Associations of Heart Rate With Inflammatory Markers Are Modulated by Gender and Obesity in Older Adults

Abstract

Background.

Faster resting heart rate (HR), which is associated with inflammation and elevated cortisol levels, is a risk factor for excess cardiovascular morbidity and mortality. Obesity is associated with increased cardiovascular morbidity and mortality, inflammation, and elevated cortisol levels. The aim of the present study was to evaluate the interaction of Body Mass Index (BMI) with inflammation and cortisol in modulating HR in older subjects.

Methods.

We analyzed data of 895 participants aged 65+ enrolled in the “InCHIANTI” study, in sinus rhythm, and not taking beta blockers or digoxin. Linear regression was performed to assess the adjusted association between HR, IL-6, and cortisol levels. The model was also analyzed stratifying for BMI tertiles. Logistic regression was adopted for evaluating the association of HR exceeding the mean value with Il-6 and serum cortisol.

Results.

According to multivariable linear regression, IL-6 and cortisol levels were associated with HR (B = 1.42, 95% CI = 0.43–2.42; p = .005 and B = .34, 95% CI = 0.17–.51; p < .0001, respectively). The association was significant only among women in the highest BMI tertile (B = 4.16, 95% CI = 1.40–6.91; p = .003 for IL-6 and B = .57, 95% CI = 0.14–1.01; p = .010 for cortisol). Logistic regression confirmed that IL-6 and cortisol levels were associated with HR above the mean value in the highest BMI tertile (OR = 2.13, 95% CI = 1.15–3.97; p = .009 and OR = 1.14, 95% CI = 1.03–1.25; p = .009, respectively).

Conclusions.

Faster HR is associated with proinflammatory state in elderly patients; this association seems to be limited to women with higher BMI.

Several epidemiological studies have reported that higher resting heart rate (HR) predicts cardiovascular morbidity and mortality in general populations, independently of blood pressure and serum cholesterol levels; the risk of cardiovascular events has been found to increase even for moderate levels of tachycardia, ie for HR>80 beats per minute (bpm [1–3]). However, similar levels of HR (ie beyond 80 bpm) have also been associated with increased risk of noncardiovascular events, such as hip fracture, as well as all-cause mortality (3). Thus, faster resting HR might not directly increase the risk of cardiovascular events, as suggested by some authors (2); rather, it could reflect some underlying condition also associated with noncardiovascular events.

Interestingly, increased resting HR has also been reported among patients with chronic inflammatory conditions, such as rheumatoid arthritis (4). Inflammation, along with increased production of inflammatory cytokines, is known to increase the risk of both cardiovascular and noncardiovascular morbidity and mortality, particularly among older persons (5–8). On the other hand, in vitro and in vivo studies suggested that inflammatory cytokines might increase HR (9,10).

Also, it has been reported that elevated cortisol levels, as in Cushing’s syndrome or during physical or psychological stress, have been associated with autonomic imbalance and alteration in HR variability (11).

Thus, we assessed in a general older population the hypothesis of an independent association between serum levels of inflammatory markers and resting HR. Also, as obesity has been associated with both increased resting HR and inflammation, we verified whether such an association might be influenced by body mass index and related comorbility. This topic is of particular interest because overweight has been proven a risk factor for mortality in middle-aged, but not in older subjects (12).

Methods

Participants

The present study is based upon data from the “Invecchiare in Chianti” (Aging in the Chianti Area, InCHIANTI) study, a prospective population-based study of older persons in Tuscany, Italy. The InCHIANTI study aims to identify risk factors for late-life disability (13).

Briefly, participants were selected from the city registries of Greve in Chianti and Bagno a Ripoli using a multistage sampling method. In 1998, 1155 subjects aged 65+ randomly selected from the population agreed to participate in the project (91.6% response rate). Three and 6 years after the baseline visit (2001–2003 and 2004–2006), study participants underwent repeated phlebotomy, laboratory testing, and physical performance assessment.

The Italian National Research Council on Aging Ethical Committee ratified the study protocol, and participants provided written consent to participate.

At baseline, analyses included 895 participants aged 65 plus; 154 participants had been excluded because of missing data for study variables, or rhythm other than sinus (n = 49), or treatment with beta blockers (n = 19), or digoxin (n = 38). No subjects were taking dihydropyridine calcium channel blockers.

