Introduction

Obesity and weight gain is widespread in many developed countries. Recently, not only overweight and obese individuals, but also persons of normal weight are trying to lose weight regardless of age or gender. However, weight loss by dietary methods is often temporary, and repeated attempts at weight reduction can result in weight cycling.¹ Intentionally or unintentionally, such variability in body weight is a common phenomenon. Therefore, concern has been raised about the possible negative effects of weight change and weight fluctuation.^2,3,4,5 Some studies^2,3,4 have reported a positive relation between weight fluctuation and an increase in all-cause mortality, whereas others⁵ have not. On the other hand, a relation between weight fluctuation and mortality from cardiovascular disease (CVD) is fairly consistent.^2,3,4 The physiologic basis for this effect is uncertain, although it has been suggested that weight fluctuation may accelerate atherogenesis.

C-reactive protein (CRP) is an acute phase reactant,⁶ which is a marker for underlying systemic inflammation. With the availability of highly sensitive assay systems, it has been possible to investigate the association between plasma CRP levels below the conventional upper limit of normal (1 mg/dl) and CVD. Several prospective studies have demonstrated that an elevated CRP level is a strong independent predictor of future myocardial infarction, stroke, peripheral atherosclerosis, and vascular death among apparently healthy men and women.^{7,8,9,10,11,12}

In this study, after first examining the distribution of circulating CRP levels among Japanese, we investigated the cross-sectional relation between CRP levels and overall adiposity (body mass index, BMI). Then, we investigated the effects of long-term weight variability on circulating CRP levels.

Subjects and methods

Subjects

A survey was conducted in 1997 among employees of N Corporation, a manufacturing company in Nagoya, Japan. The participants in the 1997 survey consisted of 764 Japanese male workers aged 40–49 y who responded to a self-reported questionnaire including medical history and lifestyle characteristics. They underwent a physical examination including height and weight measurement, and provided a collection of fasting blood samples. For the present study, we have obtained the actual weights at ages 20, 25, 30 y, and that of 5 y ago (in 1992) from the respective health checkup records. Five subjects with a medical history of cancer were excluded from the present study. Finally, the current analysis is restricted to 637 men with complete available data on all of the actual weights at ages 20, 25, 30, and 5 y ago, together with current weight, and the serum CRP concentration of the provided blood sample. As for the characteristics such as current BMI, smoking status, and CRP levels, there was no difference between these 637 men and 127 excluded men.

Informed consent for the use of personal information for analysis was given by the population of this study. The study protocol was approved by the Ethics Committee of Nagoya University Graduate School of Medicine, Nagoya, Japan.

Body mass index and weight variability indices

Weight and height were measured in the morning fasting state to the nearest 0.1 kg and 0.1 cm, respectively, with subjects wearing light clothing and no shoes at all five time points. BMI was calculated as weight in kilograms divided by the square of height in meters and used as the relative weight. According to clinical guidelines, underweight was defined as <18.5, normal range as 18.5–<25.0, overweight as 25.0–<30.0, and obese as 30.0 kg/m². Since approximately 70% of the subjects meet the criteria of normal weight, the normal category was further divided into two groups in this study: lower-normal (18.5–<22.0) and upper-normal (22.0–<25.0).

Weight variability was divided into two distinct components, trend over time and fluctuation over time, using a simple linear regression model in which each value of the subject's five actual weights (aged 20, 25, 30 y, five years ago, and current) was a dependent variable and the subject's ages at examination independent ones. The slope coefficient of this model was used to represent an individual's weight trend of direction and magnitude (weight-slope), while the root mean square error (RMSE), a standard deviation around this slope, of this model was used to represent the weight fluctuation magnitude (weight-RMSE). The coefficient of variation (CV) of weight is commonly used to describe weight fluctuation, but a person with a slight and steady weight gain over a long period of time can have the same CV as a person with large weight gains and losses with no overall weight gain. As CV and slope are correlated, one cannot discriminate between a nonlinear slope effect and the instability of changes by CV itself. In contrast, RMSE can be a more sensitive measure of instability regardless of the overall trend.

