Endocrinology and Metabolism

Sexual dimorphism in the acute effects of secondhand smoke on thyroid hormone secretion, inflammatory markers and vascular function

Andreas D. Flouris, Giorgos S. Metsios, Athanasios Z. Jamurtas, Yiannis Koutedakis

Abstract

Experimental evidence for the physiological effects of secondhand smoke (SHS) is limited, although it affects millions of people globally and its prevalence is increasing, despite currently adopted antismoking measures. Also, scarce evidence suggests that the effects of SHS may be more pronounced in men. We conducted a randomized single-blind crossover study to investigate the sex-specific SHS effects in a controlled simulated bar/restaurant environment on gonadal and thyroid hormones, inflammatory cytokines, and vascular function. Twenty-eight (women = 14) nonsmoking adults underwent a 1-h exposure to moderate SHS and a 1-h control trial. Serum and urine cotinine, gonadal and thyroid hormones, inflammatory cytokines, heart rate, and arterial blood pressure were assessed before exposure and immediately after in both trials. Results showed that testosterone (P = 0.019) and progesterone (P < 0.001) in men and 17β-estradiol (P = 0.001) and progesterone (P < 0.001) in women were significantly decreased after SHS. In men, SHS was accompanied by increased free thyroxine (P < 0.001), triiodothyronine (P = 0.020), and decreased the triiodothyronine-to-free thyroxine ratio (P = 0.033). In women, significant SHS-induced change was observed only in free thyroxine (P = 0.010), with considerable sex variation in free thyroxine and triiodothyronine and a decrease in luteinizing hormone (P = 0.026) and follicle-stimulating hormone (P < 0.001). After SHS, IL-1β (P = 0.001) and systolic blood pressure (P = 0.040) were increased in men but not women. We concluded that a 1-h SHS exposure at bar/restaurant levels is accompanied by decrements in gonadal hormones in both sexes and marked increases in thyroid hormone secretion, IL-1β production, and systolic blood pressure in men.

  • environmental tobacco smoke
  • cotinine
  • estrogen
  • cytokines

more than 126 million american and 130 million Chinese adult nonsmokers are currently exposed to secondhand smoke (SHS) on a daily basis, while global estimates include 700 million children and 50 million pregnant women (31). These figures generate major concerns given the overwhelming evidence on the adverse health effects of SHS (26) together with recent reports showing that, despite currently adopted measures, the prevalence rates of smoking are increasing (30). Experimental data from our (17) and other (14, 18, 19, 26) laboratories suggest that even brief exposures to SHS cause marked changes in thyroid hormone secretion, platelet aggregation, endothelial function, arterial pressure waveform, inflammatory markers, as well as other hemodynamic alterations involved in the development of ischemic heart disease. At least some of these effects appear to be more pronounced in men compared with women (14, 20), but the mechanism underlying this phenomenon has yet to be elucidated. We conducted a randomized single-blind crossover study to investigate the sex-specific effects of a 1-h exposure to SHS on gonadal and thyroid hormones, inflammatory cytokines, and specific parameters of vascular function. Gonadal hormones exert considerable influence on thyroid hormone regulation (6, 7) and immune function (5), which, in turn, vastly affect cardiovascular haemodynamics. Hence, our hypothesis was that the smoke-induced effect on gonadal hormones generates a sexual dimorphism in thyroid hormone secretion and immune function response to SHS, resulting in augmented changes in vascular function in men.

