Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

Endotoxins are associated with visceral fat mass in type 1 diabetes


Bacterial lipopolysaccharides (LPS), potent inducers of inflammation, have been associated with chronic metabolic disturbances. Obesity is linked to dyslipidemia, increased body adiposity, and endotoxemia. We investigated the cross-sectional relationships between serum LPS activity and body adiposity as well as inflammation in 242 subjects with type 1 diabetes. Body fat distribution was measured by DXA and serum LPS activity by the limulus amebocyte lysate end-point assay. Since no interaction between visceral fat mass and sex was observed, data were pooled for the subsequent analyses. LPS was independently associated with visceral fat mass, when adjusted for traditional risk factors (age, sex, kidney status, hsCRP, insulin sensitivity). In the multivariate analysis, serum LPS activity and triglyceride concentrations had a joint effect on visceral fat mass, independent of these factors alone. A combination of high LPS and high hsCRP concentrations was also observed in those with the largest visceral fat mass. In conclusion, high serum LPS activity levels were associated with visceral fat mass in subjects with type 1 diabetes strengthening its role in the development of central obesity, inflammation and insulin resistance.


Bacterial endotoxins or lipopolysaccharides (LPS) are potent inducers of systemic inflammation. Endotoxin transport and clearance has been linked to lipoprotein metabolism. Endotoxins are redistributed towards VLDL and LDL particles especially in conditions of lower HDL cholesterol concentrations, like in atherosclerosis and insulin resistance1,2. Toll-like receptor 4 (TLR4) is a pattern recognition receptor that binds LPS and initiates an inflammatory response. Compared to lean individuals, TLR4 expression is up-regulated in the adipose tissue of obese subjects3. A low dose LPS infusion augments adiposity, inflammation, insulin resistance, hyperglycemia and dyslipidemia in mice4. In humans, a low dose LPS infusion induces insulin resistance and potentiates the expression of inflammatory markers in adipose tissue5.

Studies in non-diabetic children and obese adults, have shown an adverse association between endotoxin levels, central adiposity and insulin resistance6,7. We have previously shown that a high serum LPS activity is associated with features of the metabolic syndrome and the progression of kidney disease in type 1 diabetes8,9. Recent studies have further demonstrated a putative link between endotoxins and visceral fat10. Excess visceral fat is associated with impaired insulin sensitivity, and the development of vascular complications in diabetes11. It has also been suggested that the distribution of body fat could better predict insulin resistance than BMI itself 7,12. Evidently, certain human genetic variants may also have a significant impact on BMI and body adiposity13.

Our aim was to investigate the association between serum LPS activity and visceral fat mass, measured by dual-energy x-ray absorptiometry (DXA), in subjects with type 1 diabetes.

Subjects and Methods

Data from individuals with type 1 diabetes were collected during the Finnish Diabetic Nephropathy (FinnDiane) Study visits taking place in Helsinki between years 2011 and 2016. In the current analyses, we included data from all individuals with completed LPS and body composition measurements, and either normal urinary albumin excretion rate (T1D-normo, n = 156), microalbuminuria (T1D-micro, n = 42) or macroalbuminuria (T1D-macro, n = 46). Renal status was assessed by the albumin excretion rate (AER) in two out of three consecutive timed urine collections using the following criteria: normal AER <20 μg/min or <30 mg/24 h, and macroalbuminuria ≥200 μg/min or ≥300 mg/24 h. Kidney transplant recipients and those undergoing dialysis were excluded from the analysis. Type 1 diabetes was defined as disease onset before the age of 40 and the initiation of insulin injections within one year of diagnosis.

