The High Serum Monocyte Chemoattractant Protein-1 in Obesity Is Influenced by High Parathyroid Hormone and Not Adiposity

Serum monocyte chemoattractant protein-1 (MCP-1) and parathyroid hormone (PTH) are elevated in obesity, but MCP-1 is only higher in the presence of increased PTH, and is independent of adiposity.

Abstract

Context:

Chronic high levels of PTH may be associated with up-regulation of proteases and cytokines. Monocyte chemoattractant protein-1 (MCP-1) is an inflammatory cytokine, produced predominantly by macrophages and endothelial cells, and is expressed in adipose tissue. More recently it has been shown that PTH administration increases MCP-1 expression in osteoblasts.

Objectives:

Because both PTH and MCP-1 levels are higher in obesity, the goal was to determine whether the high MCP-1 occurs only in the presence of high serum PTH and is independent of adiposity and examine its relationship with bone mineral density (BMD) and turnover.

Design, Setting, and Participants:

In this case-control clinical design, 111 eligible women were categorized into four groups: leaner women [body mass index (BMI) 23 ± 2 kg/m²] with normal or higher PTH and obese (BMI 44 ± 7 kg/m²) with normal or higher PTH.

Results:

Serum MCP-1 levels were higher (P < 0.01) in the high (PTH = 74.9 ± 27.0 pg/ml, MCP-1 = 421.5 ± 157.0 pg/ml) compared with normal PTH (PTH = 32.5 ± 10.4 pg/ml, MCP-1 = 322.5 ± 97.8 pg/ml) group, independent of BMI. C-reactive protein and adiponectin were influenced only by BMI and not PTH. MCP-1 was positively associated with osteocalcin and propeptide of type 1 collagen in the leaner (r > 0.3, P < 0.05) but not the obese women and was not associated with BMD in either group.

Conclusions:

Together these data suggest that MCP-1 is higher only in the presence of increased PTH and that adiposity alone cannot explain the higher MCP-1 levels in obesity.

Serum PTH plays a central role in the regulation of calcium homeostasis. In addition to its well-established role in the skeleton, reports suggest an association between high PTH and risk of hypertension, insulin resistance, dyslipidemia, cardiovascular mortality, and cancer (1) but is not a consistent finding in all studies (2). Chronic high levels of PTH in patients with hyperparathyroidism are associated with the up-regulation of proteases and cytokines and also increased bone turnover (3). More recently it has been shown that treatment of osteoblasts with PTH increases expression of monocyte chemoattractant protein-1 (MCP-1) (4).

MCP-1 is a proinflammatory cytokine that is produced predominantly by macrophages and endothelial cells (5). Serum MCP-1 levels are higher in patients with atherosclerosis (6), and both protein and mRNA levels of MCP-1 are higher in atherosclerotic lesions (7). In addition, MCP-1 levels are increased in obesity (8, 9), showing higher mRNA expression and protein levels in the adipose tissue, and MCP-1 levels decrease with weight loss (10, 11). MCP-1 is also expressed by the osteoblast, and both mRNA and protein expression in bone as well as serum levels increases with PTH administration in a rodent model (4, 12). The relationship between the elevated PTH and MCP-1 in obesity has not been linked, although both high PTH and MCP-1 have similar effects on cardiovascular disease (13), insulin resistance (8, 14), and mortality (15).

Obesity is associated with an increase in serum PTH, and a positive relationship is observed between fat mass and serum levels of PTH (16–18). In addition, patients with hyperparathyroidism are often heavier than their peers (19). The higher PTH may be a result of the lower serum level of 25-hydroxyvitamin D (25-OHD), possibly due to increased sequestration of vitamin D into adipose tissue or decreased exposure to sunlight. Low 25-OHD levels may thus promote secondary hyperparathyroidism and may contribute to the increase in serum PTH in this population. However, higher serum levels of PTH levels are also associated with lower whole-body fat oxidation and suppresses insulin signaling in the adipocyte and may thus promote adiposity and insulin resistance (14, 20–22). Obesity is described as a state of low-grade inflammation due to the presence of elevated levels of proinflammatory cytokines such as MCP-1, TNF-α, IL-6, and C-reactive protein (CRP) together with lower adiponectin. Both serum levels of PTH and MCP-1 are higher in obese adults, yet whether the higher PTH increases MCP-1 is not known, and it is not clear whether this occurs independent of adiposity. We examined whether PTH influences MCP-1 in obese and lean women with normal and higher levels of PTH in a case-control design. In addition, because our laboratories have found that PTH increases MCP-1 in osteoblasts, we also examined the relationship with bone turnover and bone mineral density (BMD) in this study.

