Skip Navigation

European Journal of Heart Failure 2005 7(6):958-965; doi:10.1016/j.ejheart.2004.06.009
This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Ärnlöv, J.
Right arrow Articles by Lithell, H.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Ärnlöv, J.
Right arrow Articles by Lithell, H.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

© 2005 European Society of Cardiology

hUNC-93B1, a novel gene mainly expressed in the heart, is related to left ventricular diastolic function, heart failure morbidity and mortality in elderly men

Johan Ärnlöva,*, Johan Sundströma, Lars Lindb, Bertil Andrénb, Maria Anderssona, Richard Renelanda,d, Lars Berglundc, Vladimir Kashubae, Alexei Protopopove, Eugene Zabarovskye and Hans Lithella

a Department of Public Health and Caring Sciences/Geriatrics PO Box 609, S-75125 Uppsala, Sweden
b Department of Medical Sciences, Uppsala University Sweden
c Uppsala Clinical Research Center, Uppsala University Sweden
d Sequenom Inc., San Diego USA
e Microbiology and Tumor Center and Center for Genomics and Bioinformatics, Karolinska Institute Sweden

* Corresponding author. Tel.: +46 18 6117972; fax: +46 18 6117976. E-mail address: Johan.Arnlov{at}pubcare.uu.se


   Abstract
Top
 Abstract
1. Introduction
 2. Subjects and methods
 3. Results
 4. Discussion
 5. Conclusions
 References
 
Aims: The hUNC-93B1 gene has the highest expression in the heart. We aimed to explore relationships between the hUNC-93B1 gene and cardiac function, morbidity and mortality in elderly men.

Methods and results: Two sub-samples of the population-based ULSAM-cohort (n=330, mean age 71 years and n=152, mean age 75 years, respectively) were used to explore and validate relationships between genotypes of the hUNC-93B1 gene and cardiac phenotypes (ejection fraction, E/A-ratio, left ventricular mass index and relative wall thickness). In the two samples, subjects homozygous for haplotype H3 had 34% and 35% higher level of E/A-ratio compared to non-carriers (p=0.0002 and 0.017, respectively) independent of cardiovascular disease and medication. Using national cause-of-death and hospital-discharge register data with 29 years of follow-up, no heart failure patients homozygous for haplotype H3 were hospitalised for heart failure before the age of 75 years, compared to 25% for heterozygous and 55% for non-carriers (p<0.03). No homozygous subjects died during follow-up while 17% of the heterozygous and 15% of the non-carriers died (p=0.01).

Conclusion: Haplotype H3 of the hUNC-93B1 gene seems related to E/A-ratio in elderly men. The relationship between the hUNC-93B1 gene and the age at onset of heart failure and mortality support a view of a clinically relevant impact of the gene.

Key Words: Heart failure • Echocardiography • Genes • Left ventricular diastolic function

Received April 6, 2004; Revised May 14, 2004; Accepted June 10, 2004


   1. Introduction
Top
 Abstract
 1. Introduction
2. Subjects and methods
 3. Results
 4. Discussion
 5. Conclusions
 References
 
The human member of the UNC (uncoordinated) gene family (hUNC-93B1) was recently identified and cloned [1]. hUNC-93B1 is a homologue to the Caenorhabditis elegans UNC-93 gene, which have been suggested to be involved in excitation–contraction coupling in muscle or in coordinating muscle contraction between muscle cells by affecting the functioning of gap junctions [2]. Altered-function alleles of UNC-93 in C. elegans result in sluggish movement and a characteristic "rubber band" uncoordinated phenotype.

The hUNC-93B1 gene has the highest level of expression in the heart but is expressed in all human tissues [1]. Due to the high gene expression in the heart, we hypothesized that the impairment of skeletal muscle function seen in C. elegans might also apply to human cardiac muscle and thus affect the function of the heart.

