Effect of Vitamin D₃ Supplementation in Combination with Weight Loss on Inflammatory Biomarkers in Postmenopausal Women: A Randomized Controlled Trial

Overweight and obesity are associated with the risk of developing several types of cancer (1–3). The underlying mechanisms linking adiposity to cancer risk are still unknown, but it is likely that the tissue microenvironment in the obese state is involved in dysregulation of a number of systems including inflammatory pathways (4–6), and in the expression of adipokines, secreted by adipose tissue, such as adiponectin and leptin (7). The inflammatory response is mediated via cytokines, proteins that can either promote (e.g., c-reactive protein (CRP), IL1β, TNFα, or suppress inflammation (e.g., IL10). A variety of proinflammatory cytokines including IL1β and IL6 have been implicated in the etiology of cancer (8–12). Adiponectin is a peptide hormone whose levels inversely correlate with body mass index (BMI; ref. 13). It has insulin-sensitizing, antiangiogenic, and anti-inflammatory effects (14, 15). Leptin is a hormone that regulates appetite, food intake, and body weight. Both adiponectin and leptin play a central role in energy homeostasis, metabolism, and regulation of adiposity (16, 17).

The definitions of vitamin D [25-hydroxyvitamin D (25(OH)D)] deficiency and insufficiency have been debated (18), with classifications for “deficiency” ranging from <10 ng/mL to <20 ng/mL (19–21). The Workshop Consensus for Vitamin D Nutritional Guidelines estimated that between 50% to 60% of older populations worldwide are vitamin D deficient (19–21). A recent meta-analysis and systematic review of both observational cohort studies of circulating 25(OH)D, and randomized clinical trials of vitamin D supplementation, reported inverse associations of circulating 25(OH)D with risk of death due to cardiovascular disease, cancer, and other causes, as well as reduced overall mortality with Vitamin D₃ supplementation (22). Furthermore, several investigations suggest that vitamin D deficiency may be associated with increased levels of circulating inflammatory markers (23–26).

Several studies have reported that weight loss is associated with reductions in inflammation and in adipokines: in a previous randomized controlled trial (RCT), a behavioral weight-loss intervention significantly reduced circulating CRP, serum amyloid A, and IL6; increased adiponectin, and decreased leptin (27, 28). However, whether vitamin D administration can increase the effect of weight loss on inflammation-related biomarkers and adipokines is unknown.

This study is ancillary to the Vitamin D diet and Activity (ViDA) clinical trial which randomized overweight/obese, 25(OH)D insufficient (i.e., ≥10–<32 ng/mL), postmenopausal, healthy women to one year of 2,000 IU oral vitamin D₃ supplementation daily versus daily placebo. All randomized women underwent a concomitant weight-loss (diet + exercise) program, with a goal of 10% weight loss. We reported previously that levels of CRP decreased significantly with increasing serum 25(OH)D change from baseline to 12 months, adjusted for change in weight, only among the 55% of participants whose study medication adherence data were available (29).

The primary aims of this study are to examine, in the context of a completed RCT, whether 2,000 IU oral vitamin D supplementation in combination with a weight-loss program, reduces serum levels of leptin or the proinflammatory cytokines TNFα, IL6, IL1β, and IL8; or increases levels of adiponectin or the anti-inflammatory marker IL10, compared with a weight-loss program+placebo in postmenopausal overweight or obese women who had insufficient levels of 25(OH)D at baseline. In preplanned analyses, we also stratified analyses by categories of weight change.

This study is ancillary to the ViDA study (Trial Registration: www.clinicaltrials.gov Identifier NCT01240213), a randomized clinical trial comparing the effects of 12 months of 2,000 IU/day oral vitamin D (cholecalciferol or vitamin D₃) supplementation versus placebo on weight and other biomarkers of breast cancer risk in overweight and obese postmenopausal women partaking in a lifestyle based weight-loss program.

