AJP - Heart AJP: Renal Physiology
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Am J Physiol Heart Circ Physiol 293: H1013-H1022, 2007. First published March 30, 2007; doi:10.1152/ajpheart.00475.2006
0363-6135/07 $8.00
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
293/2/H1013    most recent
00475.2006v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via Web of Science (3)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Uusitalo, A. L. T.
Right arrow Articles by Kaprio, J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Uusitalo, A. L. T.
Right arrow Articles by Kaprio, J.

Role of genetic and environmental influences on heart rate variability in middle-aged men

A. L. T. Uusitalo,1 E. Vanninen,1 E. Levälahti,4 M. C. Battié,2 T. Videman,3 and J. Kaprio4

1Department of Clinical Physiology and Nuclear Medicine, Kuopio University Hospital and University of Kuopio, Kuopio; 4Department of Public Health, University of Helsinki, and Department of Mental Health and Alcohol Research, National Public Health Institute, Helsinki, Finland; 2Department of Physical Therapy and 3Faculty of Rehabilitation Medicine, University of Alberta, Edmonton, Alberta, Canada

Submitted 9 May 2006 ; accepted in final form 27 March 2007


   ABSTRACT
TOP
 ABSTRACT
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 GRANTS
 REFERENCES
 
Our aim was to estimate causal relationships of genetic factors and different specific environmental factors in determination of the level of cardiac autonomic modulation, i.e., heart rate variability (HRV), in healthy male twins and male twins with chronic diseases. The subjects were 208 monozygotic (MZ, 104 healthy) and 296 dizygotic (DZ, 173 healthy) male twins. A structured interview was used to obtain data on lifetime exposures of occupational loading, regularly performed leisure-time sport activities, coffee consumption, smoking history, and chronic diseases from 12 yr of age through the present. A 5-min ECG at supine rest was recorded for the HRV analyses. In univariate statistical analyses based on genetic models with additive genetic, dominance genetic, and unique environmental effects, genetic effects accounted for 31–57% of HRV variance. In multivariate statistical analysis, body mass index, percent body fat, coffee consumption, smoking, medication, and chronic diseases were associated with different HRV variables, accounting for 1–11% of their variance. Occupational physical loading and leisure-time sport activities did not account for variation in any HRV variable. However, in the subgroup analysis of healthy and diseased twins, occupational loading explained 4% of the variability in heart periods. Otherwise, the interaction between health status and genetic effects was significant for only two HRV variables. In conclusion, genetic factors accounted for a major portion of the interindividual differences in HRV, with no remarkable effect of health status. No single behavioral determinant appeared to have a major influence on HRV. The effects of medication and diseases may mask the minimal effect of occupational loading on HRV.

autonomic function; heredity; twins; physical activity; smoking; coffee

HEART RATE (HR) variability (HRV) is defined as the oscillation in the interval between consecutive heartbeats, i.e., oscillation in heart periods (HP) (54). Cardiac autonomic modulation is the main regulator of HRV. Therefore, HRV can be considered an indirect indicator of quality and quantity of autonomic nervous system function, including components under vagal and sympathetic modulation (54). However, HRV may also reflect the effect of the renin-angiotensin system and thermoregulation on cardiovascular control (54).

Low HRV has been shown to be an independent predictor of overall mortality in elderly individuals (55), to precede ischemic events (20), and to be associated with increased risk for sudden cardiac death (19). Genetic factors play a role in many cardiovascular diseases; conditions that increase risk for cardiovascular disease, such as diabetes and the metabolic syndrome, also have a genetic component (27). HRV is associated with these diseases (26) and is also, to a large extent, genetically determined (6, 7, 21, 43, 45–47, 49, 62). According to these family and twin studies in young and middle-aged adults, the estimates of heritability, i.e., genetic variability expressed as a portion of the total variation in HRV, ranged from 13% to 65%. In the Framingham Study, a genome scan has identified chromosomal regions containing genes possibly contributing to HRV in 28- to 62-yr-old siblings (44). The exact genes determining HRV are not known, but some studies of candidate genes have been published (41, 48, 53).

In addition to the influences of genes, many environmental and behavioral factors affect HRV, but the extent of their effect in healthy and diseased individuals is unclear. Some endocrinologic and cardiovascular diseases change autonomic function. The best studied of these is diabetes (61). In addition to diabetes, cardiac failure and myocardial infarction have well-established influences on cardiac autonomic function (54). Some psychiatric diseases and their medications have obvious effects on autonomic function (8, 9). Also airway (16), kidney (33), and gastrointestinal tract (24) diseases and their medications may influence HRV.

Various medications related to chronic diseases also affect HRV. The best-reported primary effects are due to agonists or antagonists of autonomic receptors. The most widely used group is beta-blockers, which block sympathetic cardiac stimuli and, thereby, have a marked effect on HRV (3, 4). Of the other cardiovascular medications, we have no real evidence for an effect of Ca2+ antagonists and angiotensin-converting enzyme (ACE) inhibitors on HRV.

Physical activity or exercise training has been suggested to increase HRV (25, 40, 56), but opposite findings have also been reported (31, 51, 58, 59). A previous study in the same Finnish twins included in the present investigation estimated the heritability of adulthood exercise level to be 51% (42). In addition, genetic effects on exercise behavior and respiratory sinus arrhythmia have been reported to overlap in adolescent, but not middle-aged, twins (11). De Geus et al. (11) reported also that heritability has a larger role in determining respiratory sinus arrhythmia in middle-aged than in young twins, indicating that age needs to be taken into account in genetic studies of HRV and respiratory sinus arrhythmia.

Smoking, including environmental tobacco smoke, seems to acutely decrease HRV (35) and to persistently increase sympathetic activity (13). Coffee consumption increases HR and blood pressure, even in habitual users (23). In addition, some biological factors, such as body build and obesity, explain HRV (32, 36).