Subjects and Methods

Subjects

Women aged 25–71 yr who were either leaner [body mass index (BMI) <27 kg/m²] or obese (BMI >35 kg/m²) were recruited by advertising in local newspapers and e-mail: list-serves and in the medical outpatient clinics at Robert Wood Johnson University Hospital. Clinic patients who were diagnosed with hyperparathyroidism and were eligible for this study (n = 14) were referred by the endocrinologist. Those patients interested in participating in the study completed blood work and body composition analysis before treatment. Volunteers were also recruited for this study (n = 23) to serve as controls without diagnosis of hyperparathyroidism; however, some of them had higher levels of PTH. Others included in this study had been enrolled to participate in weight loss trials in the laboratory previously (n = 74), and their baseline data were used for this study. Women were excluded if they were osteoporotic (T-score <−2.5); taking osteoporosis medications known to influence bone or mineral metabolism including use of hormone replacement therapy in the past year; had evidence of metabolic bone disease, thyroid disorders, immune disease, heart attack, or stroke in the past 6 months; or were currently undergoing treatment for hyperparathyroidism, kidney stones, diabetes, active cancers, or cancer therapy within the past 12 months. Premenopausal women who were menstruating monthly were included and were excluded if they were pregnant or lactating within the past year. Postmenopausal women were required to be at least 2 yr since their last menstruation.

Women were advised to consume 1–1.2 g/d of calcium (Citracal; Bayer, Morristown, NJ), depending on the recommended intake for their age and a daily multivitamin that contained 400 IU of vitamin D for at least 4–6 wk before all the measurements were performed. This was done to attenuate variability in serum PTH due to potential large differences in dietary intake of Ca or vitamin D. This study was approved by the Institutional Review Boards at Rutgers University and the University of Medicine and Dentistry of New Jersey, and all participants signed an informed consent form before initiation of any study procedures.

Methods

Weight and height were measured with a balance beam scale and stadiometer, respectively (Detecto, Webb City, MO). Nutrient intake was estimated using 3-d food records and analyzed using nutrient software (version 10.1; Food Works, Longvalley, NJ). BMD, bone mineral content, and body composition were measured using dual-energy x-ray absorptiometry [Lunar Prodigy Advanced; GE-Lunar, Madison, WI; coefficient of variation (CV) <1% for all sites using enCORE 2004 software (version 8.10.027; GE Lunar)]. Volumetric BMD (vBMD) was measured using peripheral quantitative computed tomography (Stratec XCT 3000; Orthometrix, White Plains, NY). Sectional images were standardized at specific sites (4 and 38%) using distal tibia as the anatomical marker and analyzed for vBMD as has been described previously (18).

Laboratory methods

Blood was aliquoted and centrifuged to separate serum that was stored at −80 C until further analysis. Serum was analyzed for calcium (Arsenazo III, Endpoint; Ponte Scientific, Canton, MI) and other markers. RIA were used to analyze serum levels of PTH (Scantibodies Laboratory, Santee, CA; CV <6.8%), 25-OHD (Diasorin, Stillwater, MN; CV <12.5%), bone formation markers such as osteocalcin (OC; BTI, Sloughton, MA; CV <9%), and propeptide of type 1 collagen (P1NP; Orion Diagnostica, Espoo, Finland, CV <10.2%). Our laboratory also participates in the international Vitamin D External Quality Assessment Scheme to ensure quality and accuracy of 25-OHD analysis. ELISA were used to analyze N-telopeptide of type 1 collagen (NTx; Wampole Laboratories, Princeton, NJ; CV <14%), MCP-1 (R&D Systems, Minneapolis, MN; CV <7.8%), CRP (R&D Systems, Minneapolis, MN; CV <8.3%), and total adiponectin (Alpco Diagnostics, Salem, NH; CV <9.2%). Urinary pyridinium cross-links, pyridinoline (PYD) and deoxypyridinoline (DPD), were measured by HPLC (CV <6%), as described previously (23), and corrected for creatinine. Serum creatinine was measured using Pointe Scientific reagent (CV <3.6%). Estimated glomerular filtration rate (eGFR) was calculated based on serum creatinine and calculated body surface area [eGFR = glomerular filtration rate (Modification of Diet in Renal Disease Study equation) BSA/1.73 m²].