In the first phase of the study, the aim was to identify single nucleotide polymorphisms (SNPs) and haplotypes in the hUNC-93B1 gene. In the second phase, the aim was to investigate the relations between these genetic variations and five different indices of myocardial function and geometry in a sample of a population-based cohort of elderly men. In the third phase of the study, we wanted to validate the findings in a different, slightly older, sample of the same cohort. Finally, as a fourth phase we aimed at investigating the relation between the genotypes and heart failure morbidity and mortality using data from the national cause-of-death and hospital discharge registers.


   2. Subjects and methods
Top
 Abstract
 1. Introduction
 2. Subjects and methods
3. Results
 4. Discussion
 5. Conclusions
 References
 
2.1. Study population
A reinvestigation of the population-based ULSAM-cohort (The Uppsala Longitudinal Study of Adult Men) was performed between 1990 and 1994 when the subjects were approximately 70 years of age (n=1221, median age 71.5 years). A wide spectrum of metabolic, hemodynamic and anthropometric measurements was performed [3–6] (http://www.pubcare.uu.se/ulsam/). DNA was obtained in 944 of these subjects. All subjects gave written consent and the Ethics Committee of Uppsala University approved the study. The investigation conforms with the principles outlined in the Declaration of Helsinki.

2.1.1. Primary sample
The first 398 consecutive subjects of the cohort had an echocardiographic examination performed by an experienced physician. The technical quality was sufficient to assess all indicators of left ventricular function and geometry in 334 of the subjects. In order to treat the E/A-ratio as a continuous variable, subjects with a restrictive pattern (E/A-ratio>2.5) or with suspected ‘pseudonormalization— (LV systolic dysfunction with E/A-ratio>1.2) were excluded from the analysis (n=4) leaving 330 subjects who comprised the "primary sample" of this study. The primary sample was used in the haplotype detection and in the initial analyses of the relation between the genotypes and cardiac morphology and function.

2.1.2. Validation sample
Of the remaining subjects that were not examined with echocardiography at the investigation in 1990–1994, 426 subjects had an echocardiographic examination 3–8 years later (median age 75.1), performed by three different examiners. The technical quality of the echocardiographic examination was sufficient to assess all indicators of left ventricular function and geometry in 157 subjects. Due to ‘pseudonormalization— or a restrictive pattern 5 subjects were excluded from the analysis. The remaining 152 subjects comprised the population for the validation analyses and will hereafter be referred to as the validation sample.

2.2. Genetic analyses
2.2.1. SNP discovery
To identify polymorph positions in the hUNC-93B1 gene, 22 fragments covering selected regions of the gene were amplified using genomic DNA from 15–30 unrelated anonymous individuals. GenBank accession number AC004923 [GenBank] covers the genomic sequence of the hUNC-93B1 gene and was selected as master sequence. The 3{dagger}'-end of the gene was avoided, since that region is extremely homologous to other sequences in the human genome and there is risk of unspecific amplification. After amplification, detection of genetic variation in the hUNC-93B1 gene was performed using solid phase sequencing (AutoLoadTM Solid-Phase Sequencing kit, Amersham Pharmacia Biotech) and gel electrophoresis on ALFexpressTM sequencers (Amersham Pharmacia Biotech).

2.2.2. SNP typing
Fragments covering each of the selected SNPs were amplified from genomic DNA from the 330 individuals in the primary sample. Detection of SNPs using the PSQ platform (Pyrosequencing AB) was performed according to the manufacturers instructions. PSQ sample preparation VP93, VP99, VP101 and VP102: Capture at 60 °C for 30 min and annealing at 80 °C for 2 min. PSQ sample preparation VP94: Capture at 60 °C for 30 min and annealing at 95 °C for 2 min. Sequence of PCR and SNP typing primers are given in Appendix 1.