MATERIALS AND METHODS

Participants

The experimental protocol conformed to the standards set by the Declaration of Helsinki and was approved by the University of Thessaly ethical review board. Twenty-eight (women = 14) healthy adults (mean ± SD: age = 26.4 ± 4.4 yr; height = 171.63 ± 10.28 cm; and weight = 64.83 ± 13.12 kg) volunteered and signed informed consents. This sample size was deemed adequate based on statistical power calculations (see Statistical Analysis for details). The participants were recruited from the general population and had not participated in previous studies involving SHS exposure. Exclusion criteria included smoking, pregnancy, previous/current thyroid disorders, evidence of cardiac or pulmonary disease, as well as current disease and medications known to affect the thyroid, the pituitary function, or the metabolic status. All women participants were premenopausal with regular menstruation and were tested during the late luteal phase of their menstrual cycle. To ensure that the findings of the study were based on euthyroid individuals, participants who revealed concentrations of thyroid hormones above or below normal levels [i.e., free thyroxine (fT4): 0.8–1.9 ng/dl; triiodothyronine (T3): 0.8–1.8 ng/ml; and thyroid stimulating hormone (TSH): 0.5–5 μIU/ml] in the first baseline measurement (see Experimental Design) were excluded from further analyses. Based on the adopted criteria, all volunteers were deemed euthyroid.

Experimental Design

Participants were given a detailed verbal description of the protocol, followed by extensive familiarization with all data collection procedures and instruments during an initial familiarization session performed at least 3 days before testing. Anthropometrical measurements were also performed at this time. The study adopted a randomized single-blind crossover design with participants visiting the laboratory on 2 consecutive days (to minimize day-to-day variation in the examined parameters) in a 12-h postabsorptive state where they randomly underwent two different trials. In the experimental trial, participants were exposed to 1 h of SHS inside an environmental chamber. In the control trial, they remained in the same chamber for 1 h while breathing normal sea-level air. In both trials, participants were assessed for cotinine levels, gonadal and thyroid hormone secretion, cytokine levels, heart rate, and arterial blood pressure before they entered the chamber (baseline measurement) and immediately after their exit (follow-up measurement). These measurements were conducted in a quiet room maintained at 22–25°C air temperature. Participants consumed no food or fluid during their time inside the chamber and until after each of the follow-up measurements. To ensure that the comparatively long half-life of specific indexeses (e.g., cotinine and fT4) did not influence our results, a randomly chosen 50% of the men and the women participants (i.e., 7 men and 7 women) underwent the experimental trial first in sequence, while the remaining participants performed the experimental trial second in sequence. All participants were instructed to refrain from strenuous physical activity and other excessive stressors for 72 h before the first trial but also during the 2 assessment days, particularly during the period between awakening and experimentation and during transit from home to the laboratory. All testing was conducted by the same trained investigators who were unaware of the specific trial that each participant was undergoing. To ensure that the investigators would not be able to differentiate between trials, all participants were given a shirt and athletic pants to wear upon exiting the chamber after both trials. These clothes had been previously exposed to SHS inside the chamber and emitted strong tobacco scent. Both trials began at 7:30 am and were conducted using identical precalibrated equipment.

SHS Exposure

During both trials (i.e., experimental and control), participants were instructed to remain seated at rest (i.e., reading a book or magazine) for 1 h inside a 6 × 5 × 4 m environmentally controlled chamber with air temperature maintained at 24°C (air velocity: 0.05 m/s) and humidity at 45%. In the experimental trial, participants were exposed to SHS adjusted at a carbon monoxide (CO) concentration of 23 ± 1 ppm to meet levels previously reported for bar/restaurant environments (13). We checked for gradients of gas concentrations and particle density by continuously measuring different areas inside the chamber using a Horiba (MEXA-311GE) CO-CO2 analyzer. The desired CO concentration of the gas mixture was achieved by combustion of cigarettes from various popular brands. In the control trial, the air inside the chamber was identical to normal sea-level environmental conditions (O2: 20.93%; CO2: 0.04%; and N: 78.1) for the entire length of the trial.

Data Collection

Serum and urine cotinine.

Serum and urine cotinine, a preferable biomarker to carboxyhemoglobin for assessing SHS exposure (12), was measured before participants entered the chamber (i.e., baseline data) and immediately after (i.e., follow-up data) in both trials. Both serum and urine samples were obtained to ensure validity since, although urine is the specimen of choice for SHS exposures (urine cotinine concentration is ∼10× higher than in circulation), such samples may be subject to contamination.

For serum cotinine analyses, 5 ml of whole blood were used from the total 20 ml was collected by a certified phlebotomist from an antecubital vein into plain evacuated test tubes at baseline and follow-up in both trials. Immediate sample processing included allowing the blood to clot at room temperature (i.e., ∼23°C) for 30 min and centrifugation at 1,000 g for 10 min. The serum layer was then removed and frozen in multiple aliquots at −20°C until analyzed.