Body composition was measured by DXA (Lunar iDXA by GE Healthcare, Scanex Medical Systems, Finland). The iDXA software calculates the body fat distribution including visceral-, android-, and gynoid fat mass. BMI was calculated (weight/height2). Insulin sensitivity was assessed calculating the estimated glucose disposal rate (eGDR): eGDR = 24.4 − 12.97 × WHR − 3.39 × AHT − 0.60 × A1C, where WHR stands for waist-to-hip ratio and AHT for antihypertensive treatment and/or blood pressure ≥140/90 mmHg (yes = 1, no = 0). Daily insulin dose was self-reported. Serum lipid concentrations were analyzed centrally by automated enzymatic methods (Hoffmann-La Roche, Basel, Switzerland). Serum high-sensitive C-reactive protein (hsCRP) concentration was measured by immunoassay (Modular analyzer, Roche). LPS activity was measured by the Limulus amebocyte lysate (LAL) assay from 1:5 diluted fasting serum samples (Hycult Biotechnology, Uden, the Netherlands). For the LAL assay, inter- and intra-assay coefficients of variation were 16.1% and 4.5%, respectively. The sensitivity limit for the assay was 0.02 EU/ml.

Statistical analyses

Normality of variable distribution was assessed with the Kolmogorov-Smirnov test. Normally distributed variables are reported as mean ± standard deviation, non-normally distributed variables as median [25th–75th quartile]. Correlation coefficients were calculated by Spearman or Pearson’s correlation test as appropriate. Differences in frequencies were assessed by Pearson’s Chi squared analysis. Differences in variation between groups were calculated using Mann-Whitney U-test, Kruskal-Wallis test and ANOVA, as appropriate. The population was divided into tertiles based on serum LPS activity, triglyceride-, or hsCRP concentrations to further explore the role of endotoxins, dyslipidemia, and inflammation on visceral fat mass.

The interaction between LPS and gender on visceral fat mass was assessed using a generalized linear model. Since no interaction was evident, all data were analyzed together. The interaction between LPS and triglyceride concentration on visceral fat mass was analyzed similarly. As LPS and triglyceride concentration showed significant interaction, the interaction term was included in the multivariable model. Because LPS and triglycerides are highly correlated, we furthermore visualized their relative relationship using the generalized additive modeling (GAM) without a priori assumptions of the shape of the relation. The GAM modelling allows the inclusion of non-parametric smoothing functions to identify a potential non-linearity in the relationship between the independent and the dependent variables14,15. The generalized cross-validation function (GCV) was used as a criterion for selection of the smoothing parameters to determine an appropriate level of smoothing. Analyses were carried out using IBM SPSS Statistics for Windows, Version 22.0 (IBM Corp, Armonk, NY, USA), and R open source software ( GAM models were fitted using the mgcv library in R16.


The study protocol was approved by the Ethics Committee of the Hospital District of Helsinki and Uusimaa. The study was carried out in accordance with the approved guidelines. Participants gave their written informed consent prior to participating in the study.


Patient characteristics divided by renal status are shown in Table 1. The proportion of men was highest among T1D-macro. The diabetes duration and visceral fat mass increased with worsening renal status. Insulin sensitivity (eGDR) was highest in T1D-normo. No differences were observed in HbA1c, HDL cholesterol concentration, LPS activity, insulin dose, use of insulin pump, BMI or hsCRP concentration amongst the three groups.

Table 1 Clinical characteristics of study participants divided by renal status.

Serum LPS activity correlated positively with BMI (r = 0.274, p < 0.001), visceral fat mass (r = 0.248, p < 0.001), android to gynoid fat ratio (r = 0.231, p < 0.001), serum triglyceride concentration (r = 0.477, p < 0.001), and hsCRP concentration (r = 0.276, p < 0.001). Negative correlation was observed between insulin dose and eGDR (r = −0.205, p = 0.002). Supplementary Figure S1a and S1b show that triglyceride concentration seems to overrun the effect of LPS on the visceral fat mass. However, the relationship between LPS and triglycerides is complicated and interaction between LPS and triglyceride concentration was found, indicating that LPS has an effect on certain levels of triglyceride concentration. Therefore, the effect of serum LPS activity and triglyceride concentration on visceral fat mass were further explored in LPS and triglyceride tertiles (low/medium/high). Subjects in the high triglyceride tertile presented the highest visceral fat mass. Notably, within each LPS tertile, the visceral fat mass was observed to increase towards the higher TG tertiles (p < 0.001, all) (Fig. 1A). Moreover, the visceral fat mass among those with high LPS activity and high triglyceride level was higher compared to the group with both low LPS activity and low triglyceride level (high/high vs. low/low, median [IQR]: 1452 [767–2001] vs. 343 [144–515] g, p < 0.001). The Supplementary Figure S2 provides a deeper view on the interaction effect of LPS and triglyceride levels on the visceral fat mass focusing on the 75% of the population in the area with the densest accumulation of observations. Similarly, subjects in the high-LPS/high-hsCRP tertile presented greater visceral fat mass compared to those with low-LPS/low-hsCRP concentrations (median [IQR]: 1452 [813–2014] vs. 379 [156–818] g, p < 0.001) (Fig. 1B).