The study was performed with the approval of the Fred Hutchinson Cancer Research Center (Seattle, WA) Institutional Review Board, in accordance with an assurance filed with and approved by the U.S. Department of Health and Human Services. Written informed consent was obtained from each participant.

The study has been described in detail elsewhere (29). Briefly, 218 postmenopausal, overweight or obese (BMI ≥ 25 kg/m²), ages 50 to 75 years, not taking hormonal therapy, with insufficient serum 25(OH)D concentrations (i.e., ≥10–<32 ng/mL), were enrolled between 2010 and 2012, and randomly assigned to a 1-year intervention arm consisting of 2,000 IU/day of oral vitamin D₃ supplementation + a lifestyle-based weight-loss program (n = 109; ‘Vitamin D’); or to a control arm of a daily placebo + a lifestyle-based weight-loss program (n = 109; “placebo”). Ineligibility criteria included contraindications to taking >2,000 IU vitamin D₃/day; already participating in a diet or exercise intervention; history of bariatric surgery; use of weight loss medications; or serum 25(OH)D <10 ng/mL (deficient); and ≥32 ng/mL (replete). Randomization was stratified according to BMI (<30 kg/m² or ≥30 kg/m²) and consent for optional breast and abdominal subcutaneous fat biopsies. The number of women randomized to each study arm did not differ by season (χ² P > 0.99). All staff except study statisticians were blinded to randomization status. Eighty-six percent of women returned for their 12-month appointment to provide biospecimens and survey data (N = 94, placebo arm; N = 93, vitamin D arm). Of these women, only 55% remembered to return their unused pills. Thus, complete pill counts are available only for a subset of women: 54% of women in the vitamin D arm and 56% of women in the placebo arm. Of these, vitamin D-randomized women consumed 98% of capsules compared with 96% in placebo-randomized women. At baseline, 59 (27.1%) women reported taking vitamin D supplements on a regular basis, compared with 41 (21.8%) at 12 months. The 12-month change in vitamin D intake from supplements did not differ between study arms (P = 0.60).

The ViDA lifestyle-based weight-loss program was administered to all participants. It included a diet and exercise component, with a weight-loss goal of 10% of participants' baseline weight, which was adapted from a successful intervention that we have previously used in a similar population of overweight and obese postmenopausal women (30), based upon the Diabetes Prevention Program and Look Ahead lifestyle change weight loss programs (31, 32).

The goals of the diet program were: total daily energy intake of 1,200 to 2,000 kcal/day based on baseline weight, less than 30% daily energy intake from fat, and a 10% reduction in body weight by 6 months with maintenance thereafter to 12 months. The nutrition program, led by behaviorally trained registered dieticians, was delivered in groups and individual sessions.

The goal of the exercise program was: ≥45 minutes of moderate-to-vigorous intensity exercise, 5 days per week (225 minutes/week) for 12 months. Women attended two sessions per week at our study facility where they were supervised by an exercise physiologist, and performed their remaining sessions at home. Facility-based exercise consisted of treadmill walking or jogging, stationary bicycling, and use of other aerobic machines, while a variety of home exercises were encouraged including walking/hiking, aerobics, and bicycling.

At baseline and 12 months, participants provided a 12-hour fasting 50 mL sample of blood, which was processed within 1 hour of collection and stored at −70°C. Participants were instructed to refrain from alcohol (48 hours) and vigorous exercise (24 hours) before clinic appointments. Of 218 participants randomized, baseline and 12-month serum was available for 187 (86%) participants (93 control; 94 intervention).

