Viruses and bacteria like Chlamydia pneumoniae and Helicobacter pylori have been suggested to have a role in pathogenesis of overweight and obesity.
We studied whether C. pneumoniae-specific IgG antibodies are associated with elevated body mass index (BMI), waist and hip circumference, and/or waist–hip ratio (WHR), and whether the risk is more pronounced in the simultaneous presence of an ongoing inflammation as measured by elevated high-sensitive C-reactive protein (hsCRP) levels.
Subjects and methods:
Our study population was derived from the Northern Finland Birth Cohort 1966 (NFBC1966), a general population sample of 12 058 live-born children. This cross-sectional study consisted of 5044 persons at 31 years of age. Serum C. pneumoniae IgG titers were measured by microimmunofluorescence test, and hsCRP levels by immunoenzymometric assay.
C. pneumoniae IgG positivity (titer 32), both alone and jointly with elevated hsCRP (1.64 mg l−1, an upper quartile), was found to significantly associate with elevated BMI in the whole study population and with elevated hip and waist circumference in women, yet no association with WHR was seen. The analyses were adjusted for sex (when appropriate), smoking, socioeconomic position, glucose, insulin, high- and low-density lipoprotein cholesterols, triglycerides, leukocytes and pulse pressure.
These findings suggest that especially in women, persistent C. pneumoniae infection may be associated with overweight/obesity, independently of more traditional risk factors.
Overweight and obesity are increasing worldwide; according to World Health Organization (WHO), in 2005 1.6 billion adults were overweight and at least 400 million were obese.1 WHO further projects that by 2015, around 2.3 billion adults will be overweight and more than 700 million will be obese. With increasing overweight and obesity, also the prevalence of associated diseases, such as insulin resistance, type 2 diabetes, metabolic syndrome and cardiovascular diseases increase. Obesity is a multifactorial condition, with, for example, diet, amount of exercise, socioeconomic position, social networks and genetics having a role.2 Lately, it has been suggested that the obesity epidemic cannot be solely explained by changes in diet and lifestyle or due to genetic factors, but that infectious agents could be, at least partly, linked.3, 4 Several viruses have been suggested to associate with obesity, and in humans, associations have been reported at least for SMAM-1 avian adenovirus and human adenovirus Ad-36. Also some bacteria, especially Helicobacter pylori and Chlamydia pneumoniae, have been reported to associate with elevated body mass index (BMI), overweight, obesity or metabolic syndrome in humans, both alone and in combination with other microbes.5, 6, 7, 8, 9, 10, 11, 12 However, the causality between infections and these conditions is not currently known.
C. pneumoniae is an obligatory intracellular pathogen that causes upper and lower respiratory tract infections. Most adults have been exposured to this pathogen, and its seroprevalence increases with age due to, for example, recurring infections.13 C. pneumoniae also has a tendency to cause persistent infections,14 and these have been associated, for example, with coronary heart disease, stroke, abdominal aortic aneurysm, Alzheimer's disease, asthma and reactive arthritis.15, 16 The pathogen is capable of infecting and surviving in several different cell types, including monocytes/macrophages, vascular endothelial cells and smooth muscle cells.17, 18 The possible route of C. pneumoniae from lungs to other tissues could be through bronchoalveolar macrophages,18 and recent reports suggest that C. pneumoniae is also capable of infecting both murine and human pre-adipocytes and adipocytes.19, 20
Slightly elevated levels of serum C-reactive protein (CRP) are considered as a marker of systemic low-grade inflammation associated with obesity, insulin resistance and metabolic syndrome,21 and future cardiovascular events.21, 22, 23, 24, 25, 26
Previous studies concerning the association between C. pneumoniae infection markers and overweight/obesity have focused only on BMI and/or have had rather small sample size,5, 6, 7, 8, 10, 12, 27, 28, 29, 30, 31, 32 the results being contradictory. In our large (n=5044), well-defined population, our aim was to study whether C. pneumoniae-specific IgG antibodies are associated with elevated BMI, and whether the risk is more pronounced in those individuals with an ongoing inflammation, measured by elevated high-sensitive CRP (hsCRP) levels. We have previously reported that simultaneous presence of markers of C. pneumoniae infection and inflammation associate with pre-eclampsia requiring elective pre-term delivery,33 and with increased risk of coronary heart disease.34 Such a combination of markers may also implicate that C. pneumoniae infection is present as a persistent form and promotes low-grade inflammation. In addition, we would examine whether C. pneumoniae infection associates with distribution of fat on specific locations, by analyzing the association of infection markers with waist and hip circumference and waist–hip ratio (WHR).
