Serum Vitamin D Is Inversely Associated with Blood Eosinophil Count Among Adults with Allergic Asthma

Abstract

OBJECTIVE:

To evaluate the role that serum vitamin D concentration plays in the number of eosinophils in peripheral blood in adults with allergic asthma.

MATERIAL AND METHODS:

A total of 142 patients were categorized based on different cutoff points for eosinophils: ≥200, ≥300, ≥400, and ≥500 cells/mL. The vitamin D concentration was stratified into <20 and ≥20 ng/mL. The association between vitamin D (independent variable) and eosinophils (dependent variable) was explored using multivariate analysis.

RESULTS:

The average number of eosinophils in the included patients was 418 cells/mL, and 33.8% of the included patients had vitamin D concentrations ≥20 ng/mL. Asthmatic patients with vitamin D< 20 ng/mL had a higher mean concentration of eosinophils than did asthmatic patients with vitamin D ≥20 ng/mL (464 ± 377.7 eosinophils/mL vs. 327.8 ± 247.2 eosinophils/mL, P = .025). We also observed that vitamin D was inversely correlated with eosinophil count (rho = 0.244, P = .003). In the multivariate analysis, vitamin D <20 ng/mL showed a significant inverse association with each cutoff value for eosinophilia (odds ratio >1).

CONCLUSION:

Vitamin D concentrations <20 ng/mL are associated with a significant increase in the number of eosinophils in blood. Studies that analyze the use of vitamin D supplements as complementary therapy for the treatment of asthma are needed.

MAIN POINTS

• The frequency of vitamin D concentration ≥20 ng/mL in patients with asthma was 33.8%.

• The prevalence of eosinophilia in blood based on the different cutoff values were: eosinophils ≥200 cells/µL, 75.4%; eosinophils ≥300 cells/µL, 54.9%; eosinophils ≥400 cells/µL, 40.8%; and, eosinophils ≥500 cells/µL, 30.3%.

• Asthmatic patients with vitamin D concentration <20 ng/mL had higher blood eosinophil concentrations than patients with asthma and vitamin D ≥20 ng/mL.

• There was an inverse association between vitamin D concentration and blood eosinophils.

INTRODUCTION

Among the most frequent hypovitaminosis worldwide is vitamin D (VD) deficiency. In the adult population, the prevalence of VD deficiency (<20 ng/mL) is approximately 40%.^1-3 Vitamin D is mostly recognized for its influence on calcium and phosphorus metabolism as well as bone metabolism; however, an increasing number of extraosseous effects are increasingly drawing attention. 25-hydroxyvitamin D (25[OH]D) (calcitriol), the active form of VD, participates in the regulation of innate and adaptive immune responses. It favors the production of defensins and cathelicidins and improves chemotaxis and the function of physical barriers, among other things. In the adaptive immune response, calcitriol inhibits the production of Th1 cytokines (IL-12, IFN-γ, IL-6, IL-8, TNF-α, and IL-9) and increases the production of anti-inflammatory cytokines Th2 (IL-4, IL-5, and IL-10); in addition, it inhibits the proliferation of Th1 cells and induces the production of Treg cells.⁴ At the level of eosinophils, VD reduces necrosis and, consequently, the secondary release of cytotoxic granules from eosinophils in tissues.⁵ Eosinophils cultured in a VD-free medium exhibit spontaneous activation; in contrast, the levels of activation products are reduced in the presence of VD.⁶

In Latin America, there is little information on the relationship between VD and blood eosinophils in adults with asthma. Our primary objectives were, first, to determine the prevalence of alterations in serum VD concentration and, second, to evaluate the role of serum VD concentration on the number of eosinophils in peripheral blood.

MATERIAL AND METHODS

Study Site and Patients

Our teaching hospital is located in the metropolitan area of Guadalajara in western Mexico. In this metropolitan area, the average elevation is 1540 m above sea level; the climate is temperate most of the year, with an average temperature of 18°C; and the months with the greatest solar radiation are from March to August, with a pronounced peak during May.⁷

In this cross-sectional study, only adults aged 18 years or older were included, and all were residents of the metropolitan area. Patients with a history of respiratory disease exacerbated by aspirin or near-fatal asthma, those who were using VD supplements, systemic steroids, anticonvulsants, rifampicin, or antiretrovirals during the previous 4 weeks, and those who had an asthma attack 1 month prior to the study were not included. The patient inclusion period was from October 2017 to April 2019.