The aim of the present study was to investigate the relationships between HRV levels and environmental and somatic factors in a genetic epidemiological design. Because our population-based sample consisted of middle-aged and older men, another goal was to study whether chronic diseases interact with the genetic influences on HRV variables. We hypothesized that we would observe substantial genetic influences on HRV but that certain environmental and somatic factors, such as body mass index (BMI), medication, smoking, coffee consumption, and physical activity, would also have significant effects.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 GRANTS
 REFERENCES
 
General Study Design

The study is part of the Twin Spine Study (5), an investigation of the role of lifetime environmental exposures and genetics and their interactions on spine degeneration and function. A twin study design permits the estimation of overall genetic influences separately from other familial effects that are environmental influences shared by family members. The greater similarity of monozygotic (MZ) pairs than dizygotic (DZ) pairs is taken as evidence for genetic influences on the trait, and the quantitative contribution of genes can be estimated using a modeling approach. A central assumption in twin studies is that of "equal environment," which implies that MZ twins are as similar as DZ twins with respect to the environmental causes of HRV. If the assumption is violated, twin studies will overestimate, to some degree, the effect of genetic influences. When this assumption has been tested, it has been found to be valid for a wide range of behavioral and morphological traits.

Subjects

The Twin Spine study subjects are a total of 294 MZ and 306 DZ male twins (300 twin pairs). The men were selected from the Finnish Twin Cohort, which is a population-based cohort including all Finnish same-sexed twin pairs born before 1958 and alive in 1975 (17). Zygosity of the twins was determined using a validated questionnaire method, with a <1.7% estimated probability of misclassification of zygosity (39); zygosity was further confirmed by genotyping. Of the original subjects, 40 MZ pairs were studied before HP data were included in the study protocol, and 22 men were excluded because of missing covariates or otherwise missing data or technical problems. The HR data of 208 MZ (104 pairs) and 296 DZ (148 pairs) men were available for analysis (Table 1). All subjects received written information about the study procedures before participation and gave their written informed consent to take part in the study. The study protocols were reviewed and approved by the Ethical Committee of the Department of Public Health at the University of Helsinki and the Human Subjects Committee at the University of Washington (Seattle, WA).


View this table:
[in this window]
[in a new window]

  		Table 1. Characteristics of the study group

 
Data Acquisition

A structured interview was done, and demographic and HP data were collected during the same day and, in a few cases, on two consecutive days.

Environmental exposure data. A structured interview (2.5 h average) was used to obtain data on lifetime exposures of interest lasting ≥3 mo from 12 yr of age to the present (5, 42). Demographic information and health history, occupational physical loading, history of regularly performed sport activities and leisure-time physical activities, coffee consumption, and smoking history were obtained. For each job held for ≥3 mo, the subjects were asked to describe the job activities and estimate the most common weight lifted, the frequency of lifting, and the number of hours spent sitting. This information, along with the job title, was used to categorize the job in terms of its general demands related to material handling and postural stress as follows: not working and sedentary, light-mixed, heavy-mixed, or heavy job.

Mode, intensity, mean session duration, and frequency of leisure-time physical activity were queried (5, 42). Lifetime history of leisure-time physical activities (after 12 yr of age) and, in particular, ongoing activities was classified as aerobic sports, power sports, and other sports (Table 1). Regular aerobic sports during the lifetime were classified as the number of years (<1, 1–15, 16–30, and 31–60) of participation in aerobic exercise on average at least twice per week. Power sports and other sports were examined for each time period as dichotomous variables of ≥1 yr of participation of at least twice per week vs. less. The reliability of the data on physical activity obtained from the structured interview has been evaluated (37), and the intraclass correlations for the questions on total years of exercise and mean lifetime exercise (h/wk) were 0.69 and 0.73, respectively.

Coffee consumption was recorded as none and 1–3, 4–6, 7–10, and 11–30 cups/day and as no regular coffee consumption and 1–30, 31–40, 41–51, and 52–69 yr of coffee consumption. Cigarette smoking was calculated in pack-years, i.e., the mean number of cigarettes smoked daily multiplied by the number of years of smoking, and the subjects were classified as nonsmokers, ≤25 pack-yr, and >25 pack-yr.

Chronic diseases and medications were queried (Table 2). The wide range of diseases and all medications that have been reported to have some effect on HRV were included. The concordance rates of the diseases were generally not higher in MZ than in DZ twins. The concordance rates (percentages) of all reported diseases in the MZ and DZ twins were as follows: 38 MZ (37%) and 53 DZ (36%) concordant (C) pairs (range of C pairs in the different diseases 0–14 for MZ and 0–17 for DZ), 33 MZ (32%) and 35 DZ (24%) discordant (D) pairs (range of D pairs in the different diseases 1–22 for MZ and 1–20 for DZ), and probandwise concordance (PC) rates of 0.70 for MZ and 0.75 for DZ (range in the different diseases 0–0.70 for MZ and 0–0.63 for DZ); data on individual diseases are available from the authors on request. The concordance rates for the presence of any medication in MZ and DZ twins were 0.64 and 0.51, respectively, on the basis of 17 MZ (16%) and 14 DZ (9%) C and 19 MZ (18%) and 27 DZ (18%) DC pairs. Interviewers were blind with respect to the specific discordance or selection criteria for the twins, and project investigators have been careful to avoid discussions with the interviewers regarding the hypothesized effects of particular factors on the outcomes of the project. The same interviewer questioned each twin pair.


View this table:
[in this window]
[in a new window]

  		Table 2. Types of diseases and medication and their lifetime risk (prevalence) in MZ and DZ twins

 
Clinical examination. The examination includes basic anthropometric measures, such as weight (to nearest the kg on calibrated research scales in light clothing) and height (to the nearest cm on a stadiometer). BMI was calculated as kg/m2. Percent body fat was measured by bioelectrical impedance analysis (Spectrum II, RJL Systems, Detroit, MI), as described in detail by Simonen et al. (42).