Subjects and methods
We used a nested case–control sample, based on the Northern Finland Birth Cohort 1966 (NFBC1966), a general population sample that includes 12 058 live-born children, from provinces of Oulu and Lapland.35 The cohort has been prospectively followed since the prenatal period. During 1997–1998 (at 31 years of age) a postal questionnaire on health, social status and lifestyle was sent to all the 11 541 living cohort members and it reached a 76% response rate (n=8767). In addition, those living in the original target area or in the capital area were invited for a clinical examination (71% attended, n=6033). Of them, 5821 had serum C. pneumoniae antibody and hsCRP measurements. For this study, we excluded pregnant women, and persons with missing information, non-fasted blood samples or hsCRP concentration 10 mg l−1 (indicating possible acute infections). After these exclusions, the total number was 5044 persons. The study population fully represented the original cohort, when considering BMI, and distribution of sex, socioeconomic status (SES) and smoking. Permission to gather data was obtained from the Ministry of Social and Health Affairs, and the study design has been approved by and is under review of the Ethical Committee of the Northern Ostrobothnia Hospital District. At age 31, cohort members were informed about the data collected so far and were asked to sign a written consent allowing the use of the data for research purpose. Ninety-three cohort members denied use of their data and have been therefore excluded from all the analyses. The study was approved by the ethics committee of the University of Oulu.
Clinical measures at the age of 31 years and data from the postal questionnaire
Systolic and diastolic blood pressure measurements were performed by trained nurses. Two measurements (sitting position) with a mercury sphygmomanometer were taken from the right arm after 15 min rest, using a standardized procedure.36 The average of the two measurements was used for calculation of pulse pressure (the difference between systolic and diastolic blood pressure) that was used in the multivariate analysis, instead of the original blood pressure values. Weight (0.1 kg accuracy), height (0.1 cm accuracy), waist circumference (at the level midway between the lowest rib margin and the iliac crest, 0.5 cm accuracy) and hip circumference (at the widest trochanters, 0.5 cm accuracy) were measured. In order to avoid measurement errors, all the measurements were performed by two trained nurses. BMI was calculated as the ratio of weight (kg) and height squared (m2), and the WHR as a ratio of waist circumference (cm) and hip circumference (cm). Data on SES and smoking (at 31 years of age) were derived from the postal questionnaire. Smoking was originally defined as regular, occasional or non-smoking, but because of the low number of occasional smokers, they were combined with the regular smokers group. SES was originally defined as professionals, skilled workers, unskilled workers, farmers and others, but because of the low number of farmers and others, these two groups were combined. Using these combined groups had no effect on the results of the statistical analyses.