Statistical analysis

We separated women into four groups based on BMI and median of PTH. The categories included normal BMI (leaner) and high (obese) BMI and normal and high PTH. A two-factor ANOVA was used to analyze the differences in the four groups [leaner and obese, higher (Hi), and or normal (NL)-PTH]. Differences among the four groups were tested by Tukey's post hoc analysis when the model F ratio was significant. Pearson's correlation coefficients (r) were used to assess relationship between MCP-1 and PTH. In addition, linear regression models were used to explain relative contributions of independent variables such as age, PTH, BMI, and 25-OHD levels on MCP-1 and other cytokines. A P ≤ 0.05 was considered significant. Data are presented as means ± sd unless otherwise indicated. All analyses were conducted using the SAS Statistical Package (SAS Institute, Cary, NC; version 9.2).

Results

One hundred eleven women who were either leaner with BMI less than 27 kg/m² (mean 23.4 ± 2.4 kg/m²) or obese with BMI >35kg/m² (mean 44.3 ± 7.0 kg/m²), were further classified as having Hi or NL serum PTH levels, according to the median PTH value (40 pg/ml for leaner and 59 pg/ml for obese) in each group. The mean age was 52 ± 10 yr (range 25–71 yr) and did not differ between groups (Table 1). Women were primarily Caucasians (n = 95) with 11 African Americans, three Hispanics, and two Asians. Women in the higher PTH groups had higher fat mass (P < 0.01) and body weight was greater (P < 0.08) than those in normal PTH groups. There was no significant interaction between BMI and serum PTH for weight, body composition, or age. Mean calcium intake was 1208 ± 360 mg/d and vitamin D intake was 464 ± 416 IU/d and was similar between the groups (Table 1). Areal BMD at femoral neck, spine (L2-L4), and hip and tibial trabecular vBMD was higher and cortical vBMD was lower in the obese women (P < 0.02) and were not affected by PTH levels. The eGFR was low in one obese subject in the Hi PTH group at 27.8 ml/min, and in 13 leaner subjects (six with NL PTH and seven with Hi PTH) with eGFR in the range of 42.9–59.8 ml/min.

Table 1.

Age, body composition, and nutrient intake of study participants (n = 111)^a,b,c

	Leaner		Obese		BMI	P2 PTH	BMI-PTH
	NL-PTH (n = 31)	Hi-PTH (n = 28)	NL-PTH (n = 26)	Hi-PTH (n = 26)
Age (yr)	49.5 ± 8.3	51.8 ± 8.8	53.2 ± 12.2	52.2 ± 11.3	0.250	0.780	0.450
Weight (kg)	60.1 ± 7.4d	63.0 ± 7.7d	109.4 ± 18.9e	117.5 ± 22.9e	<0.001	0.054	0.397
BMI (kg/m2)	23.0 ± 2.3d	23.8 ± 2.5d	44.1 ± 6.5e	44.6 ± 7.4e	<0.001	0.075	0.275
Fat mass (kg)a	19.7 ± 6.6d	22.1 ± 6.4d	48.5 ± 5.5e	50.4 ± 7.6e	<0.001	0.004	0.116
Lean mass (kg)a	36.6 ± 4.8d	38.0 ± 4.1d	49.3 ± 6.4e	49.2 ± 6.9e	<0.001	0.540	0.490
FN BMD (g/cm2)	0.886 ± 0.135d	0.876 ± 0.108d	0.974 ± 0.128e	1.010 ± 0.170e	0.001	0.649	0.408
Total Hip BMD (g/cm2)	0.923 ± 0.134d	0.907 ± 0.118d	1.062 ± 0.139e	1.077 ± 0.175e	<0.001	0.979	0.621
L2-L4 BMD (g/cm2)	1.155 ± 0.153	1.139 ± 0.230	1.252 ± 0.208	1.254 ± 0.185	0.014	0.875	0.836
Trab vBMD (mg/cm3)	197.8 ± 28.0d	190.9 ± 32.1d	240.6 ± 29.6e	252.1 ± 35.8e	<0.001	0.709	0.181
Cort vBMD (mg/cm3)	1177.9 ± 28.5d	1171.9 ± 28.6d	1142.1 ± 24.7e	1144.2 ± 28.1e	<0.001	0.744	0.513
Calcium intake (mg/d)	1216 ± 359	1285 ± 413	1194 ± 188	1131 ± 429	0.160	0.850	0.280
Vitamin D intake (μg/d)	11.3 ± 2.1	10.9 ± 1.3	10.6 ± 2.5	9.8 ± 4.9	0.119	0.307	0.716

^aFat and lean mass is reported for a subset n = 18 and n = 21 for Hi and NL-PTH, respectively, in the obese group.

