Obesity and Persistent Organic Pollutants: Possible Obesogenic Effect of Organochlorine Pesticides and Polychlorinated Biphenyls

Authors


(eveline.dirinck@uza.be)

Abstract

Persistent organic pollutants (POPs) are endocrine-disrupting chemicals associated with the development of the metabolic syndrome and type 2 diabetes. In humans, little is known about their role in the potential origin of obesity. This study aims to assess the associations between serum levels of POPs and the prevalence of obesity in a cohort of obese and lean adult men and women. POP serum samples were investigated cross-sectionally in 98 obese and 47 lean participants, aged ≥18 years. Serum samples were analyzed for the presence of polychlorinated biphenyl (PCB) congeners 153, 138, 180, and 170 and for the organochlorine pesticides, dichloro-diphenyl-dichloroethylene (pp-DDE), and β-hexachlorocyclohexane (βHCH). We established a significant negative correlation between BMI, waist, fat mass percentage, total and subcutaneous abdominal adipose tissue, and serum levels of PCB 153, 180, 170, and the sumPCBs. For βHCH, we demonstrated a positive correlation with BMI, waist, fat mass percentage, and total and subcutaneous abdominal adipose tissue. PCBs 180, 170, and the sum of PCBs correlated significantly negative with homeostasis model assessment for insulin resistance (HOMAIR). βHCH correlated significantly positively with HOMAIR. A strong correlation was established between all POP serum levels and age. We established a positive relationship between high serum levels of βHCH and BMI and HOMAIR, whereas serum PCB levels were inversely correlated with BMI and HOMAIR. Combined, these results suggest that the diabetogenic effect of low-dose exposure to POPs might be more complicated than a simple obesogenic effect.

Introduction

In recent years, the prevalence of obesity, defined as a BMI of >30 kg/m2, has reached alarming proportions with 30–80% of the European adult population currently being overweight (BMI >25 kg/m2) and obesity affecting up to a third of the population (1). Obesity and overweight are known to adversely affect health and to impact the risks and prognosis for a number of serious medical conditions such as type 2 diabetes and coronary heart disease (1,2). Traditionally, the increase in obesity is attributed to an increased calorie intake and a concomitant significant reduction in physical activity and energy expenditure (3). However, evidence has emerged recently that other mechanisms might be involved. Among different hypotheses—such as gut bacterial overgrowth (4), shortness of sleep (5), and viral theories (6)—environmental factors, the so-called endocrine-disrupting chemicals, might also affect or cause changes in fat mass and subsequent obesity (7,8).

Persistent organic pollutants (POPs) are known endocrine-disrupting chemicals. They include polychlorinated biphenyls (PCBs) and organochlorine pesticides, such as pp-dichloro-diphenyltrichloroethane and its major metabolite dichloro-diphenyl-dichloroethylene (pp-DDE), and β-hexachlorocyclohexane (βHCH). Organochlorine pesticides were abundantly used pesticides until the 1960s, whereas PCBs were used since the 1930s on a worldwide scale for various industrial purposes, such as dielectric fluids in electrical capacitors, transformers, and hydraulic systems. More than 100 individual PCB congeners have been identified in commercial mixtures, whose chemical and toxicological properties are related to the number and position of the chlorine atoms. Despite the ban on their use in the United States and Europe since the 1970s, their stability, resistance to degradation, and lipophilicity has lead to significant bioaccumulation in most compartments of the ecosystem and human tissues (9,10,11,12). This bioaccumulation leads to an ongoing human exposure to POPs through a variety of pathways, but the most important is dietary intake (13). So the present concentration in serum of POPs reflects both a release from fat storage compartments as well as an uptake from present exposure.

Recent epidemiological data suggest an association between POPs burden, diabetes, and the metabolic syndrome (14,15,16). The metabolic syndrome, characterized by a cluster of metabolic disorders including central obesity, glucose intolerance, dyslipidemia, and hypertension, is a known risk factor for the development of diabetes and cardiovascular disease. A significant association between the presence of POPs and the metabolic syndrome has previously been established (15,17). Moreover, obesity is a key factor in this metabolic syndrome, as it is a known risk factor for the development of insulin resistance and diabetes (18).

PCBs are hormonally active substances, mimicking the action of natural hormones such as the thyroid hormone and estrogens (19). Low levels of PCB 77 increase adipocyte differentiation, promote the expression of proinflammatory cytokines, and augment the expression of the peroxisome proliferator-activated receptor γ, a key promotor in regulating cell energy homeostasis (20,21). These data also indicate that even low-level exposure to PCBs, as observed today in the human population, might contribute to the development of obesity (19,20,21). Mullerova et al. showed a negative correlation between PCB 153 and adiponectin in obese women, thus suggesting a possible suppression of adiponectin release or production by PCB 153 (ref. 22).