Blood samples and laboratory measurements
Blood samples were collected between 8:00 and 11:00 AM after an overnight fast. Samples for glucose, leukocyte count and serum lipids were stored at +4 °C, until they were analyzed within 24 h. Samples for serum insulin were collected, separated and stored at −20 °C by a standardized laboratory procedure and analyzed within 7 days. Samples for serum hsCRP and C. pneumoniae antibodies were stored at −80 °C until analyzed. Serum levels of the different markers were measured as follows: hsCRP by immunoenzymometric assay (Medix Biochemica, Kauniainen, Finland; sensitivity 0.08 mg l−1; intra-assay coefficient of variation 4.2%, inter-assay 5.2%), glucose by a glucose dehydrogenase method (Granutest 250; Diagnostica Merck, Darmstadt, Germany; intra-assay coefficient of variation 1.5%, inter-assay 2.3%), insulin by radioimmunoassay (Pharmacia Diagnostics, Uppsala, Sweden; sensitivity 2.4 IU l−1, intra-assay coefficient of variation 5.3%, inter-assay 7.6%), total cholesterol and triglycerides by Hitachi 911 automatic analyzer and commercial reagents (Boehringer, Mannheim, Germany), high-density lipoprotein cholesterol (HDL-C) and low-density lipoprotein cholesterol (LDL-C) were determined with the same analyzer, as described previously.37, 38 Intra- and inter-assay coefficient of variations for total cholesterol, HDL-C, LDL-C and triglycerides were 0.7/1.5%, 0.5/3.2%, 1.6/2.6% and 0.9/2.4%, respectively. Blood leukocyte count was measured by an electronic cell counter (Counter Corporation, Miami, FL, USA). C. pneumoniae-specific IgG antibodies were measured by microimmunofluorescence test, using purified elementary bodies of C. pneumoniae strain Kajaani 6 (K6) as an antigen.39 The sera were tested in serial fourfold dilutions to the end point. An IgG titer 32 was used as a positivity limit.
For comparison of the C. pneumoniae antibody-positive and negative groups, we used χ2-test and Mann–Whitney U-test. To analyze the trend of the other variables across the BMI groups, we used χ2-linear-by-linear association for categorized variables and Jonckheere–Terpstra test for continuous variables. For the trend tests, we categorized the BMI into four groups: underweight (<18.50 kg m−2), normal weight (18.50–24.99 kg m−2), overweight (25.00–29.99 kg m−2) and obese (30 kg m−2). We used analysis of variance and analysis of co-variance to estimate geometric mean values of BMI, waist and hip circumference, and WHR by C. pneumoniae antibody positivity and elevated hsCRP. For these analyses, logarithmic transformations of the continuous variables were used, except that the effect of hsCRP was studied as a categorical variable coded as normal (<1.64 mg l−1) and elevated levels (1.64 mg l−1, an upper quartile). When the joint effect of simultaneous occurrence of C. pneumoniae IgG antibodies and elevated hsCRP levels was studied, a new categorical variable was constructed to represent the different combinations of the original variables. The possible trend across these combinations was tested by analysis of variance and analysis of co-variance. The waist and hip circumference and WHR were analyzed separately for men and women due to great gender-dependent variation in them. A P-value of 0.05 (two-sided) was considered statistically significant. For statistical analysis, we used SPSS for Windows 15.0 (SPSS, Chicago, IL, USA).
Table 1 shows the characteristics of our study population by the BMI groups. All the variables studied were strongly associated with BMI. The levels of leukocytes, hsCRP, total cholesterol, LDL-C, triglycerides, glucose and insulin all increased, and the level of HDL-C decreased with an increasing BMI. In addition, systolic and diastolic blood pressure, pulse pressure, waist and hip circumference and WHR increased by the increasing BMI, and a higher prevalence of smoking and C. pneumoniae IgG antibody positivity was observed in the higher BMI groups. There was also a difference in sex distribution between the BMI groups, the proportion of men tended to increase by increasing BMI. The proportion of unskilled workers was increased in the higher BMI groups.
Table 2 shows the associations between the other variables and C. pneumoniae IgG antibody positivity. The proportion of men and smokers was higher among the antibody positives (IgG titer 32) than among the seronegatives. The proportion of unskilled workers was higher among the antibody-positive persons. Antibody positivity associated with higher levels of insulin, slightly higher levels of triglycerides and slightly lower levels of HDL-C than seronegativity. In addition, the systolic and diastolic blood pressure, pulse pressure, BMI, waist and hip circumference, and WHR were higher in the antibody-positive group than in the seronegative group. The test statistics for glucose was significant, but there was virtually no difference in glucose levels between the two groups. There was no difference in levels of hsCRP, leukocytes, total cholesterol and LDL-C.