Figure 1

(A) Visceral fat mass (g) according to LPS and triglyceride tertiles (low/medium/high) in subjects with type 1 diabetes. Higher visceral fat mass was observed in those with a combination of high LPS activity/high triglyceride concentration (N = 50) compared to those with low LPS activity/low triglyceride concentration (N = 43, p < 0.001). (B) Visceral fat mass (g) according to LPS and hsCRP tertiles (low/medium/high) in subjects with type 1 diabetes. Higher visceral fat mass was observed in those with a combination of high LPS activity/high hsCRP concentration (N = 32) compared to low LPS activity/low hsCRP concentration (N = 38, p < 0.001).

The independent association between LPS activity and visceral fat mass was analyzed in a number of multivariable models (Table 2). With Model 1, we aimed at looking at this association while taking inflammation into account. In this model, LPS activity was positively associated with the visceral fat mass (B = 1253.2, 95% Wald Confidence Interval 386.4–2120.0, Wald Chi-Squared 8.029, p = 0.005). Further adjusting for insulin sensitivity (eGDR), LPS remained a significant predictor of visceral fat mass (Model 2: B = 848.6, 95% Wald CI 48.4–1648.8, Wald Chi Squared 4.320, p = 0.038). However, after further adjusting for triglyceride concentration (Model 3), LPS no longer predicted visceral fat mass. To the final model (Model 4) we incorporated a LPS*triglyceride interaction term, which was significantly associated with the visceral fat mass (B = −634.2, 95% Wald CI −1092.6–−175.7, Wald Chi-Squared 7.350, p = 0.007), independently of the two factors entered in the model separately.

Table 2 Factors associated with visceral fat mass.


Inflammation and insulin resistance are intertwined processes that are strongly linked to overweight and intra-abdominal obesity. We observed that serum LPS activity is associated with visceral fat mass in type 1 diabetes, independent of the traditional risk factors such as age, insulin sensitivity, and inflammation.

High serum LPS activity has been associated with central obesity and insulin resistance, and it has also been shown to predict the development of type 2 diabetes6,17. Intestinal and oral microbial communities are the most likely sources of bacterial endotoxins in humans. Frequent use of antibiotics may have adverse effects on gut homeostasis accompanied by dysbiosis and increased intestinal permeability of bacterial compounds. Childhood hospitalizations related to infections have been associated with an increased risk of obesity and metabolic syndrome in adulthood18. In conditions of intestinal inflammation, such as Crohn’s disease, visceral fat accumulation is evident19.

Dietary fat intake may increase systemic LPS activity, driving adipose inflammation20. In a recent study, circulating LPS activity levels correlated with intra-abdominal fat volumes, but only modestly with the subcutaneous fat volumes in patients undergoing bariatric surgery. Notably, a significant decrease in serum endotoxin levels was still evident one year after the operation21. Higher CRP levels may arise through a local inflammation of the adipose tissue associated with an increased bacterial burden19.

