci for metabolic syndrome in the Hypertension Genetic Epidemiology Network study. Obes Res, 2005; 13:1885–1890.

24.

Greenfield

, Samaras

, Jenkins

et al. Obesity is an important determinant of baseline serum C-reactive protein concentration in monozygotic twins, independent of genetic influences. Circulation, 2004; 109:3022–3028.

25.

Wang

, Christoffel

, Brickman

et al. C-reactive protein in adolescent twins: Patterns and relationship to adiposity. J Clin Endocrinol Metab, 2011; 96:3226–3233.

26.

Keenan

, Poznik

, Varo

et al. Identification of a locus modulating serum C-reactive protein levels on chromosome 5p15. Atherosclerosis, 2008; 196:863–870.

27.

Sung

, Cho

, Lee

et al. Healthy Twin: A twin-family study of Korea—protocols and current status. Twin Res Hum Genet, 2006; 9:844–848.

28.

Song

, Lee

et al. Validity of the zygosity questionnaire and characteristics of zygosity-misdiagnosed twin pairs in the Healthy Twin Study of Korea. Twin Res Hum Genet, 2010; 13:223–230.

29.

Matthews

, Hosker

, Rudenski

et al. Homeostasis model assessment: Insulin resistance and beta-cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia, 1985; 28:412–419.

30.

Albert

, Glynn

, Ridker

. Alcohol consumption and plasma concentration of C-reactive protein. Circulation, 2003; 107:443–447.

31.

Almasy

, Blangero

. Multipoint quantitative-trait linkage analysis in general pedigrees. Am J Hum Genet, 1998; 62:1198–1211.

32.

Bays

, Stein

, Shah

et al. Effects of simvastatin on C-reactive protein in mixed hyperlipidemic and hypertriglyceridemic patients. Am J Cardiol, 2002; 90:942–946.

33.

Faintuch

, Marques

, Bortolotto

et al.

Systemic inflammation and cardiovascular risk factors: Are morbidly obese subjects different?

Obes Surg, 2008; 18:854–862.

34.

Ridker

, Pare

, Parker

et al. Loci related to metabolic-syndrome pathways including LEPR,HNF1A, IL6R, and GCKR associate with plasma C-reactive protein: The Women's Genome Health Study. Am J Hum Genet, 2008; 82:1185–1192.

35.

Clark

, Fraser

. Biological variation of acute phase proteins. Ann Clin Biochem, 1993; 30,Pt 4:373–376.

36.

de Maat

, de Bart

, Hennis

et al. Interindividual and intraindividual variability in plasma fibrinogen, TPA antigen, PAI activity, and CRP in healthy, young volunteers and patients with angina pectoris. Arterioscler Thromb Vasc Biol, 1996; 16:1156–1162.

37.

Sakkinen

, Macy

, Callas

et al. Analytical and biologic variability in measures of hemostasis, fibrinolysis, and inflammation: Assessment and implications for epidemiology. Am J Epidemiol, 1999; 149:261–267.