We studied the possible effect of C. pneumoniae IgG antibody positivity (32) and elevated hsCRP levels (1.64 mg l−1, an upper quartile) on BMI. Both factors were found to associate with BMI; the C. pneumoniae IgG antibody-positive persons had a higher geometric mean BMI than the seronegative ones (Figure 1), and those with an elevated hsCRP had a higher geometric mean BMI than those with a normal hsCRP (P<0.001, data not shown). These effects remained statistically significant after adjusting for sex, smoking, SES, glucose, insulin, HDL-C, LDL-C, triglycerides, pulse pressure and leukocytes. Further, we studied the possible joint effect of C. pneumoniae and CRP on BMI. An increasing trend in the geometric mean BMI by the different groups was found (Figure 1). This effect remained statistically significant after the adjustments.
We further analyzed the possible association of C. pneumoniae IgG antibody positivity, both alone and as joint with elevated hsCRP levels, with waist and hip circumference and WHR. In men, the C. pneumoniae IgG antibody positivity was associated with a slight increase in geometric mean hip circumference, but this association did not remain statistically significant after adjusting for smoking, SES, glucose, insulin, HDL-C, LDL-C, triglycerides, pulse pressure and leukocytes (Figure 2). Elevated serum hsCRP levels were strongly associated with higher waist and hip circumference and WHR, and these associations remained statistically significant after the adjustments (For all P<0.001, data not shown). The joint effect of C. pneumoniae IgG positivity and elevated hsCRP was also strongly associated with the all three body shape measurements, with a trend for increasing waist and hip circumference by the groups (Figure 2). However, in the case of WHR and in the adjusted model for waist circumference, the increase was due to the effect of hsCRP. In the adjusted model for hip circumference, there was an increasing trend, but the difference between the two highest groups was small. In women, the results were somewhat different; C. pneumoniae antibody positivity and elevated hsCRP were associated with an increase in waist and hip circumference, both separately and as joint, and these associations remained statistically significant after the adjustments (Figure 3, For hsCRP all P<0.001, data not shown). However, for WHR an association was seen only for the elevated hsCRP and the joint effect. In the joint effect model, a trend for an increasing geometric mean WHR was found, but the difference between the two highest groups was small.
In this study, we demonstrated that C. pneumoniae IgG positivity and elevated serum hsCRP levels are associated separately and together with elevated BMI. Our results strongly support previous studies reporting an association between C. pneumoniae infection markers and elevated BMI.5, 6, 7, 8, 10, 12 However, several other studies did not find such association.27, 28, 29, 30, 31, 32 Our study included over 5000 individuals, which is to the best of our knowledge, the largest study of the topic. In addition, our study population is general population based and the subjects were relatively young, 31 years of age at the time of data collection. Thus, there is no need to consider the effect of an increasing age on C. pneumoniae seropositivity rate.13 In addition, the number of patients with, for example, cardiovascular diseases, which tend to have higher C. pneumoniae antibody levels,16 were rather small in our study. The effects of C. pneumoniae IgG positivity and elevated hsCRP levels, separately and jointly, on BMI remained statistically significant even after adjusting for factors such as LDL-C, HDL-C, triglyceride, glucose and insulin levels, leukocyte count, pulse-pressure, smoking and SES, demonstrating that the effects of these factors are independent of the more traditional risk factors. Our results are supported by recent findings showing that C. pneumoniae is capable of infecting murine pre-adipocytes and adipocytes,20 and human adipocytes derived from subcutaneous pre-adipocytes.19 Furthermore, in LDLR−/− mice repeatedly infected with C. pneumoniae, the pathogen was found in pre-adipocytes in the adipose tissue,20 demonstrating that C. pneumoniae is capable of disseminating into fat tissue, also in vivo, possibly through infected monocytes/macrophages.18 C. pneumoniae infection in adipocytes might also be able to induce accumulation of fat, as Chlamydial infections have been shown to influence accumulation and/or trafficking of lipids in host cells in vitro,40, 41, 42 and to induce liver steatosis in ApoE-deficient mice.43
The discrepancy between our results and those studies finding no association between C. pneumoniae infection markers and BMI may be due to, for example, different detection methods and cutoff limits used for C. pneumoniae antibodies, and to differences between the study populations.