Serum Vitamin D, IgE, and Eosinophils

Vitamin D, total serum IgE, and eosinophils in peripheral blood were quantified in the central laboratory of the hospital on the same day that the blood sample was obtained. The magnetic particle-based chemiluminescence technique was used to quantify VD (AccessTotal®, Beckman Coulter) and total IgE (AccessTotal IgE®, Beckman Coulter) concentrations; the former is expressed in ng/mL, and the latter is expressed in IU/mL. Flow cytometry (CELL-DYN Ruby^TM; Abbott Diagnostics Division, Abbott Laboratories, IL, USA) was used to quantify the total eosinophil concentration, expressed as cells per microliter (µL).

Skin Testing and Forced Spirometry

Allergic sensitization was determined by the presence of at least 1 positive skin test to 1 or more of the tested aeroallergens (house dust mites, dander, cockroach mix, pollen from grasses, trees and weeds, and fungal spores). Histamine was used as a positive control, and glycerin was used as a negative control. The aeroallergens were placed on both volar surfaces of the forearms, and a Hollister Stier® lancet was used for puncture; skin test results were interpreted in accordance with international guidelines.⁸

Respiratory function (forced expiratory volume [FEV₁] and forced vital capacity [FVC]) was assessed via spirometry using a Master Screen-Body PFT spirometer (Jaeger®, CareFusion, Baesweiler, Germany). Each participant performed a maximum of 8 maneuvers or until 3 tests met the criteria of acceptability and reproducibility. In addition, a challenge test with a short-acting bronchodilator inhaler (salbutamol 0.4 mg) was performed. This test served to assess the reversibility of airway obstruction.

Definitions

The following cutoff points were used to classify eosinophilia in blood: ≥200, ≥300, ≥400, and ≥500 cells/µL.

Serum VD concentrations ≥20 ng/mL were considered sufficient to maintain calcium homeostasis.⁹

Serum levels of total IgE ≥100 IU/mL were considered outside the normal range.

Allergic asthma was considered the presence of symptoms compatible with the disease, in addition to evidence of reversible airway obstruction and at least 1 positive skin test to one of the tested aeroallergens.

Ethics

Our study was approved by the ethics committee of the New Civil Hospital of Guadalajara "Dr. Juan I. Menchaca"; each of the patients had to sign a written informed consent form to participate in the study, another to undergo spirometry, and another to undergo skin prick tests.

Statistical Analyses

Quantitative variables with a normal distribution are presented as the mean and SD. Categorical variables are presented as frequencies and proportions. The Spearman rho correlation test was performed to detect an association between the variables. Finally, to identify variables associated with eosinophilia, different multivariate logistic regression analysis models were performed, 1 for each eosinophilia cutoff value. A value of P < .05 was considered statistically significant. Data analysis was performed with the Statistical Package for Social Sciences (SPSS) version 23.0 (IBM Corp.; Armonk, NY, USA).

RESULTS

In this study, a total of 142 patients were included (75% were women), and the mean age was 35 years (Table 1). The frequencies of allergic rhinitis and atopic dermatitis in patients with asthma were 82% and 10%, respectively; 37% had a personal history of hospitalization for asthma during the previous year. Based on the results of asthma control tests (ACT), 28% of the subjects perceived their asthma as controlled. Mites and cockroaches were the allergens to which patients were most sensitized, with atmospheric fungi being the allergen to which patients were least sensitized. Regarding respiratory function, the mean baseline FVC and FEV₁ were 88% and 71% of the predicted value. Notably, asthmatics with VD concentrations <20 ng/mL had significantly lower FVC and FEV₁ than patients with VD greater than ≥20 ng/mL. The mean total serum IgE concentration was 418 IU/mL; 67% of the patients had concentration equal to or greater than 100 IU/mL. The mean VD concentration was 18.4 ± 8.1 ng/mL, and in 33.8% of the patients, the concentration was ≥20 ng/mL. The mean and median absolute blood eosinophil counts were 418 cells/µL and 340 cells/µL, respectively. When comparing the median and mean concentration of eosinophils between patients with asthma and VD <20 ng/mL and patients with VD ≥20 ng/mL, there was a statistically significant difference, i.e., 359 eosinophils/mL vs. 240 eosinophils/µL (P = .023) and 464 ± 378 eosinophils/µL vs. 327.8 ± 247 eosinophils/µL (P = .025), respectively.

Table 1.