For urine cotinine analyses, the first morning urine void (80 ml) was collected in polyethylene specimen jars (Fisher Scientific, Pittsburgh, PA) at baseline in both trials. While inside the chamber, participants were advised to hydrate ad libitum with water. Another specimen of equal quantity was collected at follow-up in both trials. All samples were immediately frozen at −20°C until analyzed.

The serum sample preparation of cotinine biochemical analysis included a mixture of an aliquot (1 ml) of each sample with 1 ml of buffer solution (pH = 6.88). The urine sample preparation included mixture of an aliquot (5 ml) of each sample with 2.5 ml of buffer solution (pH = 6.88). All analyses were conducted via electron ionization mass spectrometric confirmatory analysis using a Finnigan Mat GCQ system equipped with an HP-5MSI (30 m × 0.25 mm ID × 0.25 μm) capillary column (J&W Scientific). Two microliters of each sample were injected into the system in the splitless mode. Analysis conditions were as follows: the column temperature program started from 90°C for 1 min and was raised to 280°C at the rate of 20°C/min. The injector temperature was 280°C. The transfer line temperature was set at 300°C. The mass spectrometer acquisition parameters were as follows: ion source: 200°C; electron impact ionization: 70 eV; and electron multiplier voltage: 1,200 V. The mass spectrometer was operated at the selected ion-monitoring mode and was programmed for the detection of a mass-to-charge ratio of 84 for nicotine (cotinine) and mass-to-charge ratio of 180,209 for ketamine. Under these conditions nicotine (cotinine) eluted at 6.16 min and ketanine at 10.15 min.

Gonadal hormones.

Blood collection and immediate sample processing for the determination of 17β-estradiol (E2), progesterone, and testosterone levels was identical to that previously described for serum cotinine. Analysis was conducted using 5 ml of whole blood (from the total of 20 ml collected). Gonadal hormones were measured in a TOSOH (Tokyo, Japan) enzyme immunoassay analyzer using reagents from the same company. Coefficient variance for all measurements was <5%.

Thyroid hormones, LH, and FSH.

Blood collection and immediate sample processing for the determination of fT4, T3, TSH, luteinizing hormone (LH), and follicle-stimulating hormone (FSH) were identical to that previously described for serum cotinine. In the current analysis, 10 ml of whole blood were used (from the total of 20 ml collected). Total fT4, T3, TSH, LH, and FSH levels were determined using an electrochemiluminescent immunoassay in a Hitachi E170 analyzer (RocheDiagnostics). Controls from Bio-Rad were used for fT4, T3, and TSH protocols (Bio-Rad laboratories, Irvine, CA). To estimate the proportion of active thyroid hormone (T3) relative to its precursor (fT4), the serum concentrations of T3 were divided by those of fT4. To avoid the use of many decimal points, this relationship was expressed as a percent [T3-to-fT4 ratio = (T3 × 100)/fT4]. Concentrations of LH and FSH were measured in a TOSOH enzyme immunoassay analyzer using reagents from the same company. Coefficient variance for all five measurements was <5%.

Inflammatory markers.

Analysis for IL-1β, IL-6, and TNF-α levels was conducted using 5 ml of whole blood (from the total of 20 ml collected) after blood collection and immediate sample processing identical to that previously described for serum cotinine. IL-6 and TNF-α levels were measured using commercially available enzyme immunosorbent assay kits (Biosource Europe) according to the manufacturer's protocol. The lower limits of detection for IL-6 and TNF-α were 2 and 3 pg/ml, respectively. The reproducibility was assessed by two consecutive measurements on the same day in samples from a subgroup of 15 individuals.

Parameters of vascular function.