Heart Rate

A standard resting 12-lead electrocardiogram (ECG) was performed in basal, standardized conditions during the technical evaluation session and interpreted by a cardiologist.

Inflammation Indices

Cortisol was assessed by a radioimmunoassay (RIA) method using a commercial kit (Active Cortisol RIA, DSL-2100; Diagnostic Systems Laboratories, Webster, TX).

Intra-assay coefficients for three different concentrations(low, medium, high) ranged between 8.4%and 11.1%. The interassay coefficients ranged between 9.1%, 8.9% and 11.5%.

Interleukin-6 was assessed by an ultra-sensitive enzyme-linked immunosorbent assay (ELISA) using a commercial kit (CytoScreen Human IL-6, Biosource International, Camarillo, CA). The minimum detectable threshold was 0.10 pg/mL. The interassay coefficient was 7%.

Covariates

Education was expressed as years of school attendance. Smoking was self-reported and expressed as total lifetime pack-years (packs smoked per day)*(years of smoking).

Weight loss in the last 12 months was investigated by the specific question: “Have you lost weight in the last 12 months?”.

Diseases were ascertained by experienced clinicians according to pre-established criteria that combined information from self-reported physician diagnoses, current pharmacologic treatment, medical records, clinical examinations, and blood tests. Diagnostic algorithms for diseases were modified versions of those created for the Women’s Health and Aging Study (14).

All drugs assumed by participants were coded according to the Anatomical, Therapeutic and Chemical codes (15).

Data on dietary intake were collected by the food-frequency questionnaire created for the European Prospective Investigation into Cancer and nutrition (EPIC) study (16). Specific software created for the EPIC study transformed data on food consumption into daily intake of energy, macronutrients, and micronutrients. Weight was measured to the nearest 0.1kg using a high-precision mechanical scale and standing height to the nearest 0.1cm based on wall measure with participants wearing light indoor clothes and no shoes. BMI was calculated as weight (kg) divided by height squared (m²).

All blood samples were obtained from participants in similar conditions, early in the morning, after at least 8-hour fasting and after resting for at least 15 minutes. Aliquots of serum were stored at –80°C and were not thawed until analyzed.

Usual physical activity was self-reported and defined as walking inside the house for at least 3 hours/day.

The Short Physical Performance Battery (SPPB) based on the lower-extremity performance tests was used to summarize objective physical performance (17).The test consists of three components: balance, timed 4-m walk, and chair stands. The standing balance portion required participants to maintain a side-by-side, semitandem, and tandem stance for 10 seconds with scores ranging from 0 to 4 (maximum score). The fastest time of two 4-m usual-pace walk attempts was used. The chair stands required participants to rise from a chair with arms across their chest for five repetitions. The sum of the three components yielded the final SPPB score which ranged from 0 to 12 (12 indicating the highest degree of lower extremity functioning).

Statistical analyses

Statistical analyses were performed using Statistical Package for the Social Sciences (SPSS for Mac version 20.0, 2011, SPSS Inc, Chicago, IL); differences were considered significant at the p < .050 level. Data of continuous variables are presented as mean values ± standard deviation (SD). Analysis of variance for normally distributed variables according to resting HR above the mean value was performed by ANOVA comparisons; otherwise the nonparametric Mann–Whitney U test was adopted. Chi-square analysis with two-tailed Fisher’s Exact Test was used for dichotomous variables. Serum IL-6 levels and smoking were analyzed after log transformation. The covariates to be entered into analyses were chosen according to available evidence from the electronic databases of PubMed (MEDLINE) and Cochrane Library.

Multivariable linear regression analysis was adopted to estimate the association of HR with age, sex, and all those variables which differed significantly according to the mean value of HR (p < .050). Multivariable logistic regression analysis was used to assess the association of HR above the mean value with the same set of variables. Also, the effect of BMI on the association of HR with IL-6 and serum cortisol was assessed by analysis of the interaction term.

To rule out any differences in results due to overall versus central obesity, we analysed the same full adjusted model after entering waist circumference instead of BMI.