Biochemical analysis

Venous blood samples were drawn from each participant after 12 h or over fasting. The samples were stored at -80°C until assay. Serum CRP level was measured by using nephelometry, a sensitive immunoassay based on latex particle agglutination (N-assay LA CRP-S C-type, Nittobo). The assay is sensitive enough to detect 0.06–20.0 mg/dl of CRP. Undetectable CRP values were recorded as 0.05 mg/dl. CV for repeated measurements of CRP was 4.80%.

Statistical analysis

The subjects were divided into two categories based on the CRP concentration, undetectable (0.05 mg/dl) and elevated (0.06 mg/dl). Outcome variable was defined as an elevated CRP level (0.06 mg/dl) which was compared with an undetectable CRP. The relation between current BMI and CRP concentration category was examined by multiple logistic regression analysis. We calculated odds ratios (ORs) and 95% confidence intervals (CIs) for BMI as a categorical variable with lower-normal weight (BMI: 18.5–<22.0 kg/m²) as the reference category. Tests for linear trend were performed on continuous variables. Adjustments were made for potential confounders, including age and smoking status (never, former, current) (model 1). A subsequent model included covariates for age, smoking status, and other diseases associated with low-grade inflammation, including stroke, myocardial infarction, diabetes mellitus, and asthma (model 2). The relations between circulating CRP, weight-slope, and weight-RMSE were examined in the same way. Using multiple logistic regression, the ORs of CRP elevation by the quartiles of weight-slope and weight-RMSE were estimated in three models: model 1, model 2, and model 3 as model 2+current BMI. The reference groups were defined in principle as the subjects in the lowest quartiles of weight-slope and weight-RMSE. Furthermore, to compare the difference and similarity of the effects of weight-slope and weight-RMSE in BMI categories, ORs stratified below and above BMI 25 kg/m² were calculated. All tests were conducted using LOGISTIC procedures in SPSS software, version 10.0 for Windows.¹³

Results

The characteristics of the subjects are shown in Table 1. Concerning BMI distribution, 71.4% of the subjects were classified as normal weight and those with BMI greater than 30 kg/m² accounted for only 1.9% in this study population. This finding is consistent with those in other studies on Japanese.¹⁴ The distribution of CRP levels was highly skewed to lower levels. The minimum value was less than 0.05 mg/dl, the maximum value was 5.31 mg/dl, with 25th, 50th, and 75th percentile values of less than 0.05, less than 0.05, and 0.09 mg/dl, respectively. A total of 52.9% of the subjects had undetectable CRP levels. The elevated CRP levels of 0.06–0.09, 0.10–0.99, and 1.00 mg/dl or more were present in 24.5, 21.7, and 0.9%, respectively. A total of 47.1% of the subjects had the elevated CRP levels. CRP levels were higher among men aged 45–49 y (geometric mean: 0.077 mg/dl) compared with men aged 40–44 y (geometric mean: 0.070 mg/dl). Smokers (geometric mean: 0.077 mg/dl) had higher CRP concentration than nonsmokers (geometric mean: 0.069 mg/dl). As for CRP levels in relation to disease, CRP levels were higher among men with a history of stroke (geometric mean: 0.135 mg/dl), myocardial infarction (geometric mean: 0.161 mg/dl), diabetes mellitus (geometric mean: 0.099 mg/dl), and asthma (geometric mean: 0.081 mg/dl) compared with men without a history of stroke (geometric mean: 0.074 mg/dl), myocardial infarction (geometric mean: 0.074 mg/dl), diabetes mellitus (geometric mean: 0.073 mg/dl), and asthma (geometric mean: 0.074 mg/dl), respectively.

Table 1 - Characteristics of study population.

Full table

The prevalence and adjusted ORs for elevated CRP by current BMI category are shown in Table 2. The prevalence of elevated CRP level increased with increasing BMI. Overweight and obese men were 2.71 and 10.58 times more likely to have mildly elevated CRP levels compared with lower normal-weight persons after adjusting for age and smoking (model 1). The test for trend by a comparison of adjacent categories was significant (P<0.001). After adjusting for age, smoking, stroke, myocardial infarction, diabetes mellitus, and asthma (model 2), the association of elevated CRP level with current BMI was unchanged.

