Exploration Of Reference Ranges And Influencing Factors Of Cardiac Autonomic Function In Chinese Han Population

Objective:Clinical manifestation of cardiac autonomic neuropathy is latent, but it can increase cardiovascular events significantly, and cause myocardial infarction, sudden death and other serious consequences, so it’s very important and meaningful to evaluate the function of cardiac autonomic nervous system. Among some detection methods commonly used in researches, HRV and BRS was considered to be quantitative, objective, simple, feasible and safe. HRV is thought to be relatively more accurate sensitive index to assess cardiac autonomic nervous function, and has been widely used in patients with coronary heart disease, congestive heart failure, hypertension, diabetes and other diseases.BRS has also been reported to be significantly related to assessment of diagnosis、treatment and prognosis of cardiovascular disease, which is sensitive prediction index to malignant arrhythmia after acute myocardial infarction and sudden death. However, although HRV and BRS as the index of cardiac autonomic nervous system have been applied to a large number of researches, they have not been widely used in clinical practice, the reasons may be as follows:still lack of well-accepted reference value of HRV and BRS, many physiological and pathological influence factors, and not completely clear physiological significance of their index, hence cause the situation that it is difficult to judge normal or abnormal autonomic function.This study screened out healthy population in community at first. Then, on one hand, by short term power spectral detection and analysis of HRV and BRS, we explored the distribution characteristics and influencing factors of their indices, and listed the reference range of each main indice in healthy people by grouping according to main influencing factors, in order to provide preliminary short term power spectral HRV and BRS reference values in population with similar characteristics and provide a basis for further clinical application. On the other hand, by long term time domain detection and analysis of HRV lasting24hours, we studied the distribution characteristics of the indices in part of people of above study, and then studied the correlation of the indices of three kinds of detection methods forementioned, in order to help to further understand clinical significance of each indice and the mechanism of autonomic nervous function.Methods:1. The object of the research was healthy people screening out in three large communities of Baoshan District in ShangHai during May2011to May2012, after random sampling, collection of general information, physical examination and measurement of biochemical indices and short term spectral HRV and BRS. In total, there were403people for the study of short term spectral HRV and287people for short term spectral BRS.2. Part of above people were taken further24h ECG recording and detection, to obtain24h HRV time domain indice, totally including45patients.3. According to the corresponding research, detection and analysis methods of short term spectral HRV, short term spectral BRS or long term time domain HRV was used to evaluate the cardiovascular autonomic nerve function of subjects respecticely.4. Using Kolmogorov-Smirnov test to assess whether short term spectral HRV and BRS indices were normal distribution or not. If indices weren’t normally distributed, natural logarithm (ln) conversion was done to further compare indices in different groups.5. Differences of short term spectral HRV and BRS indices in groups of different ages or genders were compared using independent sample T test or variance analysis.6. Factors that might be related to indices of short term spectral HRV and BRS were preliminarily screened out by univariate linear regression analysis. Then those factors with P<0.2were included into multiple linear stepwise regression analysis to explore the independent influence factors.7. The reference range of each indice of short term spectral HRV and BRS, which were mostly non-normally distributed, was defined by the5th percentile and the95th percentile.8. The correlation of the indices of short term spectral HRV and BRS and long term time domain HRV was was analyzed using Spearman rank correlation analysis.9. All statistical analyses were performed using SPSS17.0. P<0.05was considered to be statistical significant. Results:1. Normal distribution test showed:most of indices of short term spectral HRV and BRS were not normally distributed except VLFrriNU (P<0.01), and were skewed to the left (skewness>0), but after natural logarithm (ln) conversion, they were nearly of normal distribution. To indices of long term time domain, most of them were also non-normally distributed and skewed to the left except SDNN and SDANN. In total, the mean of each indice was listed as follows:TPrri1072ms2, LFrri231ms2, LFrriNU21NU, HFrri234ms2, HFrriNU22NU, LF\HFrri1.55; Total.brs