Measurements for both trials were conducted after the participant was resting for 5 min. Heart rate was obtained every 10 s for a total duration of 1 min via a portable telemetric devise (Polar Electro, Kempele, Finland), and the mean of all measurements was recorded. Systolic and diastolic arterial blood pressures were measured three times (the mean of all measurements was recorded) by the same trained investigator using the same mercury sphygmomanometer (AS007). The arterial blood pressure readings were used to calculate mean arterial pressure {i.e., diastolic + [0.333 × (systolic − diastolic)]}.

Statistical Analyses

Sample size calculations were conducted based on values from an SHS experiment that reported data separately for men and women (14). The resulting minimum required sample size was 20 for a 2-sided type 1 error of 5%. Preliminary data analyses included calculation of mean ± SD values of all the examined parameters for each measurement (i.e., baseline and follow-up) in both trials. To ensure that our results encapsulate possible decreased variability amongst repeated measurements within a single individual, we compared the mean difference in each measurement (i.e., follow-up and baseline) between the two conditions (i.e., experimental minus control) using paired t-tests (17). This approach is more appropriate than a repeated-measures analysis, as it isolates the SHS effects from normal diurnal variation and also takes into account that the change for a given participant may be less than the difference between individuals and allows for appropriately narrow measurement confidence intervals. Given the statistically significant effect of sex on T3 after the SHS exposure (see results), stepwise multiple linear regression analysis incorporating backward elimination at the P < 0.05 level was introduced to model the effect of cotinine levels, gonadal hormone secretion, and cytokine levels (independent variables) on T3 secretion (dependent variable). Given the effect of sex on systolic blood pressure after the SHS exposure (see results), the same regression procedure was also used to model the effect of cotinine levels, gonadal and thyroid hormone secretion, as well as cytokine levels (independent variables) on systolic blood pressure (dependent variable). To ensure that the comparatively long half-life of specific indexes (e.g., cotinine and fT4) did not influence our results, ANOVA was used to detect differences between the group of participants (n = 14) performing the experimental trial first in sequence and those (n = 14) performing the experimental trial second in sequence. The sample size calculations were conducted with PASS 2000 (Hintze J. Number Cruncher Statistical Systems, Kaysville, UT, 2001; www.ncss.com) software, while all other statistical analyses were carried out with SPSS (version 14, SPSS, Chicago, IL); the level of significance was set at P < 0.05.

RESULTS

Serum and urine cotinine levels (Table 1) were significantly increased (P < 0.001) in the experimental compared with the control trial in both sexes. As expected, significant sex variation was apparent in the majority of the gonadal hormones (P < 0.05). Testosterone (P = 0.019) and progesterone (P < 0.001) in men as well as E2 (P = 0.001) and progesterone (P < 0.001) in women were found significantly decreased after the SHS exposure (Table 2). In men, fT4 (P < 0.001) and T3 (P = 0.020) were increased, whereas the T3-to-fT4 ratio (P = 0.033) was statistically decreased after SHS. In contrast, only fT4 (P = 0.010) was significantly increased in women after SHS, while T3 (P = 0.074) and the T3-to-fT4 ratio (P = 0.314) remained comparatively constant, with sex-specific variation apparent in both fT4 and T3. Additionally, LH (P = 0.026) and FSH (P < 0.001) in women were significantly decreased after SHS. All studied inflammatory markers demonstrated increased values after the SHS exposure (Table 3), but only the increase in IL-1β in men reached statistical significance (P = 0.001). Although increased, heart rate and systolic and diastolic blood pressure (Table 3) were not significantly different in the experimental trial compared with the control (P > 0.05), but statistically significant sex variation was observed in systolic blood pressure in the follow-up change (P = 0.040). Finally, ANOVA detected no statistically significant differences (P > 0.05) between the groups of participants performing the experimental trial first or second in sequence in any of the studied parameters (Table 4).

View this table:
Table 1.

Cotinine concentration and thyroid hormone secretion for men and women in each trial and the calculated change between trials (i.e., experimental minus control)

View this table:
Table 2.

Gonadal hormones for men and women in each trial and the calculated change between trials (i.e., experimental minus control)

View this table:
Table 3.

Inflammatory cytokines and vascular tone parameters for men and women in each trial and the calculated change between trials (i.e., experimental minus control)

View this table:
Table 4.