Table 2 - Adjusted odds ratios (95% confidence intervals) for elevated CRP concentrations (0.06 mg/dl) by current BMI category.

Full table

Adjusted ORs for elevated CRP levels by quartile of weight-slope are presented in Table 3. In the overall subjects, the ORs for elevated CRP level had a monotonic increase with increasing weight-slope in models 1 and 2. Adjusting for age, smoking status, and other diseases associated with low-grade inflammation (model 2), the ORs for elevated CRP level were 1.63 (95% CI: 1.03–2.59), 1.73 (95% CI: 1.08–2.76), and 4.15 (95% CI: 2.57–6.71) among subjects with the second, third, and highest quartiles of weight-slope, respectively, compared to those with the lowest quartile. On further adjustment for current BMI (model 3), these ORs were attenuated, but the OR was 2.38 (95% CI: 1.28–4.43) for the highest vs the lowest quartile (P-value for linear trend <0.05). Data were dichotomized at BMI 25 kg/m², defined as overweight, since the effect of weight-slope may vary with current BMI. The higher ORs were more pronounced among subjects with BMI 25 kg/m² than in those with BMI <25 kg/m². On adjustment for current BMI (model 3), the association of elevated CRP level with weight-slope was abolished among the subjects with BMI <25 kg/m², whereas a significant increasing trend of ORs was still observed among those with BMI 25 kg/m².When the analyses were repeated after exclusion of smokers and persons with a medical history or under treatment for stroke, myocardial infarction, diabetes mellitus, or asthma, similar results were obtained.

Table 3 - Adjusted odds ratios (95% confidence intervals) for elevated CRP concentrations (0.06 mg/dl) by quartile of weight-slope stratified below and above BMI 25 kg/m² .

Full table

Table 4 shows adjusted ORs for elevated CRP levels by quartile of weight-RMSE. In the overall data set, the ORs of men whose weight had changed at the highest quartile of weight-RMSE were 2.16 (1.37–3.39) and 2.05 (1.30–3.24) in models 1 and 2, respectively (P-value for linear trend <0.001), when the lowest quartile group was referenced. On further adjustment for current BMI (model 3), this OR was attenuated to 1.50 (95% CI: 0.93–2.43), but a significant increasing trend was still found (P-value for linear trend <0.05). When data were dichotomized at BMI 25 kg/m², the ORs of elevated CRP level increased across quartile of weight-RMSE among subjects with BMI <25 kg/m² in all models (P-value for linear trend <0.01, <0.01, and <0.05 in model 1, 2, and 3, respectively), whereas there was no associations between elevated CRP level and weight-RMSE among those with BMI 25 kg/m².

Table 4 - Adjusted odds ratios (95% confidence intervals) for elevated CRP concentrations (0.06 mg/dl) by quartile of weight-RMSE stratified below and above BMI 25 kg/m².

Full table

After exclusion of smokers and persons with a medical history or under treatment of stroke, myocardial infarction, diabetes mellitus, or asthma, the results remained the same.

We also defined elevated CRP level as CRP concentration 0.10 mg/dl and repeated the same analysis. The associations of elevated CRP level with current BMI, weight-slope, and weight-RMSE were much the same.

Discussion

To the best of our knowledge, this is the first population-based study that reports the effect of long-term weight variability on serum CRP levels. Most studies¹⁵ concerning the effects of longitudinal weight changes used self-reported body weight by questionnaire and were not free from information bias. The advantage of our study is that we used actual measurements of weight variables.