Serum triglycerides and LPS are strongly correlated9. Indeed, we observed an interaction between LPS and triglycerides on visceral fat mass. Importantly, the interaction term was independent of the individual factors alone, indicating that the combination of high serum LPS and triglycerides have additive effects on visceral fat mass. This observation is in concordance with previous rat studies, showing that a low dose intraperitoneal LPS injection increases liver triglyceride production, while high doses decrease triglyceride-rich lipoprotein clearance22. Moreover, in healthy volunteers, a small intravenous dose of LPS acutely increased serum triglyceride levels23, while in mice a subcutaneous infusion of LPS for four weeks significantly increased visceral fat mass4. Cross-sectional nature and small study population are potential limitations of the current study. Moreover, due to the exclusion of individuals with end-stage renal disease, the results may not be generalizable to these individuals. Prospective studies are needed to establish the causalities related to the observations.

In conclusion, independent of traditional risk factors, serum LPS activity is associated with visceral fat mass in type 1 diabetes. The LPS-association with visceral fat is pronounced at high triglyceride concentrations. This strengthens the hypothesis that both bacterial endotoxins and dyslipidemia contribute to the development of central obesity, inflammation and insulin resistance and thus may subsequently be associated with adverse cardio-metabolic outcomes.

Additional Information

How to cite this article: Lassenius, M. I. et al. Endotoxins are associated with visceral fat mass in type 1 diabetes. Sci. Rep. 6, 38887; doi: 10.1038/srep38887 (2016).

Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.


  1. Feingold, K. R. & Grunfeld, C. The role of HDL in innate immunity. J. Lipid Res. 52, 1–3 (2011).

    CAS  Article  Google Scholar 

  2. Vreugdenhil, A. C., Snoek, A. M., van’t Veer, C., Greve, J. W. & Buurman, W. A. LPS-binding protein circulates in association with apoB-containing lipoproteins and enhances endotoxin-LDL/VLDL interaction. J. Clin. Invest. 107, 225–234 (2001).

    CAS  Article  Google Scholar 

  3. Boutagy, N. E., McMillan, R. P., Frisard, M. I. & Hulver, M. W. Metabolic endotoxemia with obesity: Is it real and is it relevant? Biochimie 124, 11–20 (2016).

    CAS  Article  Google Scholar 

  4. Cani, P. D. et al. Metabolic endotoxemia initiates obesity and insulin resistance. Diabetes 56, 1761–1772 (2007).

    CAS  Article  Google Scholar 

  5. Mehta, N. N. et al. Experimental endotoxemia induces adipose inflammation and insulin resistance in humans. Diabetes 59, 172–181 (2009).

    Article  Google Scholar 

  6. Moreira, A. P. et al. Higher plasma lipopolysaccharide concentrations are associated with less favorable phenotype in overweight/obese men. Eur. J. Nutr. 54, 1363–1370 (2014).

    Article  Google Scholar 

  7. Samsell, L., Regier, M., Walton, C. & Cottrell, L. Importance of android/gynoid fat ratio in predicting metabolic and cardiovascular disease risk in normal weight as well as overweight and obese children. J. Obes. 846578 (2014).

  8. Lassenius, M. I. et al. Bacterial endotoxin activity in human serum is associated with dyslipidemia, insulin resistance, obesity, and chronic inflammation. Diabetes Care 34, 1809–1815 (2011).

    CAS  Article  Google Scholar 

  9. Nymark, M. et al. Serum lipopolysaccharide activity is associated with the progression of kidney disease in Finnish patients with type 1 diabetes. Diabetes Care 32, 1689–1693 (2009).

    CAS  Article  Google Scholar 

  10. Lam, Y. Y. et al. Role of the gut in visceral fat inflammation and metabolic disorders. Obesity 19, 2113–2120 (2011).

    CAS  Article  Google Scholar 

  11. Gustafson, B. & Smith, U. Regulation of white adipogenesis and its relation to ectopic fat accumulation and cardiovascular risk. Atherosclerosis 241, 27–35 (2015).

    CAS  Article  Google Scholar 

  12. Peterson, M. D., Al Snih, S., Serra-Rexach, J. A. & Burant, C. Android adiposity and lack of moderate and vigorous physical activity are associated with insulin resistance and diabetes in aging adults. J. Gerontol. A. Biol. Sci. Med. Sci. 70, 1009–1017 (2015).

    CAS  Article  Google Scholar 

  13. Lu, Y. et al. New loci for body fat percentage reveal link between adiposity and cardiometabolic disease risk. Nat. Commun. 7, 10495 (2016).