^bA two-factor ANOVA with BMI and PTH as independent variables was performed.

^cValues with superscript letters d and e are significantly different using Tukey's post hoc testing.

Bone regulatory hormones and turnover markers

As expected, PTH was higher and 25-OHD was lower in the Hi than NL-PTH groups (P < 0.01) (Table 2). In addition, serum levels of PTH were higher and 25-OHD was lower in the obese compared with leaner women (P < 0.001), and there was a significant interaction between PTH and BMI for 25-OHD (P = 0.008). The levels of PTH were 40–45 pg/ml higher in both leaner and obese Hi than NL-PTH groups but only the obese group showed a lower serum 25-OHD in the Hi-PTH group (P < 0.01). Bone formation markers (OC and P1NP) were both lower in the obese than leaner women (P < 0.01), whereas the bone resorption marker, NTx, tended (P < 0.08) to be higher in the leaner population. Serum Ca and PYD was higher (P < 0.02) in the Hi than NL-PTH groups (P < 0.01).

Table 2.

Bone-regulating hormones, calcium, and turnover markers (n = 111)^a,b

	Leaner		Obese		BMI	P2 PTH	BMI-PTH
	NL-PTH (n = 31)	Hi-PTH (n = 28)	NL-PTH (n = 26)	Hi-PTH (n = 26)
PTH (pg/ml)	25.6 ± 6.9c	64.0 ± 31.8d	39.3 ± 13.8e	85.7 ± 22.1f	<0.001	<0.001	0.305
25-OHD (ng/ml)	29.0 ± 7.0c	28.8 ± 7.5c	23.3 ± 8.4d	15.5 ± 5.9e	<0.001	0.005	0.008
Serum Ca (mg/dl)	9.2 ± 0.6c	9.6 ± 0.8c,d	9.5 ± 0.4c,d	9.8 ± 0.8d	0.128	0.015	0.957
OC (ng/ml)	10.9 ± 4.9c	11.5 ± 2.8c	8.8 ± 3.1d	9.4 ± 3.2d	0.005	0.426	0.986
P1NP (μg/liter)	52.6 ± 20.8c	57.0 ± 21.7c	39.6 ± 12.7d	48.3 ± 15.6c,d	0.008	0.112	0.594
NTx (nMBCE)	16.0 ± 7.2c	15.7 ± 4.4c	12.1 ± 4.8d	15.4 ± 6.4c	0.080	0.234	0.142
PYD (nmol/mmol)	26.5 ± 7.4c	31.4 ± 12.5c,d	28.1 ± 17.9c	36.6 ± 15.8d	0.208	0.013	0.492
DPD (nmol/mmol)	10.2 ± 3.4	11.1 ± 4.6	9.9 ± 6.5	10.7 ± 3.8	0.724	0.365	0.905
Serum creatinine (mg/dl)	0.90 ± 1.14c	0.95 ± 0.15c	1.09 ± 0.30d	0.94 ± 0.15c	0.023	0.231	0.015
eGFR (ml/min)	69.3 ± 11.2c,d	66.7 ± 11.6c	76.2 ± 17.6d	93.2 ± 21.3e	<0.001	0.025	0.003

^aValues with superscript letters c, d, e, and f are significantly different.

^bA two-factor ANOVA with BMI and PTH as independent variables was performed.

MCP-1 and control cytokines

The serum levels of MCP-1 and control cytokines in the four groups are shown in Fig. 1. In the leaner women, serum levels of MCP-1 were higher in the Hi-PTH (408.3 ± 121.6 pg/ml) compared with the NL-PTH groups (300.7 ± 82.0 pg/ml) (P < 0.01). Similarly, in the obese category, MCP-1 levels were higher (P = 0.049) in the Hi-PTH (436.3 ± 191.7 pg/ml) as compared with the NL-PTH groups (348.5 ± 109.9 pg/ml). Interestingly, MCP-1 levels did not differ between the BMI categories (P = 0.13) and was also similar between Hi-PTH groups independent of BMI. In contrast, CRP levels were significantly higher and adiponectin levels lower in the obese (P < 0.01) compared with leaner women but were not different between the PTH categories (P > 0.4).

Fig. 1.

Serum levels of inflammatory cytokines (MCP-1, CRP, and adiponectin) in leaner and obese women. Values with different superscript letters are significantly different, using Tukey's post hoc testing.

Independent predictors of MCP-1 and control cytokines

The relative contribution of independent variables such as BMI, age, PTH, and 25-OHD on serum levels of MCP-1 was examined using multiple regression analysis (Table 3). PTH was the most significant predictor of MCP-1 (P < 0.02) and the β-coefficient for PTH shows that a 1 sd increase in PTH leads to a 0.26 sd increase in MCP-1 when the other variables are held constant. Interestingly, BMI was not a significant predictor of MCP-1 (P > 0.4) in the multiple regression model. However, BMI explains a significant portion of the variance (P < 0.01) for control cytokines, i.e. CRP and adiponectin, followed by 25-OHD, whereas PTH and other variables were not significant predictors of both these control cytokines.

Table 3.

Multiple regression model for the relative influence of age, BMI, PTH, and 25-OHD on cytokines

	MCP-1 (model R2 = 10.3)			CRP (model R2 = 48.2)			Adiponectin (model R2 = 21.3)
	R2	Std β	P	R2	Std β	P	R2	Std β	P
PTH	7.3	0.26	0.01	0.03	0.03	0.773	0.01	−0.03	0.718
Age	2.5	0.15	0.10	0.82	−0.11	0.220	6.91	0.26	0.004
BMI	0.3	0.08	0.45	46.4	0.61	<0.001	12.56	−0.27	0.009
25-OHD	0.2	0.06	0.61	0.91	−0.09	0.186	1.72	0.14	0.183

Std, Standard.

Cytokines and bone turnover

The relationship of MCP-1 with bone formation and resorption markers using Pearson's correlations is presented in Table 4. MCP-1 was not associated with bone turnover in the entire cohort of women or in the obese women; however, in the leaner group, MCP-1 was positively associated (r > 0.3, P < 0.05) with bone formation markers OC and P1NP and resorption marker DPD. The positive associations of MCP-1 with bone turnover in the leaner women remained, even after adjusting for PTH levels. Similarly, in all women, CRP was negatively associated with OC and P1NP (r > −0.2, P < 0.03), and adiponectin was positively associated with these bone formation markers and BMD (r > 0.3, P < 0.01) in all women, but these relationships disappeared after controlling for BMI. MCP-1 tended to be weakly associated with femoral neck and hip BMD (r > 0.17, P < 0.09) in all women.

Table 4.

Relationship of MCP-1 with bone turnover^a,b

MCP-1	All women (n = 111)		Leaner (n = 59)		Obese (n = 52)
	r	P	r	P	R	P
OCb	0.111	0.251	0.331	0.012	0.023	0.873
NTx	−0.090	0.375	0.121	0.383	−0.236	0.118
PYDb	0.209	0.031	0.216	0.106	0.184	0.205
DPD	0.132	0.175	0.267	0.044	0.064	0.66
P1NPb	0.078	0.495	0.315	0.043	−0.021	0.896

^aPearson's correlations (r) were performed between MCP-1 and bone turnover.

^bRelationship exists, even after controlling for PTH.

Discussion

It is well established that MCP-1 is elevated in obesity, and it is attributed to its synthesis in adipose tissue. Studies from our laboratories (4) have shown that PTH induces osteoblastic expression and increases serum levels of MCP-1 in rodents (12). A major role of MCP-1 in bone is to facilitate bone resorption by recruiting preosteoclasts to the remodeling sites and then promoting osteoclast differentiation. Whether the higher PTH in obesity increases MCP-1 expression and serum levels is not known. In this study we examined how two levels of serum PTH in obese and leaner individuals influence MCP-1 levels and show that MCP-1 is not high in all obese individuals and only in those with high PTH. In addition, we found that in leaner individuals with high PTH, there is a similar high level of MCP-1 that cannot be attributed to their adiposity. Unlike MCP-1, there was no positive association between PTH and other cytokines, such as CRP or adiponectin in weight-matched obese and leaner women, suggesting a unique relationship between PTH and MCP-1 that is independent of adiposity.

Obesity is associated with higher levels of serum PTH, and multiple studies have shown this relationship in both healthy individuals and in patients with primary hyperparathyroidism (16–18). Several mechanisms have been proposed for the higher PTH seen in obesity including lower 25-OHD levels (21, 24, 25) and may influence their altered BMD, as reported by our laboratory previously (18). Serum levels of 25-OHD are key regulators of serum PTH levels and are negatively correlated with BMI and body fat mass (26–28). This relationship may occur because 25-OHD is fat soluble, leading to increased deposition into excess adipose tissue and possibly other tissues in obese individuals (24, 27). It is also possible that obese people have decreased exposure to sunlight due to lower physical and outdoor activity, resulting in lower 25-OHD and higher PTH. On the other hand, high PTH levels may exacerbate obesity and its associated insulin resistance because an increase in PTH has been shown to decrease whole-body fat oxidation (21) and suppress insulin signaling in adipocytes (14). The cause-effect relationship between excess body weight and PTH is unclear, with reports showing that weight loss decreases PTH (29, 30), and some, on the other hand, show that PTH lowering via parathyroidectomy does not decrease body weight (2). Given that higher PTH exists with excess adiposity, we suggest that chronically elevated PTH contributes to higher MCP-1 levels. Higher serum MCP-1 has been previously shown to exacerbate inflammation and comorbidities associated with obesity (6, 11).

The predominant source of MCP-1 in obesity is believed to be from the macrophages of the white adipose tissue (8, 31). However, the source of MCP-1 in response to high serum levels of PTH is unclear, and indeed, our laboratories have shown that osteoblasts secrete MCP-1 in response to PTH administration (12). Because PTH suppresses insulin signaling in the adipocyte (14) and MCP-1 has been shown to have this same action (32, 33), it is not clear whether PTH and MCP-1 have independent effects on the adipocyte or whether PTH acts through MCP-1 on the adipocyte. PTH stimulates MCP-1 in another cell, the osteoblast, so we hypothesize that it may also stimulate MCP-1 in the adipocyte because both arise from the same progenitor cell. It may be that the osteoblast is a more important source of MCP-1 in lean individuals, given the positive associations of osteoblast specific bone formation markers (OC and P1NP) in the leaner population, whereas excess adipose tissue may be the primary source of MCP-1 in obesity, although future research is needed to confirm the potential role of MCP-1 in bone turnover.

Hyperparathyroid patients have a low BMD, particularly cortical bone density, compared with healthy controls (34, 35). One possible mechanism that may reduce BMD and increase bone turnover in patients with hyperparathyroidism, includes up-regulation of osteoblastic expression of cytokines (36). Because PTH influences BMD, one would expect MCP-1 to also show a similar relationship with bone; however, neither MCP-1 nor PTH was associated with BMD. This is possibly because our high PTH groups had mean PTH levels of only 64 and 86 pg/ml in leaner and obese women, respectively, and these levels are much lower than those of hyperparathyroid patients reported in the literature with low BMD (3, 37, 38).

The strengths and limitations of the study include the following. One strength is that patients that present with primary hyperparathyroidism are usually overweight or obese (19), and recruitment of a leaner population of women with higher levels of PTH was an important aspect of the current study to understand the influence of PTH in the absence of obesity. In addition, women were also advised to consume adequate calcium intake of 1–1.2 g daily before biochemical assessment to ensure that a low calcium intake did not transiently increase PTH levels in these women. Also, all subjects were recruited in the winter and early spring months to avoid the influence of sun exposure on 25-OHD levels. Some limitations to this study include the following. Because this was a retrospective case-control design, a cause-effect relationship cannot be determined, and this may also be subject to selection bias. In addition, in the obese group, the fat and lean mass was not available in 25% of the obese women because some of the severely obese subjects could not fit on the dual-energy x-ray absorptiometry table for body composition analysis. Hence, it is likely that fat mass is higher than reported in the obese groups, but because we are missing approximately the same number of women with similar BMI in the high and normal PTH groups, fat mass is unlikely to differ between the two obese groups. In addition, we did not have any data on the duration of high PTH in our subjects, and this may be important in the interpretation of these results, especially with respect to BMD. Furthermore, although the relationship of MCP-1 with bone turnover showed weak associations with formation markers in the leaner population, there was no relationship found in the obese women, and clearly further research is required to address the potential role of MCP-1 in bone turnover. Also, the small model r² for MCP-1 suggests that there are other factors influencing MCP-1 that explain a large part of its variance. However, we know from these data that PTH, and not BMI, explains a significant amount of the variance for serum MCP-1.

Together these data suggest that high PTH levels, irrespective of adiposity, are associated with higher MCP-1 levels. These findings may contribute to a better understanding of the association between high PTH levels and greater cardiovascular mortality and insulin resistance because previous studies have shown that MCP-1 plays an important role in the etiology of both of these diseases. Future studies should aim to understand whether traditional treatments to reduce hyperparathyroidism will decrease the levels of MCP-1 and thereby reduce the progression of disease.