HRV measures. The subjects entered the laboratory at 8–12 AM or 2–5 PM; both brothers of each twin pair were measured in the morning or afternoon after a ≥2-h fast. The subjects were instructed to consume a light meal before the fast. The subjects had not consumed coffee, cola beverages, tea, and chocolate for 12 h. Tobacco use was forbidden for 12 h and alcoholic beverages for 48 h before the measurements. The men were instructed to sleep enough during the previous night and avoid exercise and strenuous physical loads for 24 h before the measurements. The men took their normal medication. The room was silent, with dimmed lights and temperature of 20–22°C. After 2 min of supine rest, a 5-min ECG (MultiCare 302, Rigel Research, Surrey, UK) was registered with the subject at rest in the supine position with breathing frequency controlled at 0.20 Hz. Controlled breathing frequency during 5 min of rest in the supine position, instead of free breathing and ambulatory measurements in living conditions, was chosen to standardize the circumstances and the effects of posture, physical activity, and respiration on HRV measures. The heartbeat data were simultaneously converted from analog to digital (200 Hz, 12 bits) and saved on the hard disk of IBM PC/AT-compatible microcomputer for subsequent offline analysis. The software package CAFTS (Medikro, Kuopio, Finland) was used to evaluate HRV in time and frequency domains. HR, HP, the square root of the mean of the sum of the squares of differences (RMSSD) between adjacent HPs, low-frequency (LF, 0.04–0.15 Hz) and high-frequency power (HF, 0.15–0.40 Hz) of HP variability, and the LF-to-HF ratio (LF/HF) were calculated.

The QRS detection was completed with numerical derivation of the ECG signal followed by thresholding. In the last phase of the QRS detection, the temporal resolution of the R peaks and HPs was increased with a second-order polynomial-fit interpolation of each R wave. Spectral analysis of HRV was performed using a modified covariance autoregressive model with a fixed model order of 14. The regions of interest were selected by exclusion of ectopic beats and visual judgment of stationarity.

Statistical Analysis

Reliable HRV results for the statistical analysis were obtained from 225 of 252 twin pairs. Twenty-seven twin pairs were excluded from the final analysis because of arrhythmias or technically bad data of one or both twins. First, zero-skewness logarithmic transformation was used, because the HRV variables did not meet the criteria for normal distribution. Zero-skewness transformation is as follows: log(untransformed variable + k), where k is estimated from the data. All determinants, except percent body fat and BMI, which are continuous and were transformed, are binary or categorical variables, and some were formed by categorization of continuous variables (see above).

Pearson correlation coefficients were calculated to estimate the similarity of HRV variables in MZ and DZ twins as a measure of the contribution of genetic effects for HRV variables.

Tests of homogeneity of means and variances across twin type and twins were carried out using STATA SVY procedures for complex survey data, because the subjects were sampled as twin pairs. We used maximum likelihood based on sample covariance matrices and means, as described in detail elsewhere (28), to estimate genetic and environmental components of variance for HRV variables. Theoretically, four separate parameters (Fig. 1) can be modeled: additive genetic (A) effects, dominance genetic (D) effects, and shared (C) and nonshared, i.e., unique environmental (E) effects, but C and D cannot be estimated simultaneously using data on twins alone. The most parsimonious model (AE, ACE, ADE, or E) based on the possible combinations of the four parameters for each of the HRV variables was tested using {chi}2 goodness-of-fit statistics. Akaike's information criterion was used throughout the analysis to evaluate the relative fit of the various models. A lower Akaike information criterion indicates better fit. Under the current study design of twins reared together, it is assumed for the traits being analyzed that 1) genetic and environmental factors are not correlated, 2) there is no genotype x environment (G x E) interaction, and 3) mating is random (i.e., there is no tendency to marry "like"), which implies that the additive genetic covariance of DZ twins is half of the total additive genetic variation.


Figure 1
View larger version (9K):
[in this window]
[in a new window]

  		Fig. 1. Univariate path coefficient model of additive (A1 for twin 1, A2 for twin 2) and dominance genetic (D1 and D2) common (C1 and C2) and unique environmental (E1 and E2) effects for phenotypes of twins 1 and 2 (P1 and P2, respectively). For monozygotic (MZ) twins, {alpha} = 1 and beta = 1; for dizygotic (DZ) twins, {alpha} = 0.5 and beta = 0.25.

 
The associations between the independent variables, i.e., specific environmental and anthropometric factors, and the dependent HRV variables were analyzed separately by survey regression models for all individuals. Again, complex survey methods were used to take into account that the twin pairs were the primary sampling unit and that twins within pairs may not represent fully independent observations. The independent variables that were statistically significantly associated with dependent HRV variables were further examined by age-adjusted regression models. In those models, the explained portion (Rm2) includes also the part explained by age. Approximation for age-adjusted R2 is as follows: Radj2 = Rm2 – Rage2, where Rage2 is the portion explained by age in an unadjusted model.

Independent variables that were statistically significantly associated with the HRV outcome measure in age-adjusted univariate regression models were added to the multivariate twin model (Fig. 2). Only AE was assumed for HRV variables, because the best fit was achieved with the AE model for the most HRV variables. We evaluated the amount of variance of each HRV variable accounted for by each independent variable and then the amount of the remaining variance accounted for by age and unknown genetic and unknown unique environmental factors. The structural models were estimated using the Mx program (29).


Figure 2
View larger version (7K):
[in this window]
[in a new window]

  		Fig. 2. Multivariate twin model with regression of suspected determinant on heart rate variability (HRV) variable with age adjustment. Model is presented for 1 twin. Path coefficient t, effect of age on determinant; coefficient s, effect of age on HRV variable; coefficient beta, effect of determinant; a, c, and e, genetic, common environmental, and unique environmental path coefficients for determinant, respectively; as and es, specific genetic and unique environmental path coefficients of HRV variable (i.e., a portion of genes and environment from the variance of the dependent HRV variable, when age and another determinant are included in the model).

 
In the second phase, all analyses were repeated separately for healthy and diseased subjects to test the interaction between health status and genes and environmental and anthropometric factors and to test the effect of individual environmental and anthropometric factors on variance of the HRV variables in healthy and diseased subjects. The diseases are presented in Table 2. STATA survey mean and survey linear combination programs were used to test homogeneity of means and variances between health status and zygosity groups. Survey regression programs subpop (i.e., subpopulation) option was used to apply survey regressions to health status groups. Age-adjusted analysis was applied when the unadjusted regression was significant.

Path analysis with genotype x environment (G x E) interaction (28) was applied to study the interaction of health status with genetic effects and environmental and anthropometric factors. To test the assumption of no genotype-environment correlation, cross-twin cross-trait correlations between health status and HRV variables and between health status and environmental and anthropometric factors were estimated for MZ and DZ twins.