The same multivariable linear regression model was analyzed after stratifying for BMI tertiles by sex; analysis was repeated considering only subjects who denied any weight change over the last year. This latter analysis was performed in order to correct for the potential confounding by weight loss and underlying conditions. Also, the multivariable regression model was analyzed by logistic regression, adopting a HR above the mean value as dependent variable, stratifying for BMI tertiles which had been calculated according to sex.

In addition, the same multivariable logistic regression model was analyzed using a HR>80 bpm, which has repeatedly been shown to be associated with increased mortality, as the dependent variable.

The linear regression model was also analyzed after stratification for HR tertiles.

Results

The mean HR was 69 bpm. The main characteristics of participants according to the presence of HR above the mean value are depicted in Table 1. The distribution of gender in the study population differed significantly according to HR grouping (273 women and 157 men in the group of HR above the mean value vs 230 women and 235 men in controls, p >.0001). In addition, participants with HR above the mean value, as compared with other subjects, were more likely to be women, had more prevalent diagnosis of heart failure and diabetes; they took less frequently alpha-blockers, and they showed higher serum cortisol, IL-6, and BMI, but lower glomerular filtration rate, serum potassium, and SPPB. Values of BMI tertiles were as follows: lowest tertile was defined by a BMI lower than 25.5 kg/m² in men and 24.48 kg/m² in women; the mid tertile by a BMI between 25.5 and 28.34 kg/m² in men and between 24.48 and 29.51 kg/m² in women; the highest tertile was defined by a BMI higher than 28.34 kg/m² in men and 29.51 kg/m² in women. Women in the highest BMI tertile had a significant higher mean BMI, as compared with men (32.8±3.2 kg/m² vs 30.8±1.9 kg/m², p < .0001)

Table 1.

Characteristics of Participants Derived From the InCHIANTI Study Grouped According to the Mean Value of Heart Rate

	Heart Rate > 69 bpm (n = 430) n (%) or mean ± SD	Heart Rate ≤ 69 bpm (n = 465) n (%) or mean ± SD	p
Demographics and lifestyle habits
Age (years)	74±7	74±6	.090
Sex (female)	273 (63%)	230 (49%)	<.0001
Education (years)	5±3	6±3	.474
Smoking*	11±21	13±20	.017
Usual physical activity†	244 (57%)	240 (52%)	.140
Energy intake (kcal/die/kg)	28.4±8.6	29.0±8.3	.306
Vitamin C intake (mg/die/kg)	1.6±0.7	1.7±0.7	.093
PUFA intake (g/die/kg)‡	0.10±0.03	0.11±0.03	.545
Comorbid conditions
Chronic pulmonary disease	12 (3%)	12 (2%)	.999
Coronary disease	28 (6%)	25 (4%)	.482
Heart failure	19 (4%)	6 (1%)	.007
Parkinson’s disease	5 (1%)	4 (1%)	.745
Diabetes	66 (15%)	40 (9%)	.002
Disthyroidism	36 (8%)	39 (8%)	.999
Peripheral arterial disease	71 (16%)	87 (19%)	.430
Arthritis	126 (29%)	145 (31%)	.561
Vascular dementia	8 (2%)	6 (1%)	.594
Medications
Alpha-blockers	4 (1%)	14 (3%)	.031
Nitrate	20 (5%)	16 (3%)	.397
Clonidine	0	1 (1%)	.999
Objective tests
Glomerular filtration rate (mL/min)	64.9±18.2	67.5±19.4	.041
Hemoglobin (d/dL)	13.7±1.4	13.8±1.3	.121
White blood cell count (n, K/µL)	6.15±1.30	5.97±1.48	.071
Serum cortisol (µg/dL)	13.8±4.8	12.8±4.3	.001
Interleukin 6 (pg/mL)	2.1±2.4	1.9±3.9	.004
Serum sodium (mEq/L)	142±2	142±3	.305
Serum potassium (mEq/L)	4.2±0.4	4.3±0.3	.004
Serum calcium (mg/dL)	9.4±0.5	9.4±0.4	.143
SPPB§	9±3	10±2	<.0001
Body mass index	27.8±4.3	27.2±3.8	.030

Data Collection Was Performed in 1998.

*Total lifetime pack years.