Serum levels of TNFα, IL6, IL1β, IL8, and IL10 were assayed at the Dept. of Laboratory Medicine, University of Washington (UW), Seattle, WA, using Milliplex Map Human High Sensitivity T Cell Panel, Immunology Multiplex Assay (EMD Millipore Corp.). Internal QA samples containing high and low levels of each analyte were included in each assay to assess the intra-assay coefficient of variation (CV). Baseline and 12-month samples from each individual were included in the same batch, and participants' samples were randomly placed across batches. Laboratory personnel were blinded with regard to participant and QA sample identity. The intra-assay CVs for each assay for high and low concentrations, respectively, were as follows: TNFα, 22.6% and 31.9%; IL1β, 18.7% and 29.4%; IL8, 24.9% and 39.5%; IL10, 29.4% and 20.5%; and IL6, 36.4% and 41.8%. Inter-assay CVs were: TNFα, 14.4%; IL1β, 22.92%; IL8, 24.3%; IL10, 25.91%; and IL6, 25.5%. Serum levels of adiponectin and leptin were measured at the Northwest Lipid Research Laboratory at the UW using commercially immunoassays (Millipore Inc.), with CVs of 8.4% and 8.8% (adiponectin) and 9.1% and 14.3% (leptin), respectively. Serum levels of 25(OH)D were measured as previously described (29).

Demographics, lifestyle behaviors such as sun-exposure and dietary intake, physical activity, medical history, and supplement use were collected by questionnaire; anthropometrics were measured and BMI was calculated as kg/m², at baseline and 12 months. Total and percentage body fat and trunk fat mass was measured by a DXA whole-body scanner (GE Lunar). We generated an inflammatory biomarker composite z-score of all of the pro- and anti-inflammatory markers combined as described by Hopkins and colleagues (33). Briefly, a normalized z-score for each biomarker value, with a mean of zero and SD of 1.0, was calculated as z = (x − μ)/σ, where x is a participant's biomarker value at a given visit, and μ and σ are the study population mean and SD, respectively, at baseline. The combined z-score for each participant at each trial visit was created by summing the z-scores of each inflammatory marker. IL10 was subtracted as it is classified as an anti-inflammatory cytokine. Inclusion of previously measured CRP (29) in the biomarker score had no effect on the results, and was omitted. Medication use known to have effects on CRP and other inflammatory markers, such as (i) systemic corticosteroids and (ii) either ACE (angiotensin-converting enzyme) inhibitors, statins, or NSAIDs (34), was coded as use/non-use at baseline and 12 months.

One participant was omitted from the analysis as she was missing baseline serum measurements. A natural logarithmic transformation was applied to the outcome variables to improve the normality of the distribution.

Mean changes in IL6, IL1β, IL10, IL8, TNFα, the inflammatory composite score, and adiponectin and leptin, from baseline to 12 months, stratified by study arm, were computed. The intervention effect on these variables was examined based on the assigned treatment at randomization, regardless of adherence or study retention (i.e., intent-to-treat). Mean 12-month changes in the vitamin D group were compared with placebo using the generalized estimating equations modification of linear regression to account for intra-individual correlation over time.

Changes in weight were calculated and used to stratify observed changes in analytes between arms at 12 months. Weight-loss was categorized as no change/gained any weight (referent); lost <5% of baseline weight; lost ≥5–<10% of baseline weight; and lost ≥10% of baseline weight.

All models are presented as unadjusted; and adjusting for age, race/ethnicity, baseline serum 25(OH)D, baseline BMI, vitamin D intake (diet+supplement), calcium intake (diet+supplement), average sun exposure and medication use (use of either corticosteroids and/or CRP-lowering agents). Results are presented as (i) percent change, calculated as 100 × (follow-up geometric mean − baseline geometric mean)/baseline geometric mean; (ii) absolute treatment effect, calculated as (absolute change in vitamin D group − absolute change in placebo group); and (iii) relative treatment effect, calculated as (follow-up geometric mean in vitamin D group) × (baseline geometric mean in placebo group)/[(baseline geometric mean in vitamin D group) × (follow-up geometric mean in placebo group)].

All statistical tests were two sided. Statistical analyses were performed using SAS software (version 9, SAS Institute Inc.).