In univariate G x E interaction analysis, seven models were fitted to sample covariance matrices and means in each of six groups (healthy MZ, diseased MZ, discordant MZ, healthy DZ, diseased DZ, and discordant DZ) with use of maximum-likelihood estimation: the general G x E model, with varying A, D, and E effects for healthy and diseased groups and the additional genetic effect A' for diseased subjects to determine whether there is a different set of genes influencing the HRV of the diseased group (model I); the full common-effects G x E model, with varying A, D, and E effects for healthy and diseased groups (i.e., without the additional genetic factor, model II); the common-effects submodel 1, with varying A and E effects for healthy and diseased groups, but no D effects (model III); the common-effects submodel 2, with varying A effects for healthy and diseased groups, but equal E effects (model IV); the common-effects submodel 3, with varying E effects for healthy and diseased groups, but equal A effects (model V); the scalar G x E model, with A and E effects for the diseased group multiplied by the estimated scalar (i.e., a constant, model VI); and no G x E interaction (i.e., equal AE model for both groups, model VII).

Multivariate twin analyses with regression of suspected determinants on HRV variables with age correction and G x E interaction were applied to model the causal effects of determinants that were significant in group-specific (healthy or diseased) age-adjusted survey regression models. Models were estimated on the basis of matrices of Pearson, polychoric, and biserial correlations, with asymptotic covariance matrices in each of the six groups (MZ and DZ pairs by concordance and discordance for disease status) with use of asymptotic weighted least squares. Components of univariate model III (see above) were fitted to the HRV variable and determinant with varying regression coefficients due to health status. Model VII was chosen, because it provided the best fit for most HRV variables (see RESULTS).


   RESULTS
TOP
 ABSTRACT
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 GRANTS
 REFERENCES
 
Table 3 shows the HRV values in the MZ and DZ twin groups. Correlations in HRV variables between MZ twins are large and significant, but those between DZ twins are mostly nonsignificant (except in HP and LF/HF), which suggests a significant contribution of genetic effects for HRV variables.


View this table:
[in this window]
[in a new window]

  		Table 3. HRV and twin correlation coefficients with 95% CI

 
In univariate analyses, the model with additive and nonadditive genetic and unique environmental effects (ADE) provided the best fit to the data for RMSSD and HF, whereas the AE model was the best fit for HP, LF, and LF/HF. Accordingly, additive genetic or dominance genetic effects accounted for 31–57% of the total variance of HRV variables, and unique environmental effects (which also included measurement error) accounted for the remaining variance. LF/HF had the greatest unique environmental effect (69%) and HP the least (43%).

Univariate models with adjustment for age were used initially to identify significant determinants (independent variables) of HRV outcome variables (see above; data not shown). These values were then entered into path regression models with the following independent variables: age, BMI, percent body fat, chronic diseases, medication, smoking, and coffee consumption. We first describe results for the independent variables. As shown in Table 4, age had a significant effect on all HRV variables, except LF/HF, and all analyses were adjusted for age. The portion of variance explained by age for the HRV variables ranged from 1.6% to 9.4%. BMI, percent body fat, and smoking were significantly associated with HP, but the portions of variance were quite small: 1.2%, 0.8%, and 3.4%, respectively. Similarly, only BMI and medication were associated with RMSSD, again explaining 1.8% and 4.7% of the variance, respectively. Table 4 shows that the environmental and somatic determinants seemed to have the largest effects on the LF component. The portions of variance explained by smoking, chronic diseases, medication, and coffee consumption were 2.1%, 8.1%, 10.5%, and 2.3%, respectively. As for RMSSD, only BMI (2.2% of the variation) and medication (4.7% of the variation) were explanatory factors for HF, and only BMI was associated with LF/HF (2.2% of the variation). Occupational loading and any sport and leisure-time activities did not explain any HRV measure.


View this table:
[in this window]
[in a new window]

  		Table 4. Path models estimated using maximum-likelihood: goodness of fit statistics, portions explained by specific additive genetics, specific unique environment, age, determinant, and path coefficient

 
The portion of variance explained by genetic factors remained fairly stable when the different independent variables were entered into the model. For example, the heritability of HP was 50% in the age-adjusted model and 47% in a model with age and percent body fat (Table 4). However, for LF and HF (Table 4), inclusion of medication in the model decreased the estimate of heritability from 28% to 20% and from 37% to 27%, respectively, compared with the age-adjusted model.

In the second phase, we analyzed the data for the effect of health status, in addition to genetic effects, on the results. No significant interaction was found between health status and genetic effects on the group characteristics presented in Table 1. However, the healthy group was significantly younger (P < 0.001), and percent total body fat was lower in the healthy than in the diseased group (P < 0.001). In addition, the diseased group consisted of more long-term coffee drinkers (P = 0.001). A background factor may be the older age of the diseased group. RMSSD (P = 0.028), LF (P < 0.001), and HF (P = 0.018) were significantly higher in the healthy group. In univariate G x E interaction analysis, the no G x E interaction-equal AE model for both groups fit for all parameters except RMSSD and LF/HF (Table 5). For RMSSD, model V of varying E effects for healthy and diseased groups but equal A effects was the best fit; A effects accounted for 47% [95% confidence interval (CI) = 29–63] of the phenotypic variation of RMSSD for the healthy group and 39% (95% CI = 24–54) for the diseased group. However, for LF/HF, model IV of varying A effects for the healthy and diseased groups but equal E effects was the best fit. The heritability estimate (A) was 40% (95% CI = 22–55) for the healthy group and 21% (95% CI = 3–39) for the diseased group.


View this table:
[in this window]
[in a new window]

  		Table 5. Univariate G x E interaction analysis and its seven models

 
In the healthy group, the path analysis found that occupational loading explained 3.9% (1.0–8.7) of the variation in HP; BMI and coffee consumption (cups/day) explained 2.2% (0.4–9.2) and 1.9% (0.1–6.0), respectively, of the variation in HF, and BMI explained 2.2% (0.1–7.2) of the variation in LF/HF. In the diseased group, smoking explained HP [3.1% (0.5–7.8) of the variation], BMI and smoking explained RMSSD [3.7% (0.3–10.4) and 3.8% (0.8–9.1) of the variation], and BMI explained LF [1.1% (0.0–5.4, path coefficient = –1.361, 95% CI –2.354 to –0.379) of the variation] and HF [4.2% (0.5–11.3) of the variation]. This may indicate that, in the whole group and in the diseased group, medication and diseases mask the rather modest, but significant, effect of occupational loading on HP.