^†Walking inside the house for at least 3 hours/day.

^‡Polyunsaturated fatty acids.

^§Short Physical Performance Battery.

No weight loss in the last 12 months was reported by 690/895 (77%) of subjects.

Values of HR tertiles were as follows: lowest tertile was defined by a HR lower than 64 bpm; the mid tertile by a HR between 64 and 74 bpm; the highest tertile was defined by a HR higher than 74 bpm.

Multivariable Analyses

According to linear regression, IL-6 and cortisol levels were positively associated with HR in the unadjusted model (B = 1.47; 95% CI = 0.53 to 2.40; p = .002 and B = 0.36; 95% CI = 0.19 to 0.52; p < .0001, respectively), after adjusting for age and sex (B = 1.80; 95% CI = 0.84 to 2.76; p < .0001 and B = 0.37; 95% CI = 0.20 to 0.53; p < .0001, respectively), as well as in the multivariable model (B = 1.42, 95% CI = 0.43–2.42; p = .005 and B = 0.34, 95% CI = 0.17 to 0.51; p < .0001, respectively) adjusting for those variables which distinguished subjects grouped according to mean HR at a p < 0.05 level in univariate analyses (ie smoking, diagnosis of heart failure, and diabete, use of alpha-blockers, serum potassium levels, SPPB, and BMI) (see Table 2).

Table 2.

Association (B coefficients and 95% CI) of Heart Rate With the Variables of Interest in the Multivariable Regression Model

	B	95% CI	p
Age (years)	−0.03	−.19 to .13	.682
Sex (female)	4.24	2.27 to 6.20	<.0001
Smoking*	0.01	–.03 to .06	.589
Heart failure	3.50	–1.12 to 8.13	.137
Diabetes	4.25	1.90 to 6.61	<.0001
Alpha-blockers	–3.23	–8.49 to 2.02	.228
Glomerular filtration rate (mL/min)	–0.06	–.12 to .01	.064
Serum cortisol (µg/dL)	0.34	.17 to .51	<.0001
Interleukin 6 (pg/mL)†	1.42	.43 to 2.42	.005
Serum potassium (mEq/L)	–2.13	–4.37 to .11	.062
SPPB‡	0.04	–.32 to .41	.806
Body mass index	0.29	.06 to .51	.013

All the covariates were entered simultaneously into the regression model. Participants derived from the InCHIANTI study; data collection was performed in 1998. CI = confidence interval.

*Total lifetime pack years.

^†Log-transformed.

^‡Short Physical Performance Battery.

When waist circumference was entered instead of BMI in this fully adjusted model, IL-6 and cortisol levels were still associated with HR (B = 1.44, 95% CI = 0.43–2.45; p = .005 and B = .35, 95% CI = 0.18–0.52; p < .0001, respectively).

When this summary linear regression model was analyzed after stratification for BMI tertiles by sex, IL-6 and cortisol levels were still associated with HR only in the highest BMI tertile (B = 3.16, 95% CI = 1.18–3.13; p = .002 and B = 0.49, 95% CI = 0.20–0.78; p = .001, respectively). No significant association was found in the lowest (B = 0.94, 95% CI = –1.01 to 2.89; p = .343 for IL-6 and B = 0.21, 95% CI = –0.13 to .56; p = .226, for cortisol levels), nor in the mid BMI tertile (B = 1.18, 95% CI= –0.33 to 2.70; p = .126 for IL-6 and B = 0.37, 95% CI = 0.03–0.71; p = .033, for cortisol levels).

In addition, subanalyzes including only subjects who denied weight loss in the last 12 months confirmed that only in the highest BMI tertile, HR was associated with IL-6 and cortisol levels (B = 2.99, 95% CI = 0.60–5.38; p = .014 for IL-6 and B = 0.26, 95% CI = –0.01 to 0.52; p = .056, for cortisol levels, respectively).

When the multivariable linear model was analyzed after stratifying for sex, no significant association was found between HF and IL6 levels (B = 1.35, 95% CI = –0.19 to 2.88; p = .086), nor with cortisol levels (B = 0.23, 95% CI = –0.02 to 0.47; p = .072) in men; instead, in women we found a significant association of HR with IL-6 (B = 1.40, 95% CI = 0.01 to 2.74; p = .040) as well as cortisol levels (B = 0.46, 95% CI = 0.22–0.69; p < .0001).