At baseline, participants were on average aged 59.6 years, with a mean BMI of 32.4 kg/m² (Table 1). The majority were non-Hispanic white (86.2%), and the average baseline serum 25(OH)D level was 21.4 (SD 6.1) ng/mL. As previously reported, a total of 86% of participants completed their 12-month measurements. Serum 25(OH)D increased by a mean of 13.6 (95% CI, 11.6–15.4) ng/mL in the vitamin D arm while it decreased by a mean 1.3 (95% CI, −2.6 to −0.3) ng/mL in the placebo arm (P < 0.0001). Changes in weight, BMI, waist circumference, percentage body fat, insulin, and CRP were similar between groups (all P > 0.05; ref. 29).

Although there was a reduction in levels of all analytes comparing baseline and 12 months, there were no significant differences in the magnitude of these changes between arms (Table 2). We examined the effect of removing outlying data points that corresponded to values >99th percentile from the analyses. Three participants had IL1β levels classified as outliers; 3 had IL8 or TNFα classified as outliers (values ≥99th percentile), and 4 had IL6 or IL10 classified as outliers. Removal of these data had no effect on the results (data not shown).

We stratified women by degree of weight-loss, using cut points shown to have clinical significance (ref. 35; loss of <5%; 5%–10%; and ≥10% of baseline weight) compared with participants who experienced no change/gained weight (the referent group; Table 3). Changes in levels of analytes by randomization arm within each weight strata were compared with the referent group, adjusting for age, ethnicity, baseline serum 25(OH)D, vitamin D intake (diet+supplement), calcium intake (diet+supplement), medication use, and average sun exposure.

Participants who lost at least 5% to 10% of their baseline weight in the vitamin D₃ arm, significantly reduced IL6 levels above that of weight-loss alone, compared with the referent group (Table 3): absolute change −0.75 pg/mL (−17.2%), placebo versus −1.77 pg/mL (−37.3%) vitamin D, P = 0.004. Similarly, participants who lost ≥10% of their baseline weight randomized to the vitamin D arm, had greater decreases in IL6 compared with those in the placebo arm: −0.41 pg/mL (−13.6%) placebo versus −0.67 pg/mL (−17.3%) vitamin D, P = 0.02. Effects of vitamin D on IL1β were inconsistent when stratified by weight loss, but absolute changes across strata were very small. Finally, there were no intervention effects on the anti-inflammatory cytokine IL10, TNFα, IL8, the inflammatory biomarker composite score, adiponectin or leptin, when stratified by weight loss.

We next compared changes in levels of analytes from baseline to 12-month follow-up by intervention arm, stratified by baseline BMI (<30 kg/m², ≥30 kg/m²); no statistically significant differences were observed (all P > 0.05; data not shown). Similarly, there were no statistically significant changes in levels of analytes between arms, when stratified by baseline levels of serum 25(OH)D (<20 ng/mL, 20–<32 ng/mL; all P > 0.05, data not shown).

In an analysis limited to women randomized to vitamin D, there were no significant differences between levels of analytes comparing the 40 women who did not become replete, with 53 women who became replete [i.e., 25(OH)D ≥32 ng/mL] by 12 months (data not shown). Finally, there were no differences between arms, comparing women with and without complete pill counts (data not shown). Exclusion of outliers from any of these analyses had no effect on results.

Here, we report that a 12-month reduced calorie/increased physical activity weight loss intervention in combination with 2,000 IU of vitamin D₃ supplementation per day had no overall effect on circulating levels of inflammatory cytokines compared with the same weight loss intervention + placebo. Participants, who lost at least 5% to 10% of their baseline weight in the vitamin D₃ arm, significantly reduced IL6 levels above that of weight-loss alone. The nonsignificant trend of a reduction in levels of the anti-inflammatory cytokine IL10 with vitamin D₃ supplementation in combination with increased weight-loss is surprising, but may reflect downregulation of this cytokine in response to less localized inflammation. Weight loss in combination with vitamin D₃ supplementation had no effect on levels of IL8, TNFα, the composite biomarker score, adiponectin, or leptin.

