ANDROLOGY
ISSN: 2047-2919
ORIGINAL ARTICLE
Correspondence: Johan Svartberg, Section of Endocrinology, Division of internal Medicine, University Hospital of North Norway, 9038 Tromsø, Norway. E-mail: [email protected]
Keywords: cancer, cardiovascular disease, diabetes, epidemiology, mortality, SNP, testosterone Received: 10-Oct-2013 Revised: 7-Nov-2013 Accepted: 20-Nov-2013 doi: 10.1111/j.2047-2927.2013.00174.x
Single-nucleotide polymorphism, rs1799941 in the Sex HormoneBinding Globulin (SHBG) gene, related to both serum testosterone and SHBG levels and the risk of myocardial infarction, type 2 diabetes, cancer and mortality in men: the Tromsø Study 1,2 4
J. Svartberg, 3H. Schirmer, 4T. Wilsgaard, 5,6E. B. Mathiesen, 4I. Njølstad, M.-L. Løchen and 1,2R. Jorde
1 Tromsø Endocrine Research Group, Department of Clinical Medicine, UiT The Arctic University of Norway, 2Division of Internal Medicine, University Hospital of North Norway, 3Department of Clinical Medicine, 4Department of Community Medicine, 5Brain and Circulation Research Group, Department of Clinical Medicine, UiT The Arctic University of Norway, and 6Department of Neurology and Neurophysiology, University Hospital of North Norway, Tromsø, Norway
SUMMARY Low testosterone levels are associated with metabolic and cardiovascular disease risk factor, and have been shown to predict type 2 diabetes mellitus (T2DM), myocardial infarction (MI) and all-cause mortality. It is not known if these associations are causal or not. Recently, it has been shown that the serum testosterone levels are associated with single-nucleotide polymorphisms (SNPs), and we therefore studied the associations between one of these SNPs, rs1799941 on the Sex Hormone-Binding Globulin (SHBG) gene, and MI, T2DM, cancer and death. DNA was prepared from men who participated in the fourth survey of the Tromsø Study in 1994– 1995 and who were registered with the endpoints MI, T2DM, cancer or death and a randomly selected control group. For mortality, the observation time was set from 1994, and for the other endpoints from birth. The endpoint data were completed up to 2010–2013. Genetic analyses were successfully performed in 5309 men, of whom 1454 were registered with MI, 638 with T2DM, 1534 with cancer and in 2226 who had died. Men with the minor homozygote genotype had significantly higher levels of total testosterone (14.7%) and SHBG (24.7%) compared with men with the major homozygote genotype, whereas free testosterone levels did not differ significantly between the genotypes. The SNP rs1799941 was not significantly associated with MI, T2DM, cancer or mortality. Thus, our result does not support a causal relationship between total testosterone and SHBG and MI, T2DM, cancer or mortality, suggesting that low testosterone more likely is a marker of poor health.