   DISCUSSION
TOP
 ABSTRACT
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
GRANTS
 REFERENCES
 
The results indicate, as hypothesized and previously reported (6, 7, 21, 46, 47, 49, 62), are major genetic influences on HRV. Furthermore, HRV is known to be influenced by numerous factors, but the portion of the remaining HRV variance accounted for by any single environmental factor seemed to be quite modest at best. Most of the factors we studied should decrease HRV. Only some medications, physical activity, and, perhaps, occupational (physical) loading could be considered a priori to have a positive effect on HRV. However, on the basis of the present study and some previous studies (58, 59), it is probably difficult to produce a sustained increase in HRV, but a decrease in HRV might be prevented by a healthy lifestyle. Our novel finding was that health status did not seem to affect the heritability estimates of HRV measures, with the exception of LF/HF. For LF/HF, the genetic variance was 19% greater in the healthy than in the diseased group, even though the diseased group was older on average. Thus, in the diseased group, mostly still unknown environmental factors and their interaction influenced autonomic balance more than genetic factors; adverse environmental and anthropometric factors may be the main reasons for changes in LF/HF in the diseased twins. On the other hand, for the RMSSD, heredity was a strong and stable factor in cardiac parasympathetic modulation; in the diseased group, however, unhealthy lifestyle and medications seemed to affect it negatively.

Most previous studies of genetic influences on HRV have used ambulatory HRV recordings or, at least, longer recording time with spontaneous breathing frequency. However, Sinnreich et al. (46, 47) used 6-min laboratory recordings with spontaneous and paced breathing frequency, and the findings in heritability of HRV were identical between these two breathing models. Heritability estimates for HRV measures were 39–45% in their study, which agrees well with our results (31–54%, 57% for HP).

The age variation was more heterogeneous in the study population of Sinnreich et al. (46, 47) than in the present study population. De Geus et al. (11) reported that increasing age is linked to higher heritability estimates of respiratory sinus arrhythmia. If this is the case, our values for heritability might have been higher than those of Sinnreich et al. because of the older age groups in our study. Kupper et al. (21) also found heritability values for RMSSD similar to our values (49%), but Singh et al. (43, 45) found that genes account for only 13–23% of the HRV variance. Singh et al. studied siblings and spouse pairs, and the statistical methods diverged from our methods. The difference in degree of heredity in different studies may be explained by the fact that twin studies tend to produce higher estimates of heritability than studies of nuclear families, partly because of much better control of age effects. Also, the basic assumption of identical environmental effects relevant to HRV in MZ and DZ twins may, to some degree, overestimate the effect of heredity in twin studies.

Age not only influences heredity estimates of HRV but, also, has a direct negative effect on HRV (1, 2, 12, 52, 57). Our results also showed an inverse association of age with HRV, except for LF/HF. In further analyses with adjustment for age, BMI seemed to be the most prominent nongenetic determinant of HRV and, especially, of parasympathetic indexes and LF/HF. In our study, the impact of BMI on HRV was nonetheless modest. In contrast, previous studies showed considerably lower LF in overweight than in normal-weight subjects (32, 36). Weight loss increased HF and LF (22, 34). The strongest physiological links between weight loss and changes in autonomic nervous function may be changes in insulin and glucose metabolism and leptin concentration, rather than changes in body fat per se (30, 38). Variation in autonomic function with BMI (60) might also be a physiological link between general health and mortality and BMI (10).

The third most significant independent determinant of HRV was medication (Table 2). beta-Blockers and ACE inhibitors were the most frequently used medications among the subjects, and beta-blockers enhance HRV (3, 4). In contrast, chronic diseases did not seem to have an independent effect on HRV, except for LF. The finding is supported by Gottsater et al. (14), who found that the LF component, in particular, was changed in diabetes-related atherosclerosis. Atherosclerosis is also related to cardiovascular diseases and hypertension, which were the most common diseases among the study subjects. LF variance has been thought to represent the effect of blood pressure regulation on HP and to arise mostly via sympathetic regulation (54), which changes first also in diabetes (61). This may explain why vagal indexes (RMSSD and HF) were not affected by chronic diseases. Also, the LF component was higher in asthmatic children treated with beta2-agonists than in their controls (16).

Contrary to expectation, we did not observe an effect of cumulative life-long leisure-time physical activity on HRV. This is consistent with a recent long-term intervention study in which habitual light-to-moderate regular physical activity did not increase HRV (59) in a normal population sample of Finnish men. A limitation of the present study is that we cannot be sure whether the exercise habits have significantly changed before actual measurements, and even detraining for some weeks may significantly decrease HRV (15). However, life-long occupational physical loading was found to be a determinant of HP in the healthy group. To our knowledge, there are no previous studies of the long-term effect of occupational physical loading on HR, but it has been reported that heavy physical work increases the risk for early retirement, especially because of musculoskeletal disorders (18). In addition, smoking and coffee consumption had a significant effect on LF and, in the subgroups, also on parasympathetic indexes. The results agree with previous findings that smoking and environmental tobacco smoke and coffee consumption decrease HRV and parasympathetic activity (35, 50) and increase sympathetic activity (13, 23) and, thus, may increase the risk for morbidity and mortality.

The present study is relevant, in that it estimates and confirms the remarkable role of genetic factors on HRV, a sign of and predictive marker for cardiac health, morbidity, and mortality (19, 20, 55). This genetic effect was found in healthy and diseased twins. Another more novel aspect of the present study was its consideration of the relative role of mostly modifiable environmental factors on HRV measures. We know of many diseases and metabolic states that affect HRV, such as insulin resistance and the metabolic syndrome (26), that also have significant genetic components (27). The next step would be to determine whether the genes underlying HRV and genes underlying, e.g., insulin resistance or vulnerability for coronary heart diseases are also related to each other. If such a relationship exists, HRV could also be an indirect marker for those traits. Despite a major genetic contribution, many environmental and somatic factors had a minor influence on HRV. However, a large portion of variance remains unaccounted for by genetic factors or known environmental factors. This provides additional challenges to discover these unknown and possibly modifiable factors influencing cardiac health.