The results of the present study have several implications. First, our study found a positive association between CRP level and current BMI, although 71.4% of the subjects were classified as having normal- weight. Previous studies^7,12 in middle-aged and elderly persons have reported a higher prevalence of low-grade inflammation in overweight and obese persons compared with normal-weight persons. However, in these generations, the association between the CRP level and BMI may have been confounded by diseases such as diabetes mellitus and CVD, which were associated with both obesity and elevated CRP concentrations.^16,17 We carefully controlled for inflammatory disease and other factors known to influence CRP concentrations.¹⁸ Furthermore, we dichotomized the normal-weight group, since the percentage of normal-weight persons is high in this study population. Overweight and obese persons had a higher prevalence of low-grade inflammation compared with lower normal-weight persons. Of interest is our observation that the persons in the upper normal-weight category were 1.83 times more likely to have elevated CRP levels, respectively, compared with those in the lower normal-weight category even among those in the normal-weight range.

We also found that the ORs of elevated CRP levels increased with increasing weight-slope independent of current BMI. In other words, the ORs of elevated CRP levels decreased with decreasing weight-slope. This independent effect appeared to be greater among men with BMI 25 kg/m². This finding possibly suggests that weight gain may induce an increase in CRP levels, while weight loss may induce a reduction in them. Recent studies by Heilbronn et al¹⁹ and Tchernof et al²⁰ demonstrated that weight loss by dietary intervention induced a significant reduction in the circulating CRP level. Most importantly, we also found that the ORs of elevated CRP levels increased with increasing weight-RMSE. Weight fluctuation was related to the elevated CRP independent of current BMI. After data were dichotomized at BMI 25 kg/m², the association of elevated CRP with weight fluctuation was observed only among men with BMI <25 kg/m². This is consistent with other studies,²¹ where adverse effects of weight change on health were seen more often among those who were not overweight, than among those who were overweight.

Approximately 25% of circulating IL-6 is estimated to be released by human subcutaneous adipose tissue in vivo,²² and IL-6 stimulates the production of CRP in the liver.²³ This might explain the observed associations between current BMI and CRP. Some studies²⁴ have reported that those who gained weight increased their waist-to-hip circumference ratio or abdominal fat mass and those who reduced weight decreased them. In vitro, human abdominal visceral adipose tissue releases more IL-6 compared to subcutaneous adipose tissue,²⁵ possibly explaining why the weight gain or loss (weight-slope) was independently associated with elevated CRP level after adjustment for current BMI and several confounders. The theory that weight fluctuation may lead to an unfavorable distribution of body fat is controversial.^26,27 Metabolic factors besides increased body fatness may underlie the association between weight fluctuation and elevated CRP.

It is not yet clear how weight fluctuation increases the risk of morbidity and mortality from coronary heart disease. This may be explained in part by the positive association between weight fluctuation and elevated CRP. CRP has been reported to be an independent risk factor for CVD. Recently, it has been proposed that CRP function not only reflects the extent and inflammatory activity of atherosclerotic lesions themselves, or inflammation elsewhere in the body or both, but this acute phase protein is also directly involved in the pathogenesis of atherothrombosis, through several mechanisms such as activation of complement,²⁸ induction tissue factor,²⁹ and binding of plasma lipoprotein.²⁸

Several limitations of this study deserve mention. First, our study lacked the information on reasons for the weight changes. We were unable to distinguish intentional from unintentional weight loss. Secondly, we compared the circulating CRP level with current BMI and past body weight change. We did not measure CRP levels in the past and thus could not evaluate the effect of weight change on the change in CRP levels. The directionality of associations cannot be conclusively established because the study is cross-sectional. Third, the subjects in this study comprised apparently middle-aged Japanese men, and thus the results may not apply to women or Westerners with a high BMI.

In conclusion, we have shown that a higher BMI is associated with higher CRP concentrations that could not be explained by inflammatory disease or other factors or diseases known to increase CRP concentrations. These data suggest that a state of low-grade systemic inflammation is present in the overweight and obese. The results of this cross-sectional study have also shown that long-term weight gain is associated positively with circulating CRP, especially among overweight and obese men, and that longterm weight fluctuation is associated positively with circulating CRP, especially among normal-weight men. Both associations were independent of current BMI and other confounders. Being overweight and obese pose a considerable cardiovascular health risk, while a number of cardiovascular events occur among the normal-weight population. Since CRP has been shown to be one of the most powerful predictors of risk of cardiovascular events, several findings in this study suggest that not only avoiding weight gain, but also maintaining a stable weight even in the normal-weight range from the adolescence may help to prevent CVD.