The aim of this study was to assess the associations between serum levels of POPs and the prevalence of obesity in a Belgian cohort of lean and obese men and women.

Methods

Study population

A cohort of 98 obese men and women were retrospectively selected from the database of patients visiting the weight management clinic of the Department of Endocrinology, Diabetology and Metabolism of the Antwerp University Hospital between 1998 and 2007. To be included, BMI had to be ≥30 kg/m2. All subjects were ≥18 years. In the obese population, 12 subjects (12.2%) were type 2 diabetics.

A control group of 46 men and women, matched by age and sex, was recruited from hospital staff and volunteers during the same time period. Their BMI was in the normal range between 18 and 25 kg/m2. In the control population, one subject was diagnosed having type 2 diabetes. The subject characteristics are described in Table 1.

Table 1.  Subjects characteristics
inline image

This study was approved by the ethical committee of the Antwerp University Hospital and all participants gave their informed consent.

Physical measurements

All anthropometric measurements were performed in the morning with patients in fasting conditions and undressed. Height was measured to the nearest 0.5 cm and body weight was measured with a digital scale to the nearest 0.1 kg. Obesity was defined as a BMI ≥30 kg/m2. Waist circumference was measured at the mid-level between the lower rib margin and the iliac crest. Hip circumference was measured at the level of the trochanter major and the waist-to-hip ratio was calculated. Body composition was determined by bioimpedance analysis as described by Lukaski and Bolonchuk (23), and fat mass % was calculated using the formula of Deurenberg et al. (24). A computed tomography-scan at the L4–L5 level was performed to measure the amount of total abdominal adipose tissue, visceral abdominal adipose tissue, and subcutaneous abdominal adipose tissue according to previously described methods (25).

Blood sampling

Venous blood samples were obtained from fasting subjects from an anticubital vein between 08.00 am and 01.00 pm into sterile BD Vacutainer tubes (Plymouth, UK). Serum was centrifuged within 15 min at 2,500–3,000 r.p.m. and stored in Eppendorf Safe lock tubes at −80 °C during the study period. An oral glucose tolerance test was performed with 75 g of glucose, with blood samples taken to determine glucose and insulin in the fasting state and 2 h after the glucose load. Analysis was performed at the Antwerp University Hospital laboratory. Glucose was measured using dry chemistry and reflectometry on a Vitros Fusion. Insulin was determined with the Cobas method on a Modular 170 by Roche. Diabetes was classified according to World Health Organization criteria 1998 (26). The homeostasis model assessment (HOMA) was used to calculate insulin resistance as described previously (27).

Determination of POPs

Analyses of POPs were performed at the Toxicology Centre (University of Antwerp). The samples were analyzed for the PCB congeners CB 153, CB 138, CB 180, CB 170, together with pp-DDE and βHCH. These POPs are very persistent and are indicators for background (dietary) exposure (28). The analytical method was based on the method described by Covaci and Schepens (29) and used with minor modifications for low serum volumes (up to 1 ml). Serum samples were extracted with solid-phase extraction and analyzed by gas chromatography-mass spectrometry. POP serum levels are expressed in ng/g lipids. Lipid-normalized concentrations were calculated using the formula proposed by Bernert et al. (30).The limit of quantifications for the analyzed POPs ranged between 4 and 12 ng/g lipids.

Statistical analysis

Statistical calculations were performed using SPSS, version 16.0 (SPSS, Chicago, IL). Levels below the limit of quantification were entered in the database as 50% of limit of quantification. Normality of distribution was verified using the Kolmogorov-Smirnov test. Age, BMI, fasting and 2-h postprandial glucose and insulin, and the concentrations of POPs were not normally distributed within the entire, obese, and lean groups. Therefore, Mann-Whitney U test and Spearman rank correlation were used and results are presented as median, with minimal and maximal values. Transformation with the natural logarithm, log10, or square root did not alter the distribution (data not shown). Because PCBs were always used as a mixture of different congeners, we summed the concentrations of all four PCBs (sumPCB). Other variables were normally distributed so results are presented as mean ± s.d. with minimal and maximal values. Results were considered significant at P < 0.05.

Results

The characteristics of both study groups are summarized in Table 1. Age and sex distribution are identical between obese and lean individuals, with a mean age of 40 years. The lowest BMI was 18.8 kg/m2, so no underweight participants were included. Abdominal adiposity is more prevalent among obese participants, as indicated by the higher waist circumference and the waist-to-hip ratio.

POP serum levels of the obese and control population revealed a significantly different serum concentration for PCB 153, 180, 170, sumPCB, and βHCH (Table 2).