2.2.3. PCR components and conditions
Total reaction volume was 50 µl: GeneAmp®10X PCR-buffer II, 1.5 mM MgCl2 (Perkin Elmer), 0.125 mM dNTP (Ultrapure dNTP-set purchased from Amersham Pharmacia Biotech), 0.2 µM of each primer (Scandinavian Gene Synthesis), 0.65 U AmpliTaqGoldTM DNA polymerase (5U/µl]) (Perkin Elmer), and 0.2 ng/µl of DNA-sample. Amplification was performed using a GeneAmpTM PCR Systems 9700 from Perkin Elmer and the following conditions: 95 °C 10 min 45x(95 °C 30 s, Ta 60 °C 45 s, 72 °C 45 s) 72 °C 5 min, 22 °C.

2.2.4. Haplotype analysis
The analysis was performed in the primary sample using Haplotype Resolver, which is software based on the maximum likelihood methodology, and use of the EM algorithm [7].

In order to detect miss-genotyping, Hardy–Weinberg equilibrium, haplotype analysis and duplicate control were performed. No deviation in genotyping result could be seen in the 100 duplicates checked. The frequency of the three different genotypes for each SNP did not differ significantly from expected values in the Hardy–Weinberg calculations. In the haplotype analysis, two of the detected haplotypes appeared only once. As this was probably the result of an incorrect genotype calling, they were not considered true haplotypes. Haplotype analysis could be performed on a total of 310 individuals.

2.3. Echocardiography
A two-dimensional echocardiography was performed with a Hewlett-Packard Sonos 1500 cardiac ultrasound unit (Hewlett Packard Andover, Mass. USA). A 2.5 MHz transducer was used for the majority of 2-D, M-modes and Doppler examinations.

Using pulsed Doppler from the apical position, left ventricular inflow through the mitral valve was measured. The peak velocities of the early rapid filling (E-wave) and filling during atrial systole (A-wave) were recorded and from that the E/A-ratio was calculated.

Left ventricular ejection fraction, left ventricular mass index, relative wall thickness and left ventricular isovolumic relaxation time were measured by standardized criteria as previously described [3,4,6].

In a reproducibility study, 22 consecutive sampled subjects from the primary sample were reinvestigated approximately 1 month after the initial investigation. Intra-class-correlation coefficient for the different echocardiographic measurements: left ventricular mass index 0.65, relative wall thickness 0.72, ejection fraction 0.52, E/A-ratio 0.72 and isovolumic relaxation time 0.84.

2.4. Other measurements
A 12-lead electrocardiogram (ECG) with determination of Cornell product, QT-max and QT-dispersion as well as a 24-h ambulatory blood pressure measurement were performed as previously described [4,5].

2.5. Mortality and heart failure morbidity
The analysis of the relation between the genotypes and mortality and heart failure morbidity was performed in all subjects of the ULSAM-cohort where haplotype H3 and the SNP vp94 were obtainable (n=836 and 913, respectively). The mean age at base-line of the follow-up was 49.5 years (range 48.6 to 51.1 years) and the subjects had a mean follow-up time of 29.5 years (range 23.7 to 31.7 years), contributing to 26,923 person-years. Mortality was defined using the Swedish national cause-of-death register and heart failure was defined as first hospitalisation or death from heart failure (International Classification of Disease (ICD) 8 code 427 and 428, ICD 9 code 428 or ICD 10 code I50) using the Swedish national cause-of-death and hospital discharge registers. During follow-up 144 of 913 subjects died (rate 5/1000 PYAR), and 96 of 913 (rate 4/1000 PYAR) had a heart failure event.

2.6. Statistical methods
2.6.1. Primary sample
We used five primary phenotype end-point variables. For each combination of phenotype end-point and the nine genotypes, an analysis of variance model was estimated where the genotype was the factor. Logarithmic transformation was performed to achieve normal distribution if necessary. In order to control the probability for type 1 error in the primary endpoint analyses, the statistical significance of the results was ascertained with a permutation test [8]. This procedure holds the probability for type 1 error fixed at 5% over all tests considered. A p-value from the non-permuted data (p<0.000715), which was lower than the 5th percentile in the permuted distribution, was considered statistically significant.