   GRANTS
TOP
 ABSTRACT
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 GRANTS
REFERENCES
 
This research was supported by National Institute of Arthritis and Musculoskeletal and Skin Diseases Grant AR-40857; the Work Environment Fund, Finland; Academy of Finland Grants 38332 and 42044; the Alberta Heritage Foundation for Medical Research, Canada; and the European Union-funded EURODISC project (QLK6-CT-2002-02582). The Finnish Twin Cohort study is part of the Academy of Finland's Centre of Excellence for Complex Disease Genetics. The study was also supported by a grant from the Medical Fund of Kuopio University Hospital and by the Academy of Finland.


   ACKNOWLEDGMENTS
 
Present address of A. L. T. Uusitalo: Helsinki University Central Hospital, HUSLAB, Div. of Clinical Physiology and Nuclear Medicine, Meilahti Hospital, Helsinki, Finland.


   FOOTNOTES
 

Address for reprint requests and other correspondence: A. L. T. Uusitalo, Helsinki Univ. Central Hospital, Division of Clinical Physiology and Nuclear Medicine, Meilahti Hospital, PO Box 340, FIN-00029 HUS, Finland (e-mail: arja.uusitalo{at}hus.fi)

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
TOP
 ABSTRACT
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 GRANTS
 REFERENCES
 

  1. Agelink MW, Malessa R, Baumann B, Majewski T, Akila F, Zeit T, Ziegler D. Standardized tests of heart rate variability: normal ranges obtained from 309 healthy humans, and effects of age, gender, and heart rate. Clin Auton Res 11: 99–108, 2001.[CrossRef][Web of Science][Medline]
  2. Antelmi I, de Paula RS, Shinzato AR, Peres CA, Mansur AJ, Grupi CJ. Influence of age, gender, body mass index, and functional capacity on heart rate variability in a cohort of subjects without heart disease. Am J Cardiol 93: 381–385, 2004.[CrossRef][Web of Science][Medline]
  3. Aronson D, Burger AJ. Effect of beta-blockade on heart rate variability in decompensated heart failure. Int J Cardiol 79: 31–39, 2001.[CrossRef][Web of Science][Medline]
  4. Badilini F, Maison-Blanche P, Champomier P, Provost JC, Coumel P, Milon H. Frequency-domain heart rate variability in 24-hour Holter recordings: role of spectral method to assess circadian patterns and pharmacological autonomic modulation. J Electrocardiol 33: 147–157, 2000.[CrossRef][Web of Science][Medline]
  5. Battie MC, Videman T, Gibbons LE, Fisher LD, Manninen H, Gill K. 1995 Volvo Award in Clinical Sciences. Determinants of lumbar disc degeneration: a study relating lifetime exposures and magnetic resonance imaging findings in identical twins. Spine 20: 2601–2612, 1995.[Web of Science][Medline]
  6. Boomsma DI, van Baal GC, Orlebeke JF. Genetic influences on respiratory sinus arrhythmia across different task conditions. Acta Genet Med Gemellol (Roma) 39: 181–191, 1990.[Medline]
  7. Busjahn A, Voss A, Knoblauch H, Knoblauch M, Jeschke E, Wessel N, Bohlender J, McCarron J, Faulhaber HD, Schuster H, Dietz R, Luft FC. Angiotensin-converting enzyme and angiotensinogen gene polymorphisms and heart rate variability in twins. Am J Cardiol 81: 755–760, 1998.[CrossRef][Web of Science][Medline]
  8. Carney RM, Freedland KE, Veith RC. Depression, the autonomic nervous system, and coronary heart disease. Psychosom Med 67 Suppl 1: S29–S33, 2005.[Abstract/Free Full Text]
  9. Cohen H, Loewenthal U, Matar M, Kotler M. Association of autonomic dysfunction and clozapine. Heart rate variability and risk for sudden death in patients with schizophrenia on long-term psychotropic medication. Br J Psychiatry 179: 167–171, 2001.[Abstract/Free Full Text]
  10. Corrada MM, Kawas CH, Mozaffar F, Paganini-Hill A. Association of body mass index and weight change with all-cause mortality in the elderly. Am J Epidemiol 163: 938–949, 2006.[Abstract/Free Full Text]
  11. De Geus EJ, Boomsma DI, Snieder H. Genetic correlation of exercise with heart rate and respiratory sinus arrhythmia. Med Sci Sports Exerc 35: 1287–1295, 2003.
  12. Fagard RH, Pardaens K, Staessen JA. Influence of demographic, anthropometric and lifestyle characteristics on heart rate and its variability in the population. J Hypertens 17: 1589–1599, 1999.[Web of Science][Medline]
  13. Gerhardt U, Vorneweg P, Riedasch M, Hohage H. Acute and persistent effects of smoking on the baroreceptor function. J Auton Pharmacol 19: 105–108, 1999.[CrossRef][Web of Science][Medline]
  14. Gottsater A, Ahlgren AR, Taimour S, Sundkvist G. Decreased heart rate variability may predict the progression of carotid atherosclerosis in type 2 diabetes. Clin Auton Res 16: 228–234, 2006.[CrossRef][Web of Science][Medline]
  15. Gutin B, Barbeau P, Litaker MS, Ferguson M, Owens S. Heart rate variability in obese children: relations to total body and visceral adiposity, and changes with physical training and detraining. Obes Res 8: 12–19, 2000.[Web of Science][Medline]
  16. Jartti TT, Tahvanainen KU, Kaila TJ, Kuusela TA, Koivikko AS, Vanto TT, Valimaki IA. Cardiovascular autonomic regulation in asthmatic children evidenced by spectral analysis of heart rate and blood pressure variability. Scand J Clin Lab Invest 56: 545–554, 1996.[Web of Science][Medline]
  17. Kaprio J, Koskenvuo M. Genetic and environmental factors in complex diseases: the older Finnish Twin Cohort. Twin Res 5: 358–365, 2002.[CrossRef][Web of Science][Medline]
  18. Karpansalo M, Manninen P, Lakka TA, Kauhanen J, Rauramaa R, Salonen JT. Physical workload and risk of early retirement: prospective population-based study among middle-aged men. J Occup Environ Med 44: 930–939, 2002.[Web of Science][Medline]
  19. Kleiger RE, Miller JP, Bigger JT Jr, Moss AJ. Decreased heart rate variability and its association with increased mortality after acute myocardial infarction. Am J Cardiol 59: 256–262, 1987.[CrossRef][Web of Science][Medline]
  20. Kop WJ, Verdino RJ, Gottdiener JS, O'Leary ST, Bairey Merz CN, Krantz DS. Changes in heart rate and heart rate variability before ambulatory ischemic events. J Am Coll Cardiol 38: 742–749, 2001.[Abstract/Free Full Text]
  21. Kupper NH, Willemsen G, van den Berg M, de Boer D, Posthuma D, Boomsma DI, de Geus EJ. Heritability of ambulatory heart rate variability. Circulation 110: 2792–2796, 2004.[Abstract/Free Full Text]
  22. Laaksonen DE, Laitinen T, Schonberg J, Rissanen A, Niskanen LK. Weight loss and weight maintenance, ambulatory blood pressure and cardiac autonomic tone in obese persons with the metabolic syndrome. J Hypertens 21: 371–378, 2003.[CrossRef][Web of Science][Medline]
  23. Lane JD, Manus DC. Persistent cardiovascular effects with repeated caffeine administration. Psychosom Med 51: 373–380, 1989.[Abstract/Free Full Text]
  24. Lee YC, Wang HP, Lin LY, Lee BC, Chiu HM, Wu MS, Chen MF, Lin JT. Heart rate variability in patients with different manifestations of gastroesophageal reflux disease. Auton Neurosci 116: 39–45, 2004.[CrossRef][Web of Science][Medline]
  25. Levy WC, Cerqueira MD, Harp GD, Johannessen KA, Abrass IB, Schwartz RS, Stratton JR. Effect of endurance exercise training on heart rate variability at rest in healthy young and older men. Am J Cardiol 82: 1236–1241, 1998.[CrossRef][Web of Science][Medline]
  26. Liao D, Sloan RP, Cascio WE, Folsom AR, Liese AD, Evans GW, Cai J, Sharrett AR. Multiple metabolic syndrome is associated with lower heart rate variability. The Atherosclerosis Risk in Communities Study. Diabetes Care 21: 2116–2122, 1998.[Abstract]
  27. Loos RJ, Katzmarzyk PT, Rao DC, Rice T, Leon AS, Skinner JS, Wilmore JH, Rankinen T, Bouchard C. Genome-wide linkage scan for the metabolic syndrome in the HERITAGE Family Study. J Clin Endocrinol Metab 88: 5935–5943, 2003.[Abstract/Free Full Text]
  28. Mac Neale LR. Methodology of Genetic Studies of Twins and Families. Dordrecht: Kluwer Academic, 1992.
  29. Neale M. Mx: Statistical Modelling (Computer Program and Manual). Richmond, VA: Medical College of Virginia, Department of Psychiatry, 1999.
  30. Perciaccante A, Fiorentini A, Paris A, Serra P, Tubani L. Circadian rhythm of the autonomic nervous system in insulin resistant subjects with normoglycemia, impaired fasting glycemia, impaired glucose tolerance, type 2 diabetes mellitus. BMC Cardiovasc Disord 6: 19, 2006.[CrossRef][Medline]
  31. Perini R, Fisher N, Veicsteinas A, Pendergast DR. Aerobic training and cardiovascular responses at rest and during exercise in older men and women. Med Sci Sports Exerc 34: 700–708, 2002.