Characteristics of the Population with Asthma

	Total (n = 142)	Vitamin D		P
		<20 ng/mL (n = 94)	≥20 ng/mL (n = 48)
Age (years), mean ± SD	34.6 ± 10.4	36.0 ± 10.2	31.8 ± 10.3	.024
Sex, female, n (%)	106 (74.6)	69 (73.4)	37 (77.1)	.634
Current smoking, n (%)	8 (5.6)	6 (6.4)	2 (4.2)	.717
Alcohol consumption, n (%)	33 (23.2)	18 (19.1)	15 (31.3)	.106
Atopic comorbidity, n (%)
Allergic rhinitis	117 (82.4)	74 (78.7)	43 (89.6)	.108
Food allergy	27 (19.0)	18 (19.1)	9 (18.8)	.954
Atopic dermatitis	13 (9.2)	8 (8.5)	5 (10.4)	.709
Urticaria	7 (4.9)	2 (2.1)	5 (10.4)	.044
BMI (kg/m2), mean ± SD	29.1 ± 6.2	29.8 ± 6.7	27.9 ± 4.9	.078
Hospitalization in the previous year, n (%)	52 (36.6)	32 (34.0)	20 (41.7)	.372
ACT, points
Mean ± SD	15.2 ± 5.4	14.6 ± 5.4	16.2 ± 5.3	.103
≥20, n (%)	40 (28.2)	24 (25.5)	16 (33.3)	.328
Allergic sensitization, n (%)
Mites	94 (66.2)	60 (63.8)	34 (70.8)	.404
Cockroaches	70 (49.3)	45 (47.9)	25 (52.1)	.635
Trees	67 (47.2)	43 (45.7)	24 (50.0)	.631
Epithelia	57 (40.1)	37 (39.4)	20 (41.7)	.791
Weeds	60 (42.3)	36 (38.3)	24 (50.0)	.182
Grass	43 (30.3)	29 (30.9)	14 (29.2)	.836
Fungi	18 (12.7)	12 (12.8)	6 (12.5)	.964
Respiratory function, mean ± SD
% predicted baseline
FVC	87.9 ± 18.3	85.1 ± 18.9	93.4 ± 16.0	.011
FEV1	71.2 ± 19.2	68.8 ± 20.55	75.9 ± 15.4	.039
FEV1/FVC	67.1 ± 10.3	66.7 ± 11.2	67.9 ± 8.1	.493
Total serum IgE, IU/mL
Median (P25-P75)	211 (74-502)	204 (67-524)	224 (82-501)	.626
IgE ≥100, n (%)	95 (66.9)	61 (64.9)	34 (70.8)	.477
Vitamin D, ng/mL
Mean ± SD	18.4 ± 8.1	14.2 ± 3.7	26.5 ± 8.2	<.0001
≥20, n (%)	48 (33.8)	N/A	N/A	N/A
Eosinophils, cells/µL
Mean ± SD	418.1 ± 344.5	464 ± 378	328 ± 247	.025
Median (P25-P75)	340 (205-580)	359 (227-602)	240 (145-477)	.023

ACT, asthma control test; BMI, body mass index; FEV₁, forced expiratory volume; FVC, forced vital capacity; IgE, immunoglobulin E; N/A, not applicable; P₂₅-P₇₅, percentile 25-percentile 75.

The following provides the prevalence of eosinophilia in blood based on the different cutoff values: eosinophils ≥ 200 cells/µL: n = 107 (75.4%, 95% CI 67.6% to 81.7%); eosinophils ≥300 cells/µL: n = 78 (54.9%, 95% CI 46.7% to 62.9%); eosinophils ≥400 cells/µL: n = 58 (40.8%, 95% CI 33.1% to 49.1%); and, eosinophils ≥500 cells/µL: n = 43 (30.3%, 95% CI 23.3% to 38.3%).

As seen in the correlation matrix shown in Table 2, VD concentration was inversely correlated with blood eosinophil count (P = .003, statistical power of 84%) and age (P = .037). Blood eosinophil count was inversely correlated with age (P = .0398) and ACT (P = .006) and positively correlated with IgE (P = .002). Additionally, there was a significant correlation between age with body mass index (BMI) (positive correlation), FEV₁ (negative correlation), and IgE (negative correlation) (P < .0001). Asthma control test was significantly correlated with FEV₁. There was no significant correlation between VD and BMI, FEV₁, ACT, and IgE. There was also no significant correlation between eosinophils and BMI and FEV₁. There was no significant correlation between the other variables.

Table 2.

Magnitude of Correlation Between Variables

	Eosinophils	Vitamin D	Age	BMI	FEV1	ACT	IgE
Eosinophils	1
Vitamin D	-0.244 P = .003	1
Age	-0.071 P = .398	-0.176 P = .037	1
BMI	-0.104 P = .217	-0.048 P = .570	0.317 P < .0001	1
FEV1	-0.067 P = .428	0.096 P = .258	-0.298 P < .0001	-0.053 P = .513	1
ACT	-0.230 P = .006	0.028 P = .743	-0.017 P = .842	-0.087 P = .305	0.427 P < .0001	1
IgE	0.255 P = .002	0.070 P = .411	-0.317 P < .0001	-0.145 P = .086	0.088 P = .300	-0.160 P = .057	1

Correlation estimated through Spearman’s rho.