The other end-point variables were considered exploratory endpoints and as no account were taken for multiple testing a p-value<0.05 was considered statistically significant.

In a second round of analyses, the E/A-ratio was tested in an analysis of co-variance using previous myocardial infarction (hospital discharge record or Q-wave at ECG), 24-h diastolic blood pressure, body mass index (BMI), diabetes, heart rate at the time of the echocardiography examination, and the use of ACE-inhibitors, alpha-blockers, beta-blockers, calcium antagonists, diuretics and digitalis as covariates (seeAppendix Table 2 for distribution of covariates in the cohort).

2.6.2. Validation sample
In the validation sample our primary hypothesis was to validate the relation between the E/A-ratio and the five SNPs and haplotype H3 found in the primary sample. For each combination of the E/A-ratio and the six genotypes, an analysis of variance model was estimated where the genotype was the factor. Using the permutation test, a non-permuted p<0.019 was considered statistically significant.

2.6.3. Mortality and heart failure morbidity
Chi2-test was performed to assess differences in mortality and heart failure prevalence between genotypes. ANOVA was used to assess differences in age at first hospitalisation for heart failure.


   3. Results
Top
 Abstract
 1. Introduction
 2. Subjects and methods
 3. Results
4. Discussion
 5. Conclusions
 References
 
3.1. Genotyping
3.1.1. SNP discovery
Six SNPs were identified with a minor allele frequency of ≥10% in the investigated individuals. Due to technological problems with the assay design five of these were included in the association study (Table 1).


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

  		Table 1 SNPs included in the hUNC-93B1 gene association study

 
3.1.2. Haplotype typing
Six haplotypes were identified. The four most common haplotypes were included in the association study (Table 2).


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

  		Table 2 Detected haplotypes of the hUNC-93B1 gene

 
3.2. Phenotype association
3.2.1. Primary sample
The relationships between the E/A-ratio and the five SNPs and haplotype H3 are shown in Table 3. The genotype homozygous for H3 showed higher level of E/A-ratio compared to non-carriers after permutation test and adjustment for myocardial infarction, blood pressure, heart rate, BMI, diabetes and the use of cardiovascular medication (p=0.0002). The relations between the E/A-ratio and the genotypes vp93, vp94, vp101 and vp102 seen in Table 3 did not reach statistical significance after accounting for the multiple testing using the permutation test and after adjustments of the above confounders. No relations were seen between the genotypes H1, H2, or H4 and the E/A-ratio (data not shown).


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

  		Table 3 Associations of genotypes of the hUNC-93B1 gene and left ventricular diastolic function in the primary sample and the validation sample of the ULSAM-cohort

 
No associations were seen between any of the 5 genotypes and 4 haplotypes and the other primary phenotype variables (left ventricular ejection fraction, relative wall thickness, left ventricular mass index or QT-max) or the exploratory variables (isovolumic relaxation time, Cornell product or the QT-dispersion) (data shown in Appendix Table 3).

3.2.2. Validation sample
The relationships between the E/A-ratio and the different genotypes are shown in Table 3. The genotype homozygous for H3 had higher level of E/A-ratio compared to non-carriers (p<0.017) and the C/C allele of vp94 had higher level of the E/A-ratio compared to the T/T allele (p<0.015) after permutation test and adjustment for myocardial infarction, blood pressure, heart rate, BMI, diabetes and the use of cardiovascular medication.

3.2.3. Combined population
When combining the primary and the validation sample, the genotype homozygous for H3 had 41% higher level of E/A-ratio compared to non-carriers and the C/C allele of vp94 had 33% higher level of the E/A-ratio compared to the T/T allele after adjustment for myocardial infarction, blood pressure, heart rate, the use of cardiovascular medication, BMI, diabetes and age (ANOVA p=0. 0001 and p<0.0001 respectively, Fig. 1).