    CAS  ADS  Article  Google Scholar 

  14. Hastie, T. & Tibshirani, R. Exploring the nature of covariate effects in the propotional hazards model. Biometrics 46, 1005–1016 (1990).

    CAS  Article  Google Scholar 

  15. Wood, S. N. Generalized additive models: An introduction with R. (Chapman & Hall/CRC, Boca Raton, Florida, 2006).

  16. Wood, S. Mixed GAM computation vehicle with GVC/AIC/REML smoothness estimation. R package version 1. 8–9 (2015).

  17. Pussinen, P. J., Havulinna, A. S., Lehto, M., Sundvall, J. & Salomaa, V. Endotoxemia is associated with an increased risk of incident diabetes. Diabetes Care 34, 392–397 (2011).

    CAS  Article  Google Scholar 

  18. Burgner, D. P. et al. Infection-related hospitalization in childhood and adult metabolic outcomes. Pediatrics 136, e554–562 (2015).

    Article  Google Scholar 

  19. Peyrin-Biroulet, L. et al. Mesenteric fat as a source of C reactive protein and as a target for bacterial translocation in Crohn’s disease. Gut 61, 78–85 (2012).

    CAS  Article  Google Scholar 

  20. Caesar, R., Tremaroli, V., Kovatcheva-Datchary, P., Cani, P. D. & Bäckhed, F. Crosstalk between gut microbiota and dietary lipids aggravates WAT inflammation through TLR signaling. Cell Metab. 22, 658–668 (2015).

    CAS  Article  Google Scholar 

  21. Troseid, M. et al. Plasma lipopolysaccharide is closely associated with glycemic control and abdominal obesity: evidence from bariatric surgery. Diabetes Care 36, 3627–3632 (2013).

    Article  Google Scholar 

  22. Feingold, K. R. et al. Endotoxin rapidly induces changes in lipid metabolism that produce hypertriglyceridemia: low doses stimulate hepatic triglyceride production while high doses inhibit clearance. J. Lipid Res. 33, 1765–1776 (1992).

    CAS  PubMed  Google Scholar 

  23. Hudgins, L. C. et al. A single intravenous dose of endotoxin rapidly alters serum lipoproteins and lipid transfer proteins in normal volunteers. J. Lipid Res. 44, 1489–1498 (2003).

    CAS  Article  Google Scholar 

Download references


This study was supported by grants from the Signe and Ane Gyllenberg Foundation, Folkhälsan Research Foundation, the Wilhelm and Else Stockmann Foundation, the Liv och Hälsa Foundation, the Helsinki University Central Hospital Research Funds (EVO), the Finnish Cultural Foundation, the Diabetes Research Foundation, the Novo Nordisk Foundation (NNF14SA0003), Academy of Finland, Tekes and the Finnish Medical Society (Finska Läkaresällskapet). The skilled technical assistance of Anna Sandelin, and Jaana Tuomikangas is gratefully acknowledged. The authors also acknowledge all the physicians and nurses at each centre participating in the collection of patients (online appendix).

Author information




M.I.L. and A.J.A. contributed equally to the work. M.I.L., A.J.A., V.H., C.F., P.-H.G. and M.L. designed the study and interpreted the results. M.I.L., A.J.A., and V.H. analyzed the data. M.I.L. and A.J.A. drafted the manuscript. V.H., C.F., P.-H.G. and M.L. gave valuable comments to the manuscript. All authors read and approved the final manuscript.

Ethics declarations

Competing interests

The authors declare no competing financial interests.

Electronic supplementary material

Rights and permissions

This work is licensed under a Creative Commons Attribution 4.0 International License. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in the credit line; if the material is not included under the Creative Commons license, users will need to obtain permission from the license holder to reproduce the material. To view a copy of this license, visit

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Lassenius, M., Ahola, A., Harjutsalo, V. et al. Endotoxins are associated with visceral fat mass in type 1 diabetes. Sci Rep 6, 38887 (2016).

Download citation

Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.


Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing