Association between Serum Adiponectin and Vitamin D in Obese Pre- and Early Postmenopausal Women: A Cross-Sectional Study

Abstract

Objectives

This study aimed to investigate the possible association between vitamin D and adiponectin, a biomarker of insulin resistance, in pre- and early postmenopausal women with obesity.

Methods

This cross-sectional study included 118 obese women aged 45–55 years (60 premenopausal and 58 postmenopausal women). Anthropometric measurements were obtained. Body fat percentage and visceral fat were assessed using bioelectrical impedance analysis. Serum adiponectin and 25-hydroxyvitamin D (25(OH)D) levels were analyzed using chemiluminescent microparticle immunoassay.

Results

Serum adiponectin levels were higher in the postmenopausal group (7.7 ± 3.8 µg/mL) than in the premenopausal group (5.6 ± 3.8 µg/mL) (P = 0.003). Serum 25(OH)D levels were lower in the postmenopausal group (15.4 ± 6.4 ng/mL) than in the premenopausal group (20.2 ± 4.7 ng/mL) (P < 0.001). There was no significant correlation between serum adiponectin and 25(OH)D in either group (r = −0.008, P = 0.949 in the premenopausal group; r = −0.135, P = 0.312 in the postmenopausal group). Multiple linear regression analysis showed that, after controlling for age, waist circumference, and serum vitamin D levels, adiponectin levels increased by an average of 0.208 units for each 1-cm increase in hip circumference.

Conclusions

Serum vitamin D levels were not correlated with adiponectin levels in obese pre- and early postmenopausal women. Further studies are needed to investigate factors influencing adiponectin and determine whether vitamin D supplementation provides benefits for obese middle-aged women.

INTRODUCTION

The menopause transition is the period during which ovarian hormones change from premenopausal to postmenopausal levels in women. Menopause transition influences regional fat distribution, especially central adiposity [1]. A greater amount of visceral adipose tissue increases the risk of cardiovascular diseases and diabetes [2] because it can synthesize and secrete bioactive substances, such as adipokines, proinflammatory cytokines, reactive oxygen species, and prothrombotic and vascular constrictor factors [1], which are important in metabolic processes [3].

Adiponectin, an important adipokine, is mainly produced in the adipose tissue and has been proposed as a novel biomarker associated with reducing insulin resistance, inflammation, and atherosclerosis [4]. Additionally, it prevents injury-induced vascular stenosis and has potent antioxidant effects on endothelial function and anti-inflammatory effects [3]. Individuals with coronary artery disease, diabetes, hypertension, metabolic syndrome, or visceral obesity have low serum adiponectin levels [5].

Vitamin D influences the production and function of adipokines and has been associated with adiponectin levels [6,7]. Low serum vitamin D levels can result in endothelial dysfunction, increase inflammatory reactions, and worsen oxidative stress, contributing to multiple metabolic problems [8]. In obese postmenopausal women, a low serum vitamin D level is one of the most commonly encountered issues [9]. Vitamin D supplementation is recommended to improve bone and muscle quality for postmenopausal women with vitamin D deficiency [10]. Some studies reported an increase in adiponectin levels following vitamin D supplementation [11,12], whereas others did not [13,14,15]. A recent meta-analysis showed that vitamin D intake had no significant effect on adiponectin levels; however, the secretagogue effect was shown only in diabetic subjects [16]. This resulted from heterogeneity in the study design and populations, vitamin D status, the dose of vitamin D supplements, duration of intervention, or factors affecting vitamin D levels.

Currently, vitamin D has been studied for its association with adiponectin levels in different populations; nonetheless, the results are inconclusive [17,18,19]. Among obese middle-aged women, both pre- and postmenopausal women are susceptible to an increased risk of metabolic and cardiovascular diseases; however, limited research data are available regarding the association between vitamin D and adiponectin in this specific population. The present study aimed to investigate the possible association between serum adiponectin and vitamin D levels in obese pre- and postmenopausal women. If a positive association is observed, these findings may inform future research on vitamin D supplementation strategies in obese women during the transition period.

MATERIALS AND METHODS

This cross-sectional study was conducted at the King Chulalongkorn Memorial Hospital in Bangkok, Thailand, from December 2016 to July 2017. The study was conducted in accordance with the principles set forth in the Helsinki Declaration and was approved by the Institutional Review Board of the Faculty of Medicine, Chulalongkorn University (no. 093/59, date: April 28, 2016). It was registered in the Thai Clinical Trials Registry (TCTR20180501001). All participants provided written informed consent for their participation in this study.

Study population

All women aged 45–55 years with a body mass index (BMI) of at least 25 kg/m² were eligible for this study. We used the obesity criteria defined by the World Health Organization for Asians because a BMI of > 25 kg/m² increases the moderate risk of comorbidities in Asian ethnicities [20]. This study included premenopausal women with regular menstrual cycles and no vasomotor symptoms in the past year. We also included early postmenopausal women with a 12-month history of amenorrhea with vasomotor symptoms and had attained menopause ≤ 10 years prior. Those who underwent hysterectomy or bilateral oophorectomy; were diagnosed with diabetes, hypertension, cancer, renal impairment, and abnormal parathyroid gland function; and were on statins, vitamin D supplements, or hormone therapy at the time of the study period were excluded.

The sample size was obtained using the following formula [21]

$$N={\left(\frac{Z_{\alpha }+Z_{\beta }}{0.5\times ln\left(\frac{1+r}{1-r}\right)}\right)}^2+3$$

Using α = 0.05, β = 0.20, and the expected correlation coefficient from the relevant study [18] (r = 0.38), the sample size needed was 52 participants for each group. The sample size (60 participants per group) required for this study was achieved by adding a 10% loss to follow-up.

Outcome measurements

The main outcomes collected were serum 25-hydroxyvitamin D (25(OH)D) levels, circulating adiponectin levels, weight, height, BMI, blood pressure, waist and hip circumferences, body fat percentage, and visceral fat rating using bioelectrical impedance analysis (BIA). Demographic data, including age, menopausal status, years since menopause, exercise participation, sunscreen application, and alcohol consumption, were recorded. Exercise participation was classified as none or at least one day per week. Sunscreen application was categorized as none, face only, or face and body. Alcohol consumption was classified as none or any alcohol use.

Methods

All participants with a BMI of at least 25 kg/m² were interviewed to obtain information related to menopausal status, exercise, sunscreen application, and alcohol consumption. The participants fasted for at least 8 hours prior to the morning blood collection. A blood sample of 10 mL was collected from the participants to assess serum 25(OH)D and adiponectin levels. Waist and hip circumferences and blood pressure were recorded. Body fat percentage, visceral fat rating, and muscle mass were measured using BIA.

Regarding the study period from December 2016 to July 2017, we anticipated that seasonal variation in serum 25(OH)D levels would be minimal due to Thailand’s geographical and cultural context. Situated near the equator, Thailand experiences abundant sunshine and relatively stable ultraviolet B (UVB) radiation throughout the year.

Laboratory assessments

Serum 25(OH)D levels were measured using chemiluminescent microparticle immunoassay (Architect i-1000; Abbott^®) with a lower detection limit for the sensitivity of 3.1 ng/mL. The inter-assay and intra-assay coefficients of variation (CVs) of vitamin D were 3.0% and 2.1%, respectively. Adiponectin was measured using an enzyme-linked immunosorbent assay (Varioskan Flash Multimode Reader; Thermo Scientific^®) with a lower detection limit for sensitivity at 1.5 ng/mL. The inter-assay and intra-assay CVs for adiponectin were 3.65% and 1.08%, respectively. For analysis, we used 1.5 if any value was less than 1.5.

The body fat percentage, visceral fat rating, and muscle mass were measured using BIA technology (body composition analyzer SC-330; Tanita^®) with an accuracy of ± 2%.

Statistical analysis

Statistical analysis was performed using Stata software version 11 (StataCorp). Demographic data were analyzed using descriptive statistics. Inferential statistics (two independent-sample t tests, Pearson correlation coefficient, and multiple linear regression) were used to test the association and the influencing factors. The level of significance was set at P < 0.05.

RESULTS

A total of 120 women aged 45–55 years (60 participants per group) were recruited for this study. Among these, 118 participants (60 obese premenopausal women with a mean age of 49.9 ± 2.6 years and 58 obese early postmenopausal women with a mean age of 52.9 ± 1.7 years) were selected for analysis. Two women in the postmenopausal group were excluded from the analysis because of missing data. The time after attaining menopause was 4.1 ± 2.6 years in the postmenopausal group. Baseline lifestyle and behavioral factors were comparable between the two groups (Table 1). Exercise participation was similar across cohorts, with 43.3% of the premenopausal group and 39.7% of the postmenopausal group engaging in exercise at least one day per week (P = 0.685). Additionally, sunscreen use patterns remained consistent between groups (P = 0.957); approximately 57%–59% of all participants applied sunscreen to the face only, while only a small minority (15.5%–16.7%) reported application to both the face and body. Finally, alcohol consumption did not differ significantly, with most participants reporting no intake (85.0% premenopausal vs. 91.4% postmenopausal; P = 0.284).

Table 1. Participants’ baseline characteristics.

Factor		Premenopausal group (n = 60)	Postmenopausal group (n = 58)	P value
Age (y)		49.9 ± 2.6	52.9 ± 1.7	< 0.001a
BMI (kg/m2)		28.2 ± 3.3	28.5 ± 3.4	0.631
Waist circumference (cm)		90.3 ± 8.1	89.6 ± 8.4	0.656
Hip circumference (cm)		103.2 ± 7.3	104.7 ± 7.4	0.256
Body fat (%)		39.3 ± 4.5	39.9 ± 4.2	0.444
Visceral fat rating		8.8 ± 1.7	9.1 ± 1.7	0.310
Muscle mass (kg)		38.7 ± 3.0	38.2 ± 3.0	0.336
Systolic BP (mmHg)		129.3 ± 13.5	129.1 ± 15.7	0.957
Diastolic BP (mmHg)		76.6 ± 10.7	76.0 ± 10.6	0.773
Adiponectin (µg/mL)		5.6 ± 3.8	7.7 ± 3.8	0.003a
25(OH)D (ng/mL)		20.2 ± 4.7	15.4 ± 6.4	0.001a
Time since menopause (y)		–	4.1 ± 2.6	–
Exercise				0.685
	None	34 (56.7)	35 (60.3)
	Exercise ≥ 1 d/wk	26 (43.3)	23 (39.7)
Sunscreen use				0.957
	None	16 (26.7)	15 (25.9)
	Face only	34 (56.7)	34 (58.6)
	Face and body	10 (16.7)	9 (15.5)

Data are presented as number (%) or mean ± SD. Groups were compared using independent-sample t tests or Pearson’s chi-square tests, as appropriate.

BMI: body mass index, BP: blood pressure, 25(OH)D: 25-hydroxyvitamin D, –: not available.

^aSignificant level at 0.01 (two-tailed).

The BMI, waist and hip circumferences, blood pressure, body fat percentage, visceral fat rating, and muscle mass were not significantly different between the groups (Table 1). The average serum adiponectin level was significantly higher in the postmenopausal group (7.7 ± 3.8 µg/mL) than in the premenopausal group (5.6 ± 3.8 µg/mL) (P = 0.003). The serum 25(OH)D level was significantly lower in the postmenopausal group (15.4 ± 6.4 ng/mL) than in the premenopausal group (20.2 ± 4.7 ng/mL) (P < 0.001). Most of the participants in the postmenopausal group (81.4%) had vitamin D deficiency (25(OH)D < 20 ng/mL), whereas half of the participants in the premenopausal group had vitamin D deficiency (20 ≤ 25(OH)D < 30 ng/mL). The serum vitamin D levels were normal in only 1.7% and 3.3% of the postmenopausal and premenopausal groups, respectively (Fig. 1).

Fig. 1. Vitamin D status in the premenopausal and postmenopausal groups.

The association between serum adiponectin and serum 25(OH)D levels was tested using the Pearson correlation coefficient; however, no significant correlation was detected in the premenopausal group (r = –0.008, P = 0.949) and postmenopausal group (r = –0.135, P = 0.312) (Table 2). Furthermore, no linear correlation was observed in either group (Fig. 2).

Table 2. Correlation between adiponectin and vitamin D in obese pre- and postmenopausal groups.

Factor	Premenopausal group (n = 60)		Postmenopausal group (n = 58)
	1	2	1	2
Adiponectin (µg/mL)	–		–
25(OH)D (ng/mL)	–0.008 (P = 0.949)	–	–0.135 (P = 0.312)	–

Pearson correlation coefficient is reported.

25(OH)D: 25-hydroxyvitamin D, –: not available.

Fig. 2. Scatter plots between serum adiponectin and 25(OH)D levels in the premenopausal and postmenopausal groups. 25(OH)D: 25 hydroxyvitamin D.

Other factors were tested using multiple linear regression. The full model explained 7.7% of the variation in the adiponectin level of the pooled data (n = 118, r² = 0.109; adjusted r² = 0.077; P = 0.011). The linear regression equation was as follows: adiponectin level = –7.329 + 0.022 (age) – 0.083 (waist circumference) + 0.208 (hip circumference) – 0.078 (vitamin D). Age was not a significantly associated factor in the full model (P = 0.868) and was eliminated in the reduced model. The reduced model explained 8.5% of the variation in the adiponectin level of the pooled data (n = 118, r² = 0.108; adjusted r² = 0.085; P = 0.004). The relationship between adiponectin levels and hip circumference was statistically significant (P = 0.006). The reduced linear regression equation was as follows: adiponectin level = –6.086 – 0.084 (waist circumference) + 0.208 (hip circumference) – 0.080 (vitamin D). The average adiponectin level increased by 0.208 units per centimeter in hip circumference when age, waist circumference, and vitamin D levels were controlled (Table 3).

Table 3. Multiple linear regression of adiponectin (n = 118, pooled data from both pre- and postmenopausal groups).

Factor	Full model				Reduced model
	B	Beta	t	P value	B	Beta	t	P value
Age (y)	0.022	0.015	0.166	0.868
Waist circumference (cm)	–0.083	–0.175	–1.262	0.209	–0.084	–0.176	–1.274	0.205
Hip circumference (cm)	0.208	0.394	2.805	0.006b	0.208	0.393	2.815	0.006b
25(OH)D (ng/mL)	–0.078	–0.121	–1.311	0.193	–0.080	–0.125	–1.386	0.168
Constant	–7.329		–0.802	0.424	–6.086		–1.162	0.248
	r2 = 0.109, Adj r2 = 0.077, F = 3.443, P value = 0.011a				r2 = 0.108, Adj r2 = 0.085, F = 4.621, P value = 0.004b

Dependent variable: adiponectin (µg/mL), n = 118.

25(OH)D: 25 hydroxyvitamin D, Adj: adjusted.

^aSignificant level at 0.05 (two-tailed). ^bSignificant level at 0.01 (two-tailed).

DISCUSSION

This study focused on early postmenopausal women because any intervention or early prevention strategies would work better in these women than in those in their late postmenopausal phase. We found low serum adiponectin levels in both obese premenopausal and obese early postmenopausal women; however, adiponectin levels were higher in postmenopausal women than in premenopausal women. Our premenopausal participants could present with subclinical metabolic changes that affect adiponectin levels. However, this result was similar to that reported by Gavirila et al. [22], Jürimäe and Jürimäe [23], and Vaidya et al. [18].

Several studies have shown that adiponectin levels are low in obese subjects. Gariballa et al. [24] studied total adiponectin levels in overweight and obese individuals in the United Arab Emirates and showed that increased visceral fat in overweight and obese subjects was associated with decreased total adiponectin levels. Ahl et al. [5] studied the association of adiponectin and adiposity distribution with metabolic health and found that adiponectin levels in obese individuals were lower than those in non-obese individuals. In addition to obesity, adiponectin levels are decreased in individuals with coronary artery disease, diabetes, and hypertension [25]. Increased body fat, adipocyte hypertrophy, and macrophage infiltration of adipose tissue activate proinflammatory cytokines (tumor necrosis factor-alpha [TNF-α], interleukin-6 [IL-6], interleukin-10 [IL-10]) and nitric oxide, which can reduce the expression of adiponectin release from adipocytes [26].

Low vitamin D and low adiponectin levels are often observed in obese women, which were not consistent with the results of many studies on the association between vitamin D and adiponectin. In our study, serum adiponectin could not be associated with vitamin D levels in obese early postmenopausal or premenopausal women. Bidulescu et al. [19] conducted the META-health study, which also showed no significant correlation between serum adiponectin and vitamin D in obese women, but there was a significant direct association in lean Caucasian women. The association between vitamin D and adiponectin depends on race, sex, and BMI category. De Luis et al. [27] showed that adiponectin levels were unrelated to vitamin D and calcium levels in patients with primary hyperparathyroidism.

However, many prior studies have shown an association between serum adiponectin and vitamin D levels in different circumstances. Gennage-Yared et al. [28] observed a positive correlation between adiponectin and vitamin D in a Middle-Eastern population of young people with normal BMI. Lwow and Bohdanowicz-Pawlak [29] also found the same result in healthy postmenopausal Polish women. Kim et al. [30] showed a significant correlation between vitamin D levels and adiponectin levels among overweight and obese people. Vaidya et al. [18] confirmed a positive independent association between 25(OH)D and adiponectin in large cohorts of women and men, though they noted that BMI did not significantly modify this relationship. Nimitphong et al. [17] showed a correlation between insufficient vitamin D status and lower adiponectin in subjects with abnormal glucose tolerance. There was a positive association between serum 25(OH)D and adiponectin after adjusting for BMI in people with prediabetes and type 2 diabetes mellitus [31].

Studies on vitamin D supplementation and adiponectin levels have shown different results. Mousa et al. [32] reported that 16-week vitamin D supplementation increased the adiponectin concentration in overweight/obese adults with vitamin D deficiency. Hussain et al. [33] observed that 12-week vitamin D supplementation increased adiponectin levels in patients with non-alcoholic fatty liver disease. However, Ulutas et al. [34] showed no change in serum adiponectin levels after vitamin D replacement in patients receiving peritoneal dialysis. The meta-analysis, which included nine randomized controlled trials up to March 2020, showed that vitamin D supplementation had no statistically significant effect on serum adiponectin levels. Serum adiponectin levels increased only in obese subjects after vitamin D supplementation [16].

Most of the participants in our study had vitamin D deficiency. The premenopausal level of vitamin D is higher than postmenopausal levels. This could be attributed to the following: estrogen increases the vitamin D binding protein, some of the older postmenopausal women do not go out in the sun, or the skin could have a reduced capacity to make active vitamin D. Neither premenopausal women nor the postmenopausal women have adequate vitamin D levels overall, which may have affected our results. However, we demonstrated that only the hip circumference representing healthy adipose tissue was significantly positively correlated with serum adiponectin levels. Gynecoid adipose tissue, which is associated with low metabolic risk, may have affected the correlation between vitamin D and adiponectin in this study. Moreover, because of the complexity of the association, linear correlation or available statistical methods might not be able to detect the association. Even though all known confounders had been controlled for, there may still be many factors that could have been involved, which we could not control in this study, such as proinflammatory cytokines from adipocyte dysfunction, nitric oxide, insulin resistance, lifestyle, some physical activity, or sun exposure.

This study had some limitations. First, our study is cross-sectional, so we could not prove causality or directionality. Second, we did not measure the serum parathyroid hormone, insulin resistance, lipid profiles, plasma glucose, and prediabetes parameters that could influence the association between adiponectin and vitamin D. Although we measured total adiponectin levels, we did not measure the high molecular weight adiponectin levels that are known to be more sensitive in predicting insulin resistance than total adiponectin levels [35]. The vitamin D that we measured was the total vitamin D level. In obese people, the binding proteins are often decreased, resulting in lower levels of total vitamin D.

Another potential limitation noted was the timing of blood sampling across different seasons [36]. However, previous research in tropical regions like Thailand suggests that seasonal fluctuations in vitamin D are less pronounced than in temperate climates due to consistent year-round UVB availability [37]. In contrast to high-latitude regions where deficiency rates can surge significantly during winter months, studies in Southeast Asia have shown that factors such as being female, living in urban areas, and intentional sun avoidance are more consistent predictors of low vitamin D status. Specifically, behaviors that avoid sun exposure, thereby making 25(OH)D levels at a lower baseline regardless of season. Consequently, the seasonal distribution of our sampling is unlikely to have significantly biased the interpretation of the results.

In the present study, a high prevalence of vitamin D deficiency was observed, affecting 81.4% of postmenopausal and 50% of premenopausal women. This represents a potential limitation, as the range of 25(OH) D level concentrated primarily within the deficient and insufficient categories, which may have reduced the statistical power to detect significant correlations with other metabolic parameters. In previous epidemiological studies, the lack of participants within the optimal range of vitamin D can mask potential biological associations that might only become apparent when comparing deficient states to truly sufficient ones [1,2]. Future research involving populations with a broader distribution of vitamin D levels, including those with sufficient status, is required to further elucidate these relationships.

The strength of this study is that we focused on a specific group of early postmenopausal women with obesity who need to reduce metabolic syndrome and insulin resistance, which are important cardiovascular risks in postmenopausal women. To our knowledge, this is one of the first studies to explore the relationship between vitamin D and adiponectin specifically in Asian women across the menopausal transition. Moreover, we included other factors in the model, such as body composition and visceral fat, to determine if they affected serum adiponectin.

Conclusions

Our study showed that vitamin D is not associated with adiponectin in obese pre- and postmenopausal women with vitamin D deficiency, which is inconsistent with the results of previous studies. Further prospective cohort studies are needed to investigate other unknown factors that affect adiponectin and clarify whether vitamin D is associated with adiponectin. This may provide evidence to extend the benefits of vitamin D supplementation in middle-aged women, especially those with vitamin D deficiency. However, adiponectin is a promising therapeutic target because of its anti-inflammatory and cardioprotective effects.