Figure 1
View larger version (7K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 1 E/A-ratio in subjects who are non-carriers, heterozygous or homozygous of haplotype 3 of the hUNC-93B1 gene. The lowest, second lowest, middle, second highest and highest box points represent the minimum, 25th percentile, median, 75th percentile and maximum respectively. Means indicated by {blacksquare}. Data are adjusted for age, myocardial infarction, heart rate, blood pressure, body mass index, diabetes and the use of cardiovascular medication.

 
3.3. Heart failure morbidity and mortality
There was no difference in heart failure prevalence between the genotypes of haplotype H3 and vp94 at the end of the follow-up (Table 4). However, there was a significant difference between the genotypes regarding the age at the first hospitalisation for heart failure (Fig. 2). In fact, none of the heart failure patients homozygous for haplotype H3 had their first hospitalisation for heart failure before the age of 75 years, compared to 25% for the heterozygous heart failure patients and 55% for the non-carriers (p=0.01) and none of the heart failure patients with C/C-allele in vp94 had their first hospitalisation for heart failure before the age of 75 years, compared to 26% for the C/T-allele heart failure patients and 55% for the T/T-allele (p=0.01). If we exclude the heart failure patients for homozygous haplotype H3 or with C/C-allele in vp94 (n=2), the difference between the age at onset of heart failure remains significant between the non-carriers and the heterozygous and between the T/T and C/T allele (p<0.01).


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

  		Table 4 The relation between genotypes of the hUNC-93B1 gene and heart failure prevalence

 


Figure 2
View larger version (9K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 2 Age at first hospitalisation for heart failure in subjects who are non-carriers, heterozygous or homozygous of haplotype 3 of the hUNC-93B1 gene. The lowest, second lowest, middle, second highest and highest box points represent the minimum, 25th percentile, median, 75th percentile and maximum respectively. Means indicated by {blacksquare}.

 
None of the subjects homozygous for haplotype H3 or with C/C-allele in vp94 died during follow-up compared to 15.5–17.7% mortality in the other genotype groups (Table 4).


   4. Discussion
Top
 Abstract
 1. Introduction
 2. Subjects and methods
 3. Results
 4. Discussion
5. Conclusions
 References
 
As the novel hUNC-93B1 gene has the highest expression in the heart we wanted to investigate the relation between different hUNC-93B1 genotypes and cardiac phenotypes. In the first phase of the study we identified the most common SNPs and haplotypes. In the second phase, five primary end-point variables were chosen to reflect different aspects of cardiac function. A relation between haplotype H3 of the hUNC-93B1 gene and the E/A-ratio was found. As a third phase, in order to further strengthen the findings, we validated the results in another, slightly older, sample of the same cohort. And as a final phase, we found a relation between these genotypes and both the age at first hospitalisation for heart failure as well as total mortality, using data from the national cause-of-death and hospital discharge registers.

The E/A-ratio, one of the most widely used measurements of diastolic function, is affected by several cardiovascular risk factors like age, previous myocardial infarction, diabetes, hypertension, heart rate, obesity and the use of cardiovascular medication [9,11]. That the present associations between genotypes of hUNC-93B1 and the E/A-ratio were independent of these factors suggest that there may be alternative mechanisms that mediate the effects of the gene.

The physiology of diastolic function is characterized by complex interactions between left atrial and LV pressures, LV cellular derangements, and myocardial relaxation and compliance. From the echocardiographic view, LV diastolic function could be divided into three distinct parts. The isovolumic relaxation time is the earliest part of diastole and related to removal of Ca2+ from the cytoplasm by Ca2+-ATPases. Thereafter the transmitral filling of the left ventricle begins giving rise to the E-wave measured by Doppler. After this early filling period of the left ventricle, the flow into the left ventricle is further enhanced by the atrial contraction, giving rise to the A-wave. The relationship between the early and atrial filling periods is quantified by the E/A-ratio. Several factors may influence this ratio, but a reduced compliance of the left ventricle is suggested to be one of the major factors determining a reduced E/A-ratio. Such a reduction in LV compliance could either be due to structural changes, as typically seen in LV hypertrophy or functional, as seen in the ischaemic myocardium or during stunning. With the limited knowledge of the action of the hUNC-93B1 gene it seems most likely that the gene would influence the functional part of LV compliance.