INTRODUCTION The search for eternal youth has created a market for treatments that might affect the process of ageing, and testosterone is one of the hormones that have been in focus. It is well accepted that testosterone levels decline with increasing age, although the individual variation is large (Svartberg et al., 2003). Male hypogonadism is characterized by a low serum testosterone level in combination with a diversity of symptoms and signs such as reduced libido and vitality, decreased muscle mass, increased fat mass and depression. Similar symptoms in 212
Andrology, 2014, 2, 212–218
combination with subnormal testosterone levels are seen in some elderly men, and several attempts have been made to identify symptoms and corresponding testosterone levels that would define late-onset hypogonadism as a syndrome (Bhasin et al., 2010; Wu et al., 2010). However, symptoms of testosterone deficiency in elderly men are non-specific and are difficult to discriminate from symptoms of other conditions and diseases common in older men. Despite this uncertainty, the sale of testosterone has increased tremendously over the last few years (Handelsman, 2012). © 2013 American Society of Andrology and European Academy of Andrology
SNP RELATED TO TESTOSTERONE AND SHBG LEVELS
In observational studies, low testosterone levels tends to predict the development of metabolic syndrome and type 2 diabetes mellitus (T2DM) (Oh et al., 2002; Laaksonen et al., 2004; Vikan et al., 2010) and have been associated with both all-cause mortality and cardiovascular disease (CVD) mortality (Vikan et al., 2009a; Araujo et al., 2011). It is, however, not known whether these testosterone-disease associations are causal or not. It has recently been shown that the serum level of testosterone also is determined by genetic factors (Ahn et al., 2009; Nenonen et al., 2009; Skjaerpe et al., 2009; Ohlsson et al., 2011), and in the study by Ahn et al. (2009), men with the wild-type singlenucleotide polymorphism (SNP) rs1799941 in the Sex HormoneBinding Globulin (SHBG) gene, had significantly lower testosterone levels then men with the homozygous variant. One could therefore assume that men genetically predisposed for having lower testosterone levels, would be at risk for developing conditions and diseases associated with low serum testosterone levels. The Tromsø Study is a longitudinal epidemiological population-based health study. In the fourth survey in 1994–1995, blood samples for preparation of DNA were collected in approximately 27 000 subjects. The participants were followed up with registration of incident myocardial infarction (MI), T2DM, cancer and death and we therefore had the opportunity to test if a polymorphism affecting the testosterone level in men could be associated with these hard endpoints.
MATERIALS AND METHODS The Tromsø Study The Tromsø Study, conducted by UiT The Arctic University of Norway in cooperation with the National Health Screening Service, is a longitudinal population-based multipurpose study focusing on lifestyle-related diseases. The fourth survey was performed in 1994–1995, the fifth in 2001–2002 and the sixth in 2007–2008; 27 158, 8130 and 12 984 subjects attended respectively (Thelle et al., 1976; Jacobsen et al., 2012). Definition endpoints The definition of endpoints has been described thoroughly in a previous publication (Jorde et al., 2012). Shortly, possible cases of T2DM and MI were identified through self-reported diabetes and/or MI in questionnaires in the fourth, fifth and sixth surveys of the Tromsø Study and by linkage of the fourth survey participant list to the University Hospital of North Norway digital discharge diagnosis registry. The hospital medical record was then retrieved for case validation. In addition, a systematic manual and electronic search on all participants registered with cardiovascular diagnose codes were performed. The T2DM and MI endpoints were included till the end of 2011, but completely updated till the end of 2010. Information on cancer incidence and cancer location was retrieved from the Cancer Registry of Norway updated till the end of 2010. Information on death was obtained from the Causes of Death Registry, and information on moving out of the Tromsø area and emigration from Norway was obtained from the Norwegian Registry of Vital Statistics updated till 18 January 2013. Selection of study cohort The selection of the study cohort has been described previously (Jorde et al., 2012). Briefly, limited funding did not allow © 2013 American Society of Andrology and European Academy of Andrology