Cotinine levels, thyroid hormone secretion, gonadal hormones, and vascular tone parameters for the participants performing the experimental trial first and second in sequence

The stepwise multiple linear regression analysis for T3 secretion [R2 = 0.32; F(8,103) = 6.011; P < 0.001] demonstrated statistically significant associations with urine cotinine (β = 0.008; P = 0.004), E2 (β = −0.001; P = 0.044), testosterone (β = 0.000; P = 0.014), fT4 (β = 0.001; P = 0.032), FSH (β = −0.028; P = 0.015), IL-1β (β = 0.005; P = 0.010), heart rate (β = 0.003; P = 0.021), and systolic blood pressure (β = −0.004; P = 0.047). When the same statistical method was used with systolic blood pressure as a dependent variable [R2 = 0.66; F(19,101) = 19.649; P < 0.001], results demonstrated significant associations with serum (β = 0.152; P = 0.05) and urine (β=0.268; P = 0.04) cotinine, E2 (β = −0.036; P < 0.001), testosterone (β = −0.03; P = 0.004), fT4 (β = −0.114; P < 0.001), the T3-to-fT4 ratio (β = −9.59; P = 0.012), TNF-α (β = 0.289; P = 0.006), heart rate (β = 0.171; P < 0.001), and diastolic blood pressure (β = 0.445; P < 0.001).

DISCUSSION

This is the first study to directly investigate sex-specific differences in the effects of SHS on gonadal and thyroid hormones, inflammatory markers, and parameters of vascular function. Our results suggest that moderate SHS exposure is accompanied by a decrease in gonadal hormones in both sexes as well as marked increases in thyroid hormone secretion and IL-1β production in men. This sexual dimorphism in the effects of SHS on thyroid hormone secretion and IL-1β production is also associated with a significantly increased systolic blood pressure after exposure to SHS in men compared with women.

The lower T3-to-fT4 ratio, despite the increase in T3, after the SHS exposure suggests inhibited extrathyroidal T4 deiodination to T3, resulting in larger amounts of unbound hormones that are available for uptake into cells and interaction with nuclear receptors, as well as a smaller circulating hormone storage pool. In turn, this may be due to a lower circulation of binding proteins or interaction of binding proteins with fT4 and T3 (6). The absence of a significant change in the T3-to-fT4 ratio after the SHS exposure in women suggests the involvement of estrogens, which are known to influence serum total T4 and T3 concentrations by increasing the glycosylation of thyroxine-binding globulin, a protein heavily involved in T4 and T3 binding, and by slowing its clearance from the blood (6). Hence, the present association between E2 and the T3-to-fT4 ratio in the entire sample as well as the negative association between E2 and fT4 in women reflect an increase in thyroxine-binding globulin concentration. In turn, this increase in hormone binding reduces fT4 concentration by increasing the clearance of thyroxine. Analogously, the statistically significant decrease in the T3-to-fT4 ratio after the SHS exposure in men as well as the inverse association between testosterone and T3 suggests the involvement of androgens, which are known to decrease the glycosylation of thyroxine-binding globulin (6). However, it is important to also consider that the present results are not limited to extrathyroidial processes, as fT4 levels are not significantly influenced by changes in the serum concentrations of binding proteins, reflecting a purely functional thyroid state (16). Therefore, the observed effects of gonadal hormones may also indicate a downregulating effect on thyroid gland hormonogenesis. Based on these results and data from chronic active smoking, it could be suggested that chronic passive smoking (lifestyle incorporating frequent exposures to passive smoke) may have clinical implications such as thyroid and gonadal abnormalities in both sexes.