The fact that the findings were independent of myocardial infarction supports a hypothesis of a direct myocardial effect of the gene, which is not mediated by coronary atherosclerosis. It should be pointed out that this study does not tell anything about the role or function of the gene, which will be a task for further molecular genetic studies, as well as for further studies of phenotype characteristics. As the gene is expressed in all human tissue [1] it may posses other functions not connected with muscle contraction. The function of hUNC93-B1 gene in the heart and the impact of its variants on function of the protein are still largely unknown.

The fact that we did find an association to both the age at first hospitalisation for heart failure and mortality in spite of the low number of cases indicates that the hUNC-93B1 gene has a clinically relevant impact at the population level. However, a validation in a population with larger number of heart failure cases is warranted.

As haplotypes are a combination of different SNPs, they are suitable in association studies. A haplotype may be able to reveal more information about the genotype/phenotype association than a single SNP. In this study, the subjects with a C/C-allele of VP94 were more or less the same as the subjects homozygous for haplotype H3, thus the haplotype did not portray more information than the single SNP.

The cDNA of hUNC-93B1 comprises 2282 base pairs, corresponding to 597 amino acids in the hUNC-93B1 protein. Structure prediction analyses of the protein sequence suggest hUNC-93B1 to code for a transmembrane protein with potential transporter activity [1]. If the hUNC-93B1 protein is located in the cell membrane it might be a possible target site for future pharmacological intervention. Currently, there is a need for drugs that improve left ventricular diastolic function. Up to half of the patients diagnosed with heart failure have been suggested to have a primary left ventricular diastolic dysfunction [10] and at present, there are only a few randomised clinical trials on treatment of primary diastolic heart failure.

4.1. Strengths and weaknesses
Some of the previous genotype association studies have been performed in small populations or in various subgroups, which may have resulted in statistical artefacts and bias, leading to false positive or false negative conclusions. In this study, we tried to avoid these pitfalls by control for multiple testing, adjustment for possible confounders and validation of the findings in a separate group. In our view, these procedures strengthen the validity of our findings considerably. Our sample size limited the number of genotypes and phenotypes explored. With 5 genotypes and four haplotypes, there was enough statistical power to use five primary end-point variables and still account for the multiple testing. As the frequencies of the rare homozygote were low there may be an increased risk of type 1 error even though the multiple testing was taken into account. However, this risk is attenuated by the fact that the E/A-ratio level in the heterozygote group lies in between the non-carrier and the homozygous.

In an association study like this we cannot distinguish between whether the link between hUNC-93B1 and E/A-ratio and heart failure is causal, i.e. whether the gene is involved in the pathophysiology of LV diastolic function, or if it is due to linkage disequilibrium with another functional gene. However, the location of VP94 in the putative promoter region implies that this SNP may be located in a functional segment of DNA. Further studies of the influence of the genotypes on protein expression and function are needed in order to establish causality.

A homogenous population like the ULSAM-cohort is advantageous when trying to discover associations between genotype and phenotype, but as we only examined men of similar age with the same ethnic background, the results may have limited generalizability to women and other age- and ethnic groups. There is always a risk of selection bias when using subgroups of a cohort, but the two samples in this study did not differ in metabolic, anthropometric or hemodynamic variables compared to each other or compared to the rest of the cohort at the age of 70 (data not shown).


   5. Conclusions
Top
 Abstract
 1. Introduction
 2. Subjects and methods
 3. Results
 4. Discussion
 5. Conclusions
References
 
Relations between the SNP vp94 and haplotype H3 of the hUNC-93B1 gene and the E/A-ratio were found and validated in a population-based cohort of elderly men. Furthermore, these genotypes predicted mortality as well as the age at onset of heart failure, which implies a clinically significant impact of the gene at the population level, but further functional studies are needed in order to establish causality.