ACT, asthma control test; BMI, body mass index; FEV₁, forced expiratory volume in the first second, predicted baseline; IgE, immunoglobulin E.

Table 3 presents several multivariate models of factors associated with blood eosinophilia based on the following cutoff points: ≥200, ≥300, ≥400, and ≥500 cells/μL. In all adjusted models, VD <20 ng/mL exhibited a statistically significant inverse association with eosinophilia. In contrast, IgE ≥100 IU/mL was identified as a risk factor for eosinophilia for all cutoff points, except when the eosinophil count was ≥500 cells/μL. Other risk factors for eosinophilia were sex (female) with an eosinophil count ≥400 cells/μL and fungal sensitization with an eosinophil count ≥500 cells/μL. Last, age younger than 35 years was inversely associated with an eosinophil count ≥500 cells/μL.

Table 3.

Different Multivariate Models to Identify Factors Associated with Eosinophilia in Patients with Asthma

Dependent variable: eosinophils ≥200 cells/μL
	aOR	95% CI	P
Vitamin D <20 ng/mL	2.84	1.29-6.26	.020
Female	—	—	.361
Age ≥35 years	—	—	.441
Allergic rhinitis	—	—	.981
ACT ≥20 points	—	—	.578
IgE ≥100 IU/mL	3.18	1.40-7.21	.006
Sensitization to fungi	—	—	.144
Dependent variable: eosinophils ≥ 300 cells/μL
	aOR	95% CI	P
Vitamin D <20 ng/mL	2.62	1.53-6.56	.002
Female	—	—	.227
Age ≥35 years	—	—	.448
Allergic rhinitis	—	—	.806
ACT ≥20 points	—	—	.067
IgE ≥100 IU/mL	3.55	1.67-7.57	.001
Sensitization to fungi	—	—	.896
Dependent variable: eosinophils ≥400 cells/μL
	aOR	95% CI	P
Vitamin D <20 ng/mL	2.20	1.01-4.79	.046
Female	3.22	1.32-7.82	.010
Age ≥35 years	—	—	.133
Allergic rhinitis	—	—	.268
ACT ≥20 points	—	—	.431
IgE ≥100 IU/mL	2.92	1.32-6.49	.008
Sensitization to fungi	—	—	.318
Dependent variable: eosinophils ≥500 cells/μL
	aOR	95% CI	P
Vitamin D <20 ng/mL	3.16	1.27-7.85	.013
Female	—	—	.096
Age ≥35 years	0.40	0.18-0.89	.025
Allergic rhinitis	5.33	1.36-20.93	.016
ACT ≥20 points	—	—	147
IgE ≥100 IU/mL	—	—	.377
Sensitization to fungi	3.65	1.19-11.15	.023

OR obtained by binary logistic regression. All variables were entered dichotomously. The OR was not calculated for those variables that were not included in the model because they did not present a statistically significant association.

ACT, asthma control test; aOR, adjusted odds ratio; IgE, immunoglobulin E.

DISCUSSION

In this study, many patients with asthma showed alterations in VD concentration. When we compared the number of eosinophils between asthmatic patients with VD <20 ng/mL and those with VD ≥20 ng/mL, the former had a noticeably higher concentration of eosinophils in peripheral blood.

Based on the results of this study, VD deficiency in patients with allergic asthma is not infrequent, as up to 70% had VD concentrations lower than 20 ng/mL. This result contrasts markedly with those from population studies. In the United States, the National Health and Nutrition Examination Survey 2005-2006 showed that slightly more than 0% of the analyzed population had VD levels <20 ng/dL³; similar findings were observed in Mexico, 43.6% of adults also had VD levels <20 ng/dL.¹ There are different reasons to explain why VD may be insufficient in this population, including low exposure to sunlight, skin color, age, use of sunscreen, and low physical activity.¹⁰ In our study, most patients had moderate airway obstruction, which may have contributed to less outdoor physical activity; consequently, they would be less exposed to ultraviolet radiation, thus favoring less VD production. The use of systemic or topical steroids at high doses can also contribute to this same problem.