Vitamin D deficiency, which has been associated with increased risk for certain cancers (36–40), is associated with higher circulating concentrations of several markers of inflammation, including CRP and IL6 (24–26, 41). It is known that both weight-loss and increased levels of vitamin D₃ are associated with reductions in IL6. We previously reported results from a different RCT in overweight or obese postmenopausal women that used a similar behavioral weight-loss intervention as the present study, and which resulted in significant reductions in IL6 and CRP levels compared with controls (27). A variety of in vitro studies have demonstrated that the active form of vitamin D, 1,25-dihydroxyvitamin D (1,25(OH)2D₃) inhibits the production of IL6 (23), possibly by inhibiting p38 via MAPK phosphatase 5, which contains a putative vitamin D response element within its promoter that associated with the vitamin D receptor following treatment with 1,25-(OH)2 D₃ (42). Of interest, a recent study found that 1,25(OH)2D resulted in a dose-dependent downregulation of COX-2 expression in murine macrophages under both basal and lipopolysaccharide-stimulated conditions. It also reduced IL6 and TNFα secretion by suppression of Akt phosphorylation and NF-κB signaling (23). Weight loss and vitamin D₃ supplementation may act synergistically via different mechanisms to reduce levels of IL6. Alternatively, weight loss may increase the bioavailability of vitamin D in different tissues. However, in the present study, achieving 25(OH)D repletion (>32 ng/mL) in women randomized to the vitamin D arm was not associated with significant differences in levels of these analytes when compared with women who did not achieve repletion.

A variety of studies have examined the effect of vitamin D supplementation on inflammatory markers in both chronically ill and healthy individuals, with conflicting results.

A 6-month 4-arm RCT randomized 131 healthy women who were deficient in Vitamin D, to receive 60,000 IU/week of vitamin D₃ for 8 weeks followed by 60,000 IU/fortnight; calcium; dual supplementation or placebo. The study examined changes in gene expression of IFNγ, IL4, and its antagonist-IL-4δ2, and a variety of transcription factors involved in their regulation, but reported no significant effect of vitamin D₃ supplementation on their expression (43). Two groups of obese but otherwise healthy adolescents were randomized to receive 4,000 IU of vitamin D₃ daily versus placebo. After 6 months, there was no difference between groups in serum levels of CRP, IL,6 or TNFα, though the authors suggested the study was inadequately powered to detect biologically relevant changes in these three markers (44). A variety of RCTs found no effect on vitamin D₃ supplementation on inflammatory biomarkers. No effect of vitamin D₃ supplementation on CRP or IL6 was observed in a year-long RCT in 305 healthy postmenopausal women ages 60 to 70 years, comparing 400 IU versus 1,000 IU placebo supplementation (26). Similarly no effect on CRP levels were found in RCTs comparing 5,000 IU/day of vitamin D₃ versus vitamin C supplementation on 88 acutely hospitalized patients treated for an average of 8.2 days (45); 7,000 IU of vitamin D₃ daily versus placebo of 52 overweight men (>30 kg/m²) and women ages 18 to 50 years (26); or in an RCT of 438 overweight/obese subjects, 21 to 70 years old, randomized to one year of either 40,000 IU/week of vitamin D₃, 20,000 IU/week, or placebo for one year. The latter study also found no effect on IL10 (46). Finally, an RCT comparing 3 months of 1,000, 2,000, or 4,000 vitamin D3 supplementation versus placebo in 328 African Americans, found no statistically significant effect of supplementation on changes in CRP, IL6, IL10, or sTNF-R2 (41).