References

1. Kramer FM, Jeffery RW, Forster JL, Snell MK. Long-term follow-up of behavioral treatment for obesity: patterns of weight regain among men and women. Int J Obes Relat Metab Disord 1989; 13: 123–136. | ChemPort |
2. Iribarren C, Sharp DS, Burchfiel CM, Petrovitch H. Association of weight loss and weight fluctuation with mortality among Japanese American men. N Engl J Med 1995; 333: 686–692. | Article | PubMed |
3. Lissner L, Odell PM, D'Agostino RB, Stokes III J, Kreger BE, Belanger AJ, Brownell KD. Variability of body weight and health outcomes in the Framingham population. N Engl J Med 1991; 324: 1839–1844. | PubMed | ISI | ChemPort |
4. Peters ET, Seidell JC, Menotti A, Arayanis C, Dontas A, Fidanza F, Karvonen M, Nedeljkovic S, Nissinen A, Buzina R. Changes in body weight in relation to mortality in 6441 European middle-aged men: the Seven Countries Study. Int J Obes Relat Metab Disord 1995; 19: 862–868. | PubMed | ChemPort |
5. Hamm P, Shekelle RB, Stamler J. Large fluctuations in body weight during young adulthood and twenty-five-year risk of coronary death in men. Am J Epidemiol 1989; 129: 312–318. | PubMed | ISI | ChemPort |
6. Pepys MB. C-reactive protein fifty years on. Lancet 1981; 8221: 653–657. | Article |
7. Tracy RP, Lemaitre RN, Psaty BM, Ives DG, Evans RW, Cushman M, Meilahn EN, Kuller LH. 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: 1121–1127. | PubMed | ISI | ChemPort |
8. Ridker PM, Cushman M, Stampfer MJ, Tracy RP, Hennekens CH. Inflammation, aspirin, and the risk of cardiovascular disease in apparently healthy men. N Engl J Med 1997; 336: 973–979. | Article | PubMed | ISI | ChemPort |
9. Ridker PM, Cushman M, Stampfer MJ, Tracy RP, Hennekens CH. Plasma concentration of C-reactive protein and risk of developing peripheral vascular disease. Circulation 1998; 97: 425–428. | PubMed | ISI | ChemPort |
10. Ridker PM, Buring JE, Shih J, Matias M, Hennekens CH. Prospective study of C-reactive protein and the risk of future cardiovascular events among apparently healthy women. Circulation 1998; 98: 731–733. | PubMed | ISI | ChemPort |
11. Roivainen M, Viik-Kajander M, Palosuo T, Toivanen P, Leinonen M, Saikku P, Tenkanen L, Manninen V, Hovi T, Manttari M. Infections, inflammation, and the risk of coronary heart disease. Circulation 2000; 101: 252–257.
12. Koenig W, Sund M, Frohlich M, Fischer HG, Lowel H, Doring A, Hutchinson WL, Pepys MB. 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) Augsberg Cohort Study, 1984 to 1992. Circulation 1999; 99: 237–242. | PubMed | ISI | ChemPort |
13. SPSS Incorporation. SPSS Base 10.0J user's guide. SPSS INC.: Chicago, IL; 1999.
14. Yoshiike N, Matsumura Y, Zaman MM, Yamaguchi M. Descriptive epidemiology of body mass index in Japanese adults in a representative sample from the National Nutrition Survey 1990–1994. Int J Obes Relat Metab Disord 1998; 22: 684–687. | Article | PubMed |
15. Trentham-Dietz A, Newcomb PA, Egan KM, Titus-Ernstoff L, Baron JA, Storer BE, Stampfer M, Willett WC. Weight change and risk of postmenopausal breast cancer (United States). Cancer Causes Control 2000; 11: 533–542. | Article | PubMed | ISI | ChemPort |
16. Cassano PA, Rosner B, Vokonas PS, Weiss ST. Obesity and body fat distribution in relation to the incidence of non-insulin-dependent diabetes mellitus. A prospective cohort study of men in the normative aging study. Am J Epidemiol 1992; 136: 1474–1486. | PubMed |