llms/mmHg, HF.brs11ms/mmHg, LF.brs15ms/mmHg; SDNN148.33ms, SDANN140.67ms2, SDNN index54.35ms, rMSSD30.53ms, pNN509.24ms, triangular index29.07ms, respectively.2. Comparation of indices of short term spectral HRV and BRS in groups of different ages or genders showed:among different age groups, there is a significant difference for lnTPrri, lnVLFrri, lnLFrri, lnLFrriNU, lnHFrri, lnLF/HFrri, lnTotal.brs and lnHF.brs, and the values of most of indices decreased with the increase of age. LnTP.rri, lnLFrriNU, lnHFrri, lnHFrriNU and LnLF/HFrri were significantly different between men and women (p<0.05), compared with male, lnTP.rri, lnHFrri, lnHFrriNU of female were higher while lnLFrriNU and LnLF/HFrri of female is obviously lower. In addition, in different age groups, respectively corresponding values of lnTPrri, lnLFrri and lnHFrri for male or female were still in declining trend with age increasing, but in the same age group, there weren’t significantly difference of forementioned indices between male and female except lnHFrri in age group of66-75.3. Multiple linear stepwise regression analysis for indices of short term spectral HRV and BRS showed that:heart rate was negatively correlated to most of indices of HRV and BRS except lnVLFrriNU and lnLFrriNU (totally9); age was negatively correlated to most of indices except lnVLFrriNU, lnHFrriNU and lnLFbrs(totally8); systolic blood pressure was negatively correlated to most of7dependent variables including lnVLFrriNU, lnHFrri, lnHFrriNU, lnLF/HFrri, lnTotalbrs, lnHF.brs and lnLF.brs. Gender and fasting plasma glucose were respectively correlated to3dependent variables including lnHFrri, lnHFrriNU and lnLF/HFrri. HDL was negatively correlated to three indexes of BRS. Postprandial plasma glucose was positively correlated to lnVLFrriNU. The more smoking, the higher lnLFrriNU is. Besides, the correlation between age, sex, fasting plasma glucose, HDL, smoking and indices of HRV and BRS were stronger (β value was larger), and the correlation between heart rate, systolic blood pressure and indices of HRV and BRS were relatively weaker (β value was smaller).4. Because age was a major independent factors associated with most indices of short term spectrual HRV and BRS; female had obviously higher parasympathetic activity than male (higher lnHFrri, lnHFrriNU, and lower lnLF/HFrri); difference of indices of short term spectrual BRS between gender was not statistically significant, the study listed the reference value for corresponding healthy people according to age groups and gender partly.5. The correlation between long term time domain and short term spectrual HRV indices showed:SDNN and SDNNindex were correlated to TPrri (r=0.34,0.37respectively), VLFrri (r=0.32,0.37respectively); rMSSD and pNN50were correlated to TPrri (r=0.54,0.50respectively), VLFrri (r=0.42,0.40respectively), LFrri (r=0.40,0.37respectively) and HFrri (r=0.55,0.54respectively), among which the correlation with HF.rri was more closely;SDANN and HRV triangle index were not significantly associated with any short term spectrual HRV index. The correlation between long term time domain HRV and short term spectrual BRS indeices showed:rMSSD and pNN50were correlated to HF.brs (r=0.38,0.39respectively), no significant correlation between other indexes was found. There were statistically significant correlations between short term spectrual HRV and BRS indices (p<0.05), but the correlation coefficients were relatively small. Mainly three indices of BRS were certainly correlated to TPrri and HFrri. Correlations among other indices are relatively weaker.Conclusions:1. In healthy population, most of the indices of short term spectrual HRV and BRS decreased with age increasing; and the gender difference is mainly that compared to male, female had relatively stronger activity of the vagus nerve.2. In healthy population, there were certain correlations between indices of short term spectrual HRV and BRS and metabolism related factors. In this study we showed some of the indices were independent correlated to fasting plasma glucose, HDL and systolic blood.3. This study proposed reference values of indices of short term spectrual HRV and BRS in healthy Chinese han population according to age and part sex groups. Then it provided primary reference for sample estimation of similar studies later, and also certain standards of judgment for its popularization in clinical.4. As to the correlations of three abovementioned methods of detection:there was certain correlation between long term time domain HRV and short term spectral HRV. But the correlations between short term spectral BRS and long term time domain HRV or short term spectral HRV were relatively weak. This suggested that different indicators might provide different information about cardiovascular autonomic function, and these three methods related to each other, but could not replace each other.