When this summary linear regression model was analyzed after stratification for BMI tertiles in women, IL-6 and cortisol levels were still associated with HR only in the highest BMI tertile (B = 4.16, 95% CI = 1.40–6.91; p = .003 and B = 0.57, 95% CI = 0.14– 1.01; p = .010, respectively). No significant association was found in the lowest, nor in the mid BMI tertiles.

Analyzis of the interaction term in logistic regression confirmed that the association of HR above the mean value with IL-6 (p = .047) and cortisol levels (p = .011) varied according to BMI.

Logistic regression analysis confirmed that HR above the mean value was associated with IL-6 (OR= 1.75, 95% CI = 1.16–2.64; p = .007) and cortisol levels (OR = 1.09, 95% CI = 1.02–1.15; p = .009) only in the highest BMI tertile (see Table 3). Again, this association was circumscribed to women in the highest BMI tertile (OR = 2.13, 95% CI = 1.15–3.97; p = .009; for Il-6 and OR = 1.14, 95% CI = 1.03–1.25; p = .009 for cortisol levels).

Table 3.

Association (OR and 95% CI) of Heart Rate Above the Mean Value With Variables of Interest According to the Multivariable Regression Model After Stratifying for BMI Tertiles by Sex

	Lowest BMI Tertile		Medium BMI Tertile		Highest BMI Tertile
	OR	95% CI	OR	95% CI	OR	95% CI
Age (years)	0.98	0.93–1.04	1.00	0.95–1.06	0.95	0.89–1.01
Sex (female)	1.50	0.70–3.22	1.16	0.64–2.13	2.75	1.38–5.48
Smoking*	1.00	0.98–1.01	0.99	0.98–1.01	1.01	0.99–1.03
Heart failure	0.58	0.03–10.32	3.79	0.73–19.65	3.25	0.71–14.89
Diabetes	3.04	1.15–8.03	1.31	0.61–2.83	2.78	1.27–6.07
Alpha-blockers	0.26	0.03–2.51	0.22	0.02–2.28	0.25	0.04–1.54
GFR (mL/min)†	1.01	0.97–1.02	0.98	0.96–1.01	0.99	0.97–1.01
Serum cortisol (µg/dL)	0.99	0.94–1.06	1.04	0.99–1.10	1.09	1.02–1.15
Interleukin 6 (pg/mL)‡	1.28	0.90–1.82	0.98	0.73–1.32	1.75	1.16–2.64
Serum potassium (mEq/L)	0.31	0.13–0.77	0.53	0.27–1.06	1.12	0.51–2.44
SPPB§	0.87	0.76–1.01	0.93	0.83–1.05	0.97	0.86–1.10
Body mass index	1.03	0.85–1.24	1.05	0.87–1.28	1.08	0.97–1.19

All the covariates were entered simultaneously into the regression models. Participants derived from the InCHIANTI study; data collection performed in 1998. BMI = body mass index; CI = confidence interval; OR = odds ratio.

*Glomerular Filtration Rate (mL/min).

^†Total lifetime pack years.

^‡Log-transformed.

^§Short Physical Performance Battery.

In the multivariable logistic regression model having HR>80 bpm as dependent variable, faster HR was associated with IL-6 (OR = 1.35, 95% CI = 1.04–1.75; p = .023) and cortisol levels (OR = 1.08, 95% CI = 1.04–1.13; p < .0001; Appendix 1).

In the multivariable linear regression model analyzed after stratification for HR tertiles, IL-6 and cortisol levels were still associated with HR only in the highest HR tertile (Appendix 2).

Discussion

Results of this population-based study indicate that higher resting HR in older subjects is independently associated with higher circulating levels of IL-6 and cortisol; however, this association seems to be limited to women with higher BMI.