17. Rimm EB, Stampfer MJ, Giovannucci E, Ascherio A, Spiegelman D, Colditz GA, Willett WC. Body size and fat distribution as predictors of coronary heart disease among middle-aged and older US men. Am J Epidemiol 1995; 141: 1117–1127. | PubMed | ISI | ChemPort |
18. Mendall MA, Patel P, Ballam L, Strachan D, Northfield TC. C reactive protein and its relation to cardiovascular risk factors: a population based cross sectional study. BMJ 1996; 312: 1061–1065. | PubMed | ChemPort |
19. Heilbronn LK, Noakes M, Clifton PM. Energy restriction and weight loss on very-low-fat diets reduce C-reactive protein concentrations in obese, healthy women. Arterioscler Thromb Vasc Biol 2001; 21: 968–970. | PubMed | ISI | ChemPort |
20. Tchernof A, Nolan A, Sites CK, Ades PA, Poehlman ET. Weight loss reduces C-reactive protein levels in obese postmenopausal women. Circulation 2002; 105: 564–569. | Article | PubMed |
21. Harris TB, Ballard-Barbasch R, Madans J, Makuc DM, Feldman JJ. Overweight, weight loss, and risk of coronary heart disease in older women. The NHANES I Epidemiologic Follow-up Study. Am J Epidemiol 1993; 137: 1318–1327. | PubMed |
22. Mohamed-Ali V, Pinkney JH, Coppack SW. Adipose tissue as an endocrine and paracrine organ. Int J Obes Relat Metab Disord 1998; 22: 1145–1158. | Article | PubMed | ChemPort |
23. Papanicolaou DA, Wilder RL, Manolagas SC, Chrousos GP. The pathophysiologic roles of interleukin-6 in human disease. Ann Intern Med 1998; 128: 127–137. | PubMed |
24. Pihl E, Jurimae T. Relationships between body weight change and cardiovascular disease risk factors in male former athletes. Int J Obes Relat Metab Disord 2001; 25: 1057–1062. | Article | PubMed | ChemPort |
25. Fried SK, Bunkin DA, Greenberg AS. Omental and subcutaneous adipose tissues of obese subjects release interleukin-6. J Clin Endocrinol Metab 1998; 83: 847–850. | Article | PubMed | ISI | ChemPort |
26. van der Kooy K, Leenen R, Seidell JC, Deurenberg P, Hautvast JG. Effect of a weight cycle on visceral fat accumulation.Am J Clin Nutr 1993; 58: 853–857. | PubMed | ChemPort |
27. Jebb SA, Goldberg GR, Coward WA, Murgatroyd PR, Prentice AM. Effects of weight cycling caused by intermittent dieting on metabolic rate and body composition in obese women. Int J Obes Relat Metab Disord 1991; 15: 367–374. | ChemPort |
28. Bhakdi S, Torzewski M, Klouche M, Hemmes M. Complement and atherogenesis: binding of CRP to degraded, nonoxidized LDL enhances complement activation. Arterioscler Thromb Vasc Biol 1999; 19: 2348–2354. | PubMed | ISI | ChemPort |
29. Cermak J, Key NS, Bach RR, Balla J, Jacob HS, Vercellotti GM. C-reactive protein induces human peripheral blood monocytes to synthesize tissue factor. Blood 1993; 82: 513–520. | PubMed | ISI | ChemPort |

Acknowledgements

The authors are grateful to Dr Noboru Okamoto (Aichi San-no-maru Hospital), Dr Takashi Kawamura (Kyoto University Center for Student Health), and Dr Junji Toyama (Aichi Prefectural Owari Hospital) for their cooperation in conducting the survey and collecting information for this study.

This work is supported in part by a grant to Hideaki Toyoshima (09470112, 13470087), Koji Tamakoshi (12670352), and Hiroshi Yatsuya (13770192) from the Ministry of Education, Culture, Sports, Science and Technology.