  32. Piccirillo G, Vetta F, Viola E, Santagada E, Ronzoni S, Cacciafesta M, Marigliano V. Heart rate and blood pressure variability in obese normotensive subjects. Int J Obes Relat Metab Disord 22: 741–750, 1998.[CrossRef][Web of Science][Medline]
  33. Pierratos A, McFarlane P, Chan CT, Kwok S, Nesrallah G. Daily hemodialysis 2006. State of the art. Minerva Urol Nefrol 58: 99–115, 2006.[Medline]
  34. Poirier P, Hernandez TL, Weil KM, Shepard TJ, Eckel RH. Impact of diet-induced weight loss on the cardiac autonomic nervous system in severe obesity. Obes Res 11: 1040–1047, 2003.[Web of Science][Medline]
  35. Pope CA 3rd, Eatough DJ, Gold DR, Pang Y, Nielsen KR, Nath P, Verrier RL, Kanner RE. Acute exposure to environmental tobacco smoke and heart rate variability. Environ Health Perspect 109: 711–716, 2001.[Web of Science][Medline]
  36. Quilliot D, Fluckiger L, Zannad F, Drouin P, Ziegler O. Impaired autonomic control of heart rate and blood pressure in obesity: role of age and of insulin-resistance. Clin Auton Res 11: 79–86, 2001.[CrossRef][Web of Science][Medline]
  37. Ropponen A, Levalahti E, Simonen R, Videman T, Battie MC. Repeatability of lifetime exercise reporting. Scand J Med Sci Sports 11: 185–192, 2001.[Web of Science][Medline]
  38. Rosenbaum M, Goldsmith R, Bloomfield D, Magnano A, Weimer L, Heymsfield S, Gallagher D, Mayer L, Murphy E, Leibel RL. Low-dose leptin reverses skeletal muscle, autonomic, and neuroendocrine adaptations to maintenance of reduced weight. J Clin Invest 115: 3579–3586, 2005.[CrossRef][Web of Science][Medline]
  39. Sarna S, Kaprio J, Sistonen P, Koskenvuo M. Diagnosis of twin zygosity by mailed questionnaire. Hum Hered 28: 241–254, 1978.[Web of Science][Medline]
  40. Schuit AJ, van Amelsvoort LG, Verheij TC, Rijneke RD, Maan AC, Swenne CA, Schouten EG. Exercise training and heart rate variability in older people. Med Sci Sports Exerc 31: 816–821, 1999.
  41. Shihara N, Yasuda K, Moritani T, Ue H, Adachi T, Tanaka H, Tsuda K, Seino Y. The association between Trp64Arg polymorphism of the beta3-adrenergic receptor and autonomic nervous system activity. J Clin Endocrinol Metab 84: 1623–1627, 1999.[Abstract/Free Full Text]
  42. Simonen R, Levalahti E, Kaprio J, Videman T, Battie MC. Multivariate genetic analysis of lifetime exercise and environmental factors. Med Sci Sports Exerc 36: 1559–1566, 2004.
  43. Singh JP, Larson MG, O'Donnell CJ, Levy D. Genetic factors contribute to the variance in frequency domain measures of heart rate variability. Auton Neurosci 90: 122–126, 2001.[CrossRef][Web of Science][Medline]
  44. Singh JP, Larson MG, O'Donnell CJ, Tsuji H, Corey D, Levy D. Genome scan linkage results for heart rate variability (the Framingham Heart Study). Am J Cardiol 90: 1290–1293, 2002.[CrossRef][Web of Science][Medline]
  45. Singh JP, Larson MG, O'Donnell CJ, Tsuji H, Evans JC, Levy D. Heritability of heart rate variability: the Framingham Heart Study. Circulation 99: 2251–2254, 1999.[Abstract/Free Full Text]
  46. Sinnreich R, Friedlander Y, Luria MH, Sapoznikov D, Kark JD. Inheritance of heart rate variability: the Kibbutzim Family Study. Hum Genet 105: 654–661, 1999.[CrossRef][Web of Science][Medline]
  47. Sinnreich R, Friedlander Y, Sapoznikov D, Kark JD. Familial aggregation of heart rate variability based on short recordings: the Kibbutzim Family Study. Hum Genet 103: 34–40, 1998.[CrossRef][Web of Science][Medline]
  48. Sivenius K, Niskanen L, Laakso M, Uusitupa M. A deletion in the {alpha}2B-adrenergic receptor gene and autonomic nervous function in central obesity. Obes Res 11: 962–970, 2003.[Web of Science][Medline]
  49. Snieder H, Boomsma DI, Van Doornen LJ, De Geus EJ. Heritability of respiratory sinus arrhythmia: dependency on task and respiration rate. Psychophysiology 34: 317–328, 1997.[Web of Science][Medline]
  50. Sondermeijer HP, van Marle AG, Kamen P, Krum H. Acute effects of caffeine on heart rate variability. Am J Cardiol 90: 906–907, 2002.[CrossRef][Web of Science][Medline]
  51. Stein PK, Ehsani AA, Domitrovich PP, Kleiger RE, Rottman JN. Effect of exercise training on heart rate variability in healthy older adults. Am Heart J 138: 567–576, 1999.[CrossRef][Web of Science][Medline]
  52. Stein PK, Kleiger RE, Rottman JN. Differing effects of age on heart rate variability in men and women. Am J Cardiol 80: 302–305, 1997.[CrossRef][Web of Science][Medline]
  53. Stolarz K, Staessen JA, Kawecka-Jaszcz K, Brand E, Bianchi G, Kuznetsova T, Tikhonoff V, Thijs L, Reineke T, Babeanu S, Casiglia E, Fagard R, Filipovsky J, Peleska J, Nikitin Y, Struijker-Boudier H, Grodzicki T. Genetic variation in CYP11B2 and AT1R influences heart rate variability conditional on sodium excretion. Hypertension 44: 156–162, 2004.[Abstract/Free Full Text]
  54. Task Force of the European Society of Cardiology, and the North American Society of Pacing and Electrophysiology. Heart rate variability. Standards of measurement, physiological interpretation, and clinical use. Eur Heart J 17: 354–381, 1996.[Free Full Text]
  55. Tsuji H, Venditti FJ Jr, Manders ES, Evans JC, Larson MG, Feldman CL, Levy D. Reduced heart rate variability and mortality risk in an elderly cohort. The Framingham Heart Study. Circulation 90: 878–883, 1994.[Abstract/Free Full Text]
  56. Tulppo MP, Hautala AJ, Makikallio TH, Laukkanen RT, Nissila S, Hughson RL, Huikuri HV. Effects of aerobic training on heart rate dynamics in sedentary subjects. J Appl Physiol 95: 364–372, 2003.[Abstract/Free Full Text]
  57. Umetani K, Singer DH, McCraty R, Atkinson M. Twenty-four-hour time domain heart rate variability and heart rate: relations to age and gender over nine decades. J Am Coll Cardiol 31: 593–601, 1998.[Abstract/Free Full Text]
  58. Uusitalo AL, Laitinen T, Vaisanen SB, Lansimies E, Rauramaa R. Effects of endurance training on heart rate and blood pressure variability. Clin Physiol Funct Imaging 22: 173–179, 2002.[CrossRef][Web of Science][Medline]
  59. Uusitalo AL, Laitinen T, Vaisanen SB, Lansimies E, Rauramaa R. Physical training and heart rate and blood pressure variability: a 5-yr randomized trial. Am J Physiol Heart Circ Physiol 286: H1821–H1826, 2004.[Abstract/Free Full Text]
  60. Vallejo M, Marquez MF, Borja-Aburto VH, Cardenas M, Hermosillo AG. Age, body mass index, and menstrual cycle influence young women's heart rate variability—a multivariable analysis. Clin Auton Res 15: 292–298, 2005.[CrossRef][Web of Science][Medline]
  61. Vinik AI, Maser RE, Mitchell BD, Freeman R. Diabetic autonomic neuropathy. Diabetes Care 26: 1553–1579, 2003.[Abstract/Free Full Text]
  62. Voss A, Busjahn A, Wessel N, Schurath R, Faulhaber HD, Luft FC, Dietz R. Familial and genetic influences on heart rate variability. J Electrocardiol 29 Suppl: 154–160, 1996.[CrossRef][Web of Science][Medline]




This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
293/2/H1013    most recent
00475.2006v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via Web of Science (3)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Uusitalo, A. L. T.
Right arrow Articles by Kaprio, J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Uusitalo, A. L. T.
Right arrow Articles by Kaprio, J.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Visit Other APS Journals Online
Copyright © 2007 by the American Physiological Society.