Table 2.  Concentrations of investigated POPs (ng/g lipid) presented as median (minimum − maximum)
inline image

Analyzing the correlation between weight and POP serum levels, a significant inverse relationship with PCB 153, 138, 180, 170, and sumPCB was found (Table 3). BMI and βHCH serum levels correlated significantly positive, whereas a distinctly negative correlation was seen for PCB 153, 180, 170, and sumPCB. Because females and males differ in body fat mass percentage and body fat mass distribution, the correlation between waist and fat mass percentage and serum POPs levels was calculated. In the entire group, this analysis revealed a significant positive correlation between waist and βHCH, whereas a significant negative correlation was detected between waist and PCB 153, 180, 170, and sumPCB. The same pattern was established for the correlation between fat mass percentage and POP levels in the entire group. After dividing the entire group according to sex, βHCH correlates significantly positive and PCB 180 and 170 correlate significantly negative with waist and fat mass percentage in the male subgroup. In the female subgroup, fat mass percentage correlated significantly inverse with all four PCBs and their sum, but not with βHCH. Waist in females correlated significantly with βHCH, and inversely with PCB 153, 180, 170, and the sumPCB (Table 4). When analyzing the abdominal fat distribution in more detail, it is noteworthy that the significant negative correlation between the PCBs and abdominal fat was almost solely due to the subcutaneous fat mass. In contrast, βHCH correlated significantly positive with both subcutaneous and visceral abdominal fat. We could not establish a significant correlation between fasting glucose and any of the POPs levels. PCB 180, 170, and sumPCB did correlate in a significantly inverse manner with fasting insulin and HOMA for insulin resistance (HOMAIR), whereas βHCH correlated positively with HOMAIR, fasting insulin, and 2 h postprandial glucose and insulin.

Table 3.  Spearman rank correlations between POPs and measures of obesity and glucose metabolism in the entire group (n = 144)
inline image
Table 4.  Spearman rank correlations between POPs and waist and fat mass percentage according to gender
inline image

We established a strong positive correlation between all POPs serum levels and age (Table 2). Our data even indicated a sixfold increase of the sum of PCB serum concentration between the youngest and oldest subjects in our population (Figure 1). In the group with individuals aged <25 years, the mean sum of PCB serum concentration was 48.5 ng/g lipids, which increased to 315.8 ng/g lipids for individuals aged >50 years.

Figure 1.

Relationship between age and sumPCB serum levels in the entire group (n = 144).

Repetition of these analysis with the levels of POPs below limit of quantification as zero, revealed statistically identical results.

Discussion

In recent years, endocrine-disrupting chemicals have emerged as a novel cause contributing to the worldwide epidemic of diabetes. A strong link between POPs and disorders of glucose metabolism has been established in several populations (16,31,32,33,34). Obesity is a known risk factor for the development of type 2 diabetes. In this study, the relationship between POP serum levels and obesity was assessed.

We established a significant difference in BMI according to serum levels of PCB 153, 180, 170, sumPCB, and βHCH. As BMI increased, levels of PCB 153, 180, and 170 declined. The inverse was seen for βHCH. Hue et al. (35) did not detect a correlation between the serum levels of PCB 170 and BMI, although a positive relation was seen for PCB 180. In a Japanese study, the BMI was not associated with POPs, but a positive association was described between BMI and the dioxin-like PCB group. Unfortunately, PCB 170, 180, and βHCH were not included in their study (17). Interestingly, we found a negative association between BMI and all PCBs, except PCB 138. This inverse relationship between BMI and serum concentrations of PCBs has been established previously by Agudo et al. (10) and Wolff et al. (36). In a Spanish population, Agudo et al. (10) found that obese individuals (BMI >30 kg/m2) had lower PCB serum concentration compared to the lean group. In contrast, the group with a BMI between 25 and 30 kg/m2 showed higher PCB concentrations. A potential explanation may be found in the dilution capabilities of the PCBs: as these lipophilic contaminants are preferably stored in adipose tissue, a higher percentage of body fat will lead to fast and efficient storage, with lower serum levels as a consequence. Indeed, a significant negative correlation between serum levels of PCB 153, 170, 180, and sumPCB and fat mass percentage was detected, which supports this hypothesis. A significant negative association between PCB serum levels and the amount of abdominal, subcutaneous abdominal fat in particular, seems to suggest that PCBs stored in subcutaneous fat, are less easily diluted into the blood stream. A confounding factor may be the fact that the elimination time of PCBs in obese individuals might be different from that observed in lean individuals. Flesch-Janys et al. (37) indeed showed that individuals with higher BMI have reduced dioxin clearance, although this has not been clearly demonstrated for PCBs.

For βHCH, a positive relation was observed with BMI. Jakszyn et al. (38) also found a positive relationship between BMI and serum βHCH concentration in a Spanish population, which was not confirmed by others (35). Because βHCH is the most hydrophilic of all substances analyzed in this study, it is expected that it will be more readily detectable in serum. Jung et al. (39) established that βHCH is eliminated more slowly in individuals with a higher percentage of body fat. A significantly positive correlation was observed between fat mass percentage and βHCH in our study. In the current population, we did not detect any difference in BMI according to pp-DDE serum levels. Others found a significant difference in BMI, which might be due to the older age of the participants (35). Recently, Karmaus et al. reported a positive relationship between maternal DDE levels and BMI in adult female offspring (40).

In our study, no data were collected on other known obesogenic factors, such as sedentary lifestyle, diet, family history of obesity, or obesogenic medication. Moreover, the design of the study is cross-sectional. Therefore, a causal relationship between serum POPs and obesity is difficult to determine.

Our data confirm the influence of POPs on glucose metabolism, with a statistically significant higher insulin resistance (as assessed by HOMAIR) with higher serum βHCH levels. PCB 180, 170, and sumPCB are negatively correlated with HOMAIR. As previously discussed, higher PCB serum levels were found in the group with a lower BMI, thus indicating that the endocrine-disrupting effect of PCBs might involve a different pathway than the classical insulin resistance inducing effects of obesity. Henriksen et al. (41) described an increased fasting glucose with higher levels of dioxin in a large cohort of US veterans exposed to Agent Orange. We could not demonstrate a significant effect of serum levels of the sum of PCBs on fasting glucose. Our results are comparable with those of Lee et al., who found a strong association between organochlorine pesticides (such as βHCH) and HOMAIR in the National Health and Nutrition Examination Survey study (42). We did not find any statistically significant differences in fasting glucose or HOMAIR for pp-DDE. In our study sample, 13 participants were diagnosed with diabetes, of which one was a participant in the lean group. We could not detect a difference in serum POP levels between obese participants with and without diabetes (data not shown). Given the small number of diabetics, we estimate the group to be too small to detect such difference.

Our data confirm the life-long accumulation of POP in the human body. Previous studies have clearly demonstrated the capability of POPs to accumulate in the human body throughout lifespan (34,35,43). This positive relationship between age and serum levels of POPs was previously observed in lean, obese, and severely obese patients (35,44). Nichols et al. (44) even presented age-specific reference ranges for PCBs, based on the US National Health and Nutrition Examination Survey 2001–2002 data. In their analysis, the mean summed PCB serum level in a 20- to 29-year-old population was 30 ng/g lipid weight, whereas it was 163 ng/g lipid for 40- to 49-year-old participants, and up to 302 ng/g lipid for 60- to 69-year-old individuals. In a recent Spanish study (10), a mean serum concentration of 431 ng/g lipid in an age group of 35–44 years of age was observed. In the age group of 55–64, the mean serum concentration rose to 498 ng/g lipid. Our data, as shown in Figure 1, also indicate a sixfold increase of cumulative PCB burden between the youngest and oldest subjects in our population. Exposure to POPs has decreased in recent decades, due to the cessation of production and/or use of several of the investigated products. Together with the shorter duration of exposure, this can also account for the significant difference in burden of endocrine disruptors in younger vs. older obese and lean subjects.

There are some shortcomings associated with our study. Data were collected over a large period of time, spanning almost 10 years. Environmental PCB burden is known to have dropped substantially over a decade. We did not register information about possible professional exposure. It would be very interesting to investigate the relationship between concentration of POPs in serum and fat. Given the data from the literature (19,20,21,22), in particular the data on the influence via peroxisome proliferator-activated receptor γ, an effect on fat cell signaling by POPs seems indeed possible. In the present population, however, no fat samples were collected. It is therefore impossible to make an assumption on the concentration of POPs in fat and their effect on energy homeostasis in the fat cell. None of our groups displayed a normal distribution of POP levels, thus making the use of nonparametric tests necessary.

In conclusion, we were able to show a positive relationship between the serum concentration of the less lipophilic endocrine disruptor βHCH and BMI, whereas we found a negative relationship between the serum level of more lipophilic PCBs and BMI. We could not find a statistically significant relationship between serum levels of pp-DDE and BMI. Our study is concordant with the previous reports describing a positive relationship between serum levels of βHCH and insulin resistance, whereas we additionally found a negative relationship between serum levels of PCBs and insulin resistance. Combined, these results suggest that the diabetogenic effect of low-dose exposure to POPs might be more complex than a simple obesogenic effect. The exact mechanisms of influence of POPs on body energy homeostasis remain largely unknown. Given the current worldwide epidemic of obesity, the possible effects of endocrine disruptors on body weight are an imperative field of future research.

ACKNOWLEDGEMENT

T.G. and A.C. acknowledge the Funds for Scientific Research Flanders for a PhD fellowship and a postdoctoral fellowship, respectively. The study was supported by a GOA project (FA020000/2/3565) of the University of Antwerp.

DISCLOSURE

The authors declared no conflict of interest.

Ancillary