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

  		Appendix Table 1 Sequence of PCR and SNP detection primers

 


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

  		Appendix Table 2 Distribution of covariates

 


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

  		Appendix Table 3 Primary and secondary cardiac phenotype outcome variables in the primary sample

 

   Acknowledgements
 
Axel and Margaret Ax:son Johnson Foundation, Ernfors Foundation, Glaxo Smith Kline, King Gustaf V and Queen Victoria Foundation, Pharmacia, Sequenom Inc, the Swedish Cancer Society, Swedish Medical Research Council No. 5446, STINT, the Swedish National Association against Heart and Lung Disease (Hjärt-Lungfonden), Trygg Hansa, Thuréus Foundation, Uppsala Geriatric Research Foundation, and Uppsala University.


   References
Top
 Abstract
 1. Introduction
 2. Subjects and methods
 3. Results
 4. Discussion
 5. Conclusions
 References
 

  1. Kashuba V.I., Protopopov A.I., Kvasha S.M., et al. hUNC93B1: a novel human gene representing a new gene family and encoding an UNC-93-like protein. Gene (2002) 283:209–217.[CrossRef][Web of Science][Medline]
  2. Levin J.Z., Horvitz H.R. The Caenorhabditis elegans UNC-93 gene encodes a putative transmembrane protein that regulates muscle contraction. J Cell Biol (1992) 117:143–155.[Abstract/Free Full Text]
  3. Andren B., Lind L., Hedenstierna G., Lithell H. Left ventricular systolic function in a population sample of elderly men. Echocardiography (1998) 15:315–324.[CrossRef][Web of Science][Medline]
  4. Sundstrom J., Lind L., Arnlov J., Zethelius B., Andren B., Lithell H.O. Echocardiographic and electrocardiographic diagnoses of left ventricular hypertrophy predict mortality independently of each other in a population of elderly men. Circulation (2001) 103:2346–2351.[Abstract/Free Full Text]
  5. Bjorklund K., Lind L., Lithell H. Twenty-four hour ambulatory blood pressure in a population of elderly men. J Intern Med (2000) 248:501–510.[CrossRef][Web of Science][Medline]
  6. Andren B., Lind L., Hedenstierna G., Lithell H. Left ventricular diastolic function in a population sample of elderly males. Echocardiography (1998) 15:443–450.[CrossRef][Web of Science]
  7. Excoffier L., Slatkin M. Maximum-likelihood estimation of molecular haplotype frequencies in a diploid population. Mol Biol Evol (1995) 12:921–927.[Abstract]
  8. Churchill G.A., Doerge R.W. Empirical threshold values for quantitative trait mapping. Genetics (1994) 138:963–971.[Abstract]
  9. Lee M., Gardin J.M., Lynch J.C., et al. Diabetes mellitus and echocardiographic left ventricular function in free-living elderly men and women: The Cardiovascular Health Study. Am Heart J (1997) 133:36–43.[CrossRef][Web of Science][Medline]
  10. Vasan R.S., Larson M.G., Benjamin E.J., Evans J.C., Reiss C.K., Levy D. Congestive heart failure in subjects with normal versus reduced left ventricular ejection fraction: prevalence and mortality in a population-based cohort. J Am Coll Cardiol (1999) 33:1948–1955.[Abstract/Free Full Text]
  11. Ärnlöv J., Lind L., Sundström J., Andrén B., Vessby B., Lithell H. Insulin resistance, dietary fat intake and blood pressure predict left ventricular diastolic function 20 years later. Nutr Metab Cardiovasc Dis (2005) 15:242–249. [Published online ahead of print, June 6, 2005].[CrossRef][Web of Science][Medline]

Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us    What's this?



This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Ärnlöv, J.
Right arrow Articles by Lithell, H.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Ärnlöv, J.
Right arrow Articles by Lithell, H.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?