Notably, in adult patients with allergic asthma and VD deficiency, there was an inverse association with eosinophil count in peripheral blood. This finding is consistent with previous studies conducted in children and adults with allergic and nonallergic asthma.^11,12 The use of VD supplements in asthmatic patients with VD deficiency has been shown to significantly reduce allergic inflammation.² Similar to asthmatic patients, patients with allergic rhinitis and VD deficiency also have increased eosinophil concentrations in blood compared with that in controls.¹³ Last, the results herein indicate that a cutoff point of VD ≥20 ng/mL serves as a good predictor of eosinophilia in blood when the eosinophil count is 200 cells/μL or more. The modulating role of VD in the immune response seems to be responsible for its effect on eosinophils by favoring an increase in the production of Th2 cytokines in addition to inhibiting the proliferation of Th1 cells.¹⁴ According to The Global Initiative for Asthma, VD supplementation may reduce the number of asthma exacerbations or improve asthma control, especially when VD levels are <20 ng/mL.¹³

Among other relevant findings in this study is the positive association between IgE concentration and eosinophils in blood. Notably, the relationship was stronger when the amount of eosinophils was ≥200 or ≥300 cells/μL, slightly diminished at ≥400 cells/μL, and finally disappeared at ≥500 cells/μL. A wide network of cytokines, including IL-4, IL-5, IL9, IL-13 and IL-31, is involved in the regulation of IgE expression and eosinophils; however, IL-4 and IL-13 are the main cytokines involved in T2 inflammation.¹⁵ In our study population, all asthmatic patients were sensitized to allergens; thus, different mechanisms that favor eosinophil expression or fungal sensitization, as we present below, are probably involved in the lack of an association between IgE and eosinophil quantities ≥500 cells/μL. Although IgE is a biomarker of T2 inflammation, it has been suggested that IgE concentration does not help identify patients with asthma who will respond better to treatment with type 2 monoclonal antibodies¹⁵; therefore, eosinophils in blood become a better therapeutic target for the control of asthma.

In this study, an eosinophil count ≥500 cells/μL was 60% less frequent in patients with asthma and was only observed in those over 35 years old. The emergence of biological products for the control of T2 asthma redefined the role of eosinophils as a therapeutic target, marking the need to define the best cutoff points to achieve the greatest benefit of their use. Thus, there are some biologics that are indicated when the eosinophil count is ≥150 cells/μL and others that are indicated when the count is ≥400 cells/μL.¹⁶ In our study population, 75% of patients had eosinophil counts ≥200 cells/μL, but only 30% had counts ≥500 cells/μL. It seems that with increasing age, the number of eosinophils decreases, as do their effector functions.¹⁷

We also observed that allergic rhinitis in asthmatic patients was significantly associated with blood eosinophilia; however, this occurred only when eosinophil counts were ≥500 cells/μL. The mechanisms that could explain this phenomenon are not clear, but it has been shown that the concentration of eosinophils in blood is higher when allergic rhinitis is not controlled than when it is controlled.¹⁸

Finally, in our study, slightly more than 10% of the patients were sensitized to fungi; this finding is consistent with previously published data.^19,20 Notably, this event was a factor associated with blood eosinophilia, especially when the eosinophil count was ≥500 cells/μL. Colonization by fungi in patients with asthma is recognized for its ability to induce vigorous immune responses, which justifies the intentional search for allergic bronchopulmonary mycosis in patients with allergic sensitization to fungi; for this, new diagnostic criteria have been recently published.²¹

Due to the design characteristics of the study, we were not able to establish if the association between VD and eosinophilia is constant; it has been shown that the levels of eosinophils in blood can vary throughout the day and that, when treating allergic patients, they can be modified by exposure to sensitizing allergens. However, having used different cutoff points to define eosinophilia, it is unlikely that the association remained stable. Another limitation is the use of topical or systemic steroids by some participants at the time of inclusion in the study and the potential effect on blood eosinophils. Notably, we were careful not to include those patients with a history of systemic steroid use 1 month before participants were selected for this study. Regarding the use of inhaled corticosteroids, the patients were not using them regularly; but once the diagnosis of asthma was established, they were instructed to use them regularly and at adequate doses. Finally, we did not investigate the presence of parasitosis concomitant with asthma, diet characteristics, or the use of solar radiation blockers, all factors likely to modify the association between the variables.

CONCLUSION

In summary, several patients with asthma have VD concentrations below 20 ng/mL (more than 70%). In this study, VD levels <20 ng/mL were associated with a greater number of total eosinophils in peripheral blood in adults with allergic asthma. These findings fully justify the analysis of the repercussions of VD deficiency at the eosinophil level, i.e., eosinophil count, and the functional level. Similarly, more research that analyzes the use of VD supplements as complementary therapy for the treatment of asthma is needed.