Sixty-three children and adolescents with inflammatory bowel disease were randomized to receive either 400 IU of vitamin D₂/day or 1,000 IU in summer/2,000 IU in winter. Participants who received the lower dose had significantly elevated levels of CRP and IL6 compared with those who received the higher dose CRP 31% versus 10%, P = 0.04, and IL6 54% versus 27%, P = 0.05; respectively (47). In 267 patients with systemic lupus erythematosus, a chronic multisystem inflammatory autoimmune disease, a year-long supplementation with daily 2,000 IU/day of vitamin D₃ versus placebo for 12 months was associated with a reduction in inflammatory markers (48). Finally, an RCT which randomized 92 patients with colorectal adenoma to receive either 6 months of 2 g/day calcium and/or 800 IU/day of vitamin D₃ versus placebo. The study reported a significant reduction in the combined inflammatory biomarker score (77%, P = 0.003) in the vitamin D₃ group compared with placebo (33).

Our previous analysis found no effect of vitamin D₃ supplementation on levels of CRP (either overall or to repletion >32 ng/mL; ref. 29). However, we did observe greater reductions in CRP levels with greater increases in vitamin D₃, suggesting a possible dose-response effect, although this effect was no longer significant after adjusting for percentage weight loss.

Strengths of our study include its prospective design, the use of a double-blind–controlled intervention, the broad set of inflammatory markers that were evaluated, and the year-long time-frame.

Limitations include the fact that we did not include women with a 25(OH)D concentration <10 ng/mL, and that complete pill counts were only available for 55% participants over the course of the study. All participants received a weight-loss intervention, thus we were unable to determine independent effects of vitamin D₃ supplementation. Finally our study population was relatively homogeneous, and these results may not be generalizable to other racial/ethnic groups. We used multiplex immunoassays to measure concentrations of these cytokines, and inter- and intra-assay CVs were high. A number of studies have cited this issue as a potential limitation of bead-based multiplex immunoassays, compared with ELISAs (49, 50). However, these and other reports state that despite high CV values, excellent correlations were observed between ELISAs for IL1β, IL10, IL6, and TNFα when compared with multiplex assays including Milliplex (49–54), and have demonstrated that these immunoassays correlated well with ELISA measurements in serum (53, 54). Finally a study examining the utility of a fluorescent bead-based assay to simultaneously measure a variety of cytokines reported that inter-individual differences outweighed substantial assay variation for these assays (55).

In conclusion, 12 months of vitamin D₃ supplementation in combination with a weight-loss program in overweight or obese postmenopausal women, was not associated with overall reductions in inflammatory cytokines compared with women undergoing a weight-loss program alone. However, participants who lost at least 5% of their baseline body-weight, and who were randomized to the vitamin D arm, had significantly greater reductions in IL6, compared with those randomized to the placebo arm. Further investigations into the relationship between weight-loss, vitamin D₃ supplementation, and inflammation are warranted.

No potential conflicts of interest were disclosed.

Conception and design: C. Duggan, L.A. Korde, C.-Y. Wang, A. McTiernan

Development of methodology: L.A. Korde, A. McTiernan

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): C. Duggan, C. Mason, I. Imayama, L.A. Korde, A. McTiernan

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): C. Duggan, J. de Dieu Tapsoba, C.-Y. Wang

Writing, review, and/or revision of the manuscript: C. Duggan, J. de Dieu Tapsoba, C. Mason, I. Imayama, L.A. Korde, C.-Y. Wang, A. McTiernan

Study supervision: L.A. Korde, A. McTiernan

This work was supported by Breast Cancer Research Foundation 2013-2014, Susan G. Komen for the Cure Scientific Advisory Council Award 2010-12, NIH R03 CA162482-01, Seattle Cancer Consortium Breast Cancer SPORE (P50 CA138293), and Fred Hutchinson/University of Washington Cancer Consortium (all to A. McTiernan) and the Safeway Foundation to (to C. Duggan). Clinical Trial Registration: www.clinicaltrials.gov identifier: NCT01240213.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.