In order to investigate whether C. pneumoniae IgG antibody positivity, both alone and jointly with elevated hsCRP levels, has effect on the site of fat accumulation, we studied their association with waist and hip circumference and WHR. C. pneumoniae IgG antibody positivity alone is associated with an increase in hip and waist circumference in women, and a tendency towards an increase in hip circumference was seen also in men. Elevated hsCRP alone was strongly associated with all the three measurements, in both genders, thus reflecting the strong linkage between overweight/obesity and inflammation.23 The joint effect of C. pneumoniae antibody positivity and elevated hsCRP is associated with an increase in hip and waist circumference in women. In addition, a tendency for association was found for increased hip circumference in men. No association between C. pneumoniae antibody positivity and WHR was seen. These findings imply that C. pneumoniae infection may be associated with overall distribution of fat, but not with fat localization. On the other hand, the sites where fat is primarily stored are mainly regulated by genetic factors and thus, factors like infections could be expected to have little or no effect on it. The association between C. pneumoniae infection markers and waist and hip circumference was stronger in women than in men. As the causality between infections and overweight/obesity is unknown, it is difficult to speculate the reasons behind this gender difference. The seroprevalence of C. pneumoniae is higher among men than women,10, 13 thus implying that men may be more susceptible for this pathogen and/or that their antibody response against it is more efficient. Possibly, women have less efficient clearance of C. pneumoniae infections, which may predispose them for persistent infections and for dissemination of the pathogen from lungs to other tissues.
Our finding that the geometric mean BMI and the geometric mean hip and waist circumference in women were highest in the group with both C. pneumoniae IgG positivity and an elevated hsCRP, suggests that the C. pneumoniae infection is associated with accumulation of body fat, especially in those individuals having at the same time ongoing low-grade inflammation. However, it is not clear if the type of C. pneumoniae infection associated with overweight/obesity is persistent or acute. As most of the studies, we also used a single measurement of C. pneumoniae IgG as a marker of infection.5, 6, 8, 9, 11, 12, 27, 28, 29, 30, 31 A single IgG positive serum sample may indicate recent, past or persistent infection.44 The simultaneous presence of elevated hsCRP with C. pneumoniae antibodies may, however, indicate that the infectious agent is currently present, possibly in a persistent form and maintains low-grade inflammation. On the other hand, in our study also individuals with positive C. pneumoniae IgG but with normal levels of hsCRP tended to have higher BMI, and in women also higher hip and waist circumference, than those who were IgG negative with normal hsCRP, suggesting that also a past infection could have a role.
In conclusion, we have shown in cross-section of a large, general population-based birth cohort study that individuals with C. pneumoniae IgG positivity and elevated hsCRP levels have, at 31 years of age, higher BMI and in women a higher hip and waist circumference than those who are IgG negative and/or have normal hsCRP. These findings suggest that especially persistent C. pneumoniae infection may be associated with overweight/obesity, independently of the more traditional risk factors. However, the causality between these two remains unsolved. Further studies about the effects of C. pneumoniae infection on adipocytes, and about the presence of the pathogen in the adipose tissue are needed to enlighten this question.
This study was supported by the Academy of Finland (Microbes and Man, MicMan project, Salve consortium No 129418), the Biocenter Oulu (Molecular Determinants of the Metabolic Syndrome, MetSyn consortium), the Sigrid Juselius Foundation, the Finnish Funding Agency for Technology and Innovation (TEKES).
About this article
Different associations of subscapular and triceps skinfold thicknesses with pathogen load: An ecogeographical analysis
American Journal of Human Biology (2013)