Over the recent years, the relevant incidence of major cardiovascular events among subjects with normal serum cholesterol levels prompted research on other determinants of atherosclerosis and its complications (6,18). In this setting, faster HR has repeatedly been proven an independent predictor of cardiovascular and noncardiovascular morbidity and mortality (1–3,19). This association has been ascribed either to increased sympathetic output (3), or to excess mechanical stress on the coronary artery wall (2). However, some observations are not easily explained by these hypotheses. In most epidemiological studies, the risk of cardiovascular and noncardiovascular events has been found to increase for HR levels just above 80 bpm (3,19,20). Thus, even subjects whose HR is below the upper “normal” limit of 90 bpm are at higher risk. Most recently, an increased risk of mortality has been reported for HR > 70 bpm, with an 11% increase in the risk for each 10 beats increase (19,21). This indicates that the higher event rate is not due to HR per se, but rather to cardiovascular risk factors which are reflected by HR. Our study suggests that subclinical inflammation, as reflected by increased Il-6 and cortisol serum levels, might be the link between increased HR and incident cardiovascular events. In fact, higher serum IL-6 levels have been associated with faster HR even in transplanted hearts (22); also, addition of IL-6 has been found to increase the spontaneous beating rate of myocytes in serum-free medium (9). Accordingly, administration of IL-6 has been reported to increase HR in humans (10).

On the other hand, both IL-6 and cortisol levels have been associated with increased risk of unstable angina and cardiovascular mortality in large studies (5,7). Indeed, all available evidence supports a central role for inflammation in all phases of the atherosclerotic pathway, from early atherogenesis to its thrombotic complications (6,23). Likewise, increased IL-6 and cortisol levels also predict noncardiovascular events, such as hip fracture, as well as disability, and noncardiovascular mortality (3,7).

Of notice, serum proinflammatory cytokines and corticosteroids have been found to increase HR, either directly or through increased sympathetic output. However, it has also been hypothesized that autonomic dysregulation might yield proinflammatory effects (24,25); due to its design, our study does not allow to clarify the cause-effect relationship.

Of interest, the association of inflammation with HR seems to be limited to females in the highest tertile of BMI. This might simply reflect the higher prevalence of severe obesity among females.

Alternatively, the lower prevalence of men whose HR exceeded the mean value might account for our results due to insufficient power of this group. Indeed, it has been hypothesized that obesity corresponds to a subclinical inflammatory condition that might promote the production of pro-inflammatory factors (26) and that cardiac autonomic imbalance would correlate with the adipose tissue-derived inflammation (27). It has also been reported that central obesity may play a prominent role as compared with overall obesity in general population (28), but our results did not support this hypothesis. Furthermore, we only assessed the association between HR and inflammatory indices in a cross-sectional model. This hinders any interpretation of our data from the perspective of the obesity paradox.

In-line with previous studies, we found an independent association between diabetes and HR. Although an additive effect of diabetes and obesity on overactivity of the sympathetic system has been reported (29), our data suggest that the relationship between diabetes and HR is not mediated by BMI.

Strengths and Limitations

Due to its cross-sectional design, the present study does not allow to assess any cause-effect relationships. The nutrient intake was normalized by body weight (see Table 1); this might underestimate the real intake in obese subjects. However, this renders our analyses more conservative. Also, although analyses have been corrected for multiple confounders, we cannot exclude other variables potentially associated with HR, such as psychological or genetic factors. Nonetheless, this study includes a representative community-dwelling population, with high participation rate and with extensive information regarding risk factors, comorbid conditions, and objective parameters.

Conclusions

In summary, our data raise the hypothesis of an independent association between faster HR and proinflammatory state. This association seems to be limited to women with higher BMI. Thus, the simple measurement of resting HR might give information on proinflammatory state in elderly subjects. However, research is still needed to verify whether weight control might affect the inflammatory status in elderly obese women.

Funding

The InCHIANTI study baseline (1998–2000) was supported as a targeted project (ICS110.1/RF97.71) by the Italian Ministry of Health and in part by the U.S. National Institute on Aging (Contracts 263 MD 9164 and 263 MD 821336); the InCHIANTI Follow-Up 1 (2001–2003) was funded by the U.S. National Institute on Aging (Contracts N.1-AG-1-1 and N.1-AG-1-2111); the InCHIANTI Follow-Ups 2 and 3 studies (2004–2010) were financed by the U.S. National Institute on Aging (Contract N01-AG-5-0002).

Conflict of Interest

The authors declare that they have no conflict of interest in this study.