ANDROLOGY DNA preparation and genetic analyses of the entire Tromsø Study cohort and we therefore, in 2010, decided on a case-cohort design. With this approach, the same ‘control cohort’, randomly selected from the entire cohort who attended the fourth survey in 1994–1995, could be used as ‘control cohort’ for all the different endpoints (Kulathinal et al., 2007). A total of 9528 subjects were selected for participation in 2011. The final data set included 9471 subjects (4500 men) and among these 4175 subjects were in the control cohort. In 2013 the data set was updated and we also added SNP data on an additional 2509 subjects (869 men) from the Tromsø Study fourth survey cohort. With this extended data set, we now had SNP data including all subjects participating in the second phase of the fourth survey in 1994–1995 (Jacobsen et al., 2012). We did not have relevant information on the SNP rs1799941 to perform an accurate power calculation, and therefore set the base values at the conventionally used significance level of 5% at 80% power, with minor allele frequencies of 15% and 7% in cases and controls respectively. The sample size needed was then calculated to 239 cases and 239 controls. The inclusion of more than 5000 men in our study would clearly give more than sufficient power to detect differences in interest. Measurements At the survey in 1994–1995, the participants filled in questionnaires on medical history, and lifestyle factors. Blood pressure, height and weight, serum total cholesterol and triglycerides were analysed as previously described (Mathiesen et al., 2001). Analyses of testosterone and SHBG have been described previously (Svartberg et al., 2003), but shortly sera from the second visit in 1994–1995 were stored at 70 °C, and after a median storage time of 6.5 years, thawed in 2001 and analysed for sex hormones and SHBG using an automated clinical chemistry analyzer (Immulite 2000; Diagnostic Products, Los Angeles, CA, USA). Free testosterone values were calculated from total testosterone and SHBG using the Vermeulen formula (Vermeulen et al., 1999). Whole blood was collected for preparation of blood clots at the first visit (1994–1995) that were later stored at the HUNT/ NTNU Biobank in Levanger, Mid-Norway where the DNA was also prepared using a manual isolation method based on Clotted Lysate Preparation Protocol for 8 and 16 samples on the Autopure LS Instrument from Gentra (Gentra Systems Inc. MN, USA) using reagents from Qiagen (Qiagen NV, Venlo, The Netherlands). Based on the study by Ahn et al. (2009), where 874 SNPs in 37 candidate genes in the steroid hormone pathway were examined in relation to circulating levels of testosterone and SHBG in Caucasian men, one SNP, rs1799941, in the SHBG gene was reported to be most strongly associated with both testosterone and SHBG levels and was selected for analysis in this study. After we had performed our SNP analyse, information about other SNPs affecting the testosterone levels have been published (Ohlsson et al., 2011). The SNP rs1799941 was directly genotyped by KBioscience (http://www.kbioscience.co.uk) using KASP (KBioSience AlleleSpecific Polymorphism) SNP genotyping system, a competitive allele-specific polymerase chain reaction, which has previously been described in detail (Jorde et al., 2012). The call rate for the rs1799941 was 98.3%. Andrology, 2014, 2, 212–218
213
ANDROLOGY
J. Svartberg et al.
Statistics The relation between SNP genotypes and the endpoints MI, T2DM, cancer and mortality were evaluated in Cox regression analyses with age and body mass index (BMI) as covariates. For MI, the risk factors systolic blood pressure, serum cholesterol and smoking status were included. For mortality, the observation time was set from 1994, and for the other endpoints from birth. Cut-off for the observation period was by the end of 2010 for cancer, by the end of 2011 for MI and T2DM and 18 January 2013 for mortality. The pre-defined control cohort was used as controls for the men with a specific endpoint (cases). As this control cohort was randomly selected from the entire cohort, it also included a considerable number of men with one or more endpoints. When analysing a specific endpoint, men in the control cohort with that specific endpoint were moved to the case group (which only included men with that specific endpoint). Therefore, the size of the control cohort varied depending on endpoint in question. Distribution of the continuous variables total and free testosterone and SHBG, BMI, blood pressure and lipids and HbA1c were evaluated for skewness and kurtosis and by visual inspection of histograms and found normal except for triglycerides and HbA1c which were normalized by log transformation. Trends across the genotypes were evaluated with linear regression with age and BMI as covariates. The genotype frequency was examined for compliance with Hardy–Weinberg equilibrium using chi-squared analysis (Rodriguez et al., 2009). The data are shown as mean SD. All tests were performed two-sided, and a p-value
ISSN: 2047-2919
ORIGINAL ARTICLE
Correspondence: Johan Svartberg, Section of Endocrinology, Division of internal Medicine, University Hospital of North Norway, 9038 Tromsø, Norway. E-mail: [email protected]
Keywords: cancer, cardiovascular disease, diabetes, epidemiology, mortality, SNP, testosterone Received: 10-Oct-2013 Revised: 7-Nov-2013 Accepted: 20-Nov-2013 doi: 10.1111/j.2047-2927.2013.00174.x
Single-nucleotide polymorphism, rs1799941 in the Sex HormoneBinding Globulin (SHBG) gene, related to both serum testosterone and SHBG levels and the risk of myocardial infarction, type 2 diabetes, cancer and mortality in men: the