The observed effects of SHS on thyroid hormone secretion may also reflect the involvement of circulating inflammation markers, and particularly IL-1β, which have been shown to influence the hypothalamic-pituitary-adrenal axis (23, 25). IL-1β inhibits differentiated thyroid cell functions (24), including human thyroid cell adenylate cyclase and thyroglobulin release (23, 25). Consistent with this notion is the finding that IL-1β in our experiment demonstrated significant SHS-induced changes only in men, reflecting an anti-inflammatory effect of estrogens, which is well supported by previous research (28). The present SHS-induced increase in IL-1β extends the findings of a small number of previous human experiments reporting a SHS-induced increase of circulating inflammation markers including white blood cell count, C-reactive protein, homocysteine, fibrinogen (19), and leukocyte counts accompanied by an activation of the immune cells (1). In fact, given that animal models have demonstrated that the SHS-induced inflammatory reaction is dependent on IL-1β synthesis (3), it is surprising that previous human experiments did not evaluate IL-1β. This cytokine is a potent inflammatory mediator, stimulating chemokine production, recruiting leukocytes to the site of injury, and inducing the synthesis of TNF-α and IL-6 (10). In the present experiment, we observed SHS-induced increases in TNF-α and IL-6, which, however, did not reach statistical significance. Given that IL-1β up-regulates metalloproteinases (27) and fibroblast proliferation (22), features that are closely associated with the chronic inflammation and structural changes observed in patients (27), it could be suggested that chronic passive smoking may have clinical implications, especially in men, such as increased susceptibility to infection, chronic lung inflammation as well as pathological airway changes including chronic obstructive pulmonary diseases. Further, based on evidence from the mouse model showing reduction in SHS-induced inflammation, by neutralizing IL-1β (3), it could be cautiously suggested that strategies to inhibit IL-1β may have therapeutic benefit in chronic SHS-induced abnormalities.

Although the increase in systolic blood pressure after SHS was not statistically significant, the SHS-induced change was statistically significant in men compared with women, confirming the findings of a recent study (14) that used similar SHS exposure levels and duration. Sexual dimorphism in the effects of tobacco smoke has also been reported for branchial flow-mediated dilatation (4) as well as for intimamedia thickness (8). The current increase in vascular tone may be due to the confirmed smoke-induced sympathetic nervous system activation (11). Given the significant effect of thyroid hormone regulation on the sympathetic nervous system (21), it is not surprising that one of the parameters associated with systolic blood pressure in our study was the T3-to-fT4 ratio. On the other hand, our regression analysis demonstrated that gonadal hormones were also associated with systolic blood pressure. The negative link between both major gonadal hormones examined herein (i.e., E2 and testosterone) and systolic blood pressure confirms their well-known cardioprotective role against the noxious effects of SHS (14, 15, 29). Concomitantly, the observed link between systolic blood pressure and TNF-α supports a recent study (32) proposing a link between the SHS-induced increase in TNF-α and cardiovascular alterations leading to atherogenesis. Active cigarette smoking has also been associated with an imbalance in the production of TNF-α and soluble TNF receptors, leading to a relative excess of TNF-α (9). This may also explain the small SHS-induced increase in TNF-α in our experiment.

It is important to note that the current alterations in gonadal and thyroid hormone secretion, inflammatory markers, and systolic blood pressure in response to SHS pertain to acute changes and did not arise from an extreme and/or prolonged exposure. The reported cotinine levels suggest a moderate and brief SHS exposure (2), confirming a successful simulation of a bar/restaurant smoking environment. Further, the adopted design did not influence our results, as no differences were detected between participants performing the experimental trial first or second in sequence. It is concluded that a 1-h SHS exposure to levels similar to those of bars/restaurants is accompanied by a decrease in gonadal hormones in both sexes as well as marked increases in thyroid hormone secretion, IL-1β production and systolic blood pressure in men. These findings are alarming since recent reports (30) show that, despite currently adopted measures, the prevalence rates of smoking are increasing.

DISCLOSURES

This study did not receive funding contribution and/or any influence from sponsors, and the researchers declare full independence from external parties and other conflicts of interest related to this work.

Acknowledgments

We thank all participants for their time and enthusiasm; M. N. Tzatzarakis and A. M. Tsatsakis at the Centre of Toxicology Science and Research, School of Medicine, University of Crete, for conducting the cotinine analyses; and A. Germenis at the School of Health Sciences, Department of Medicine, University of Thessaly, for conducting the analysis on cytokines.

Footnotes

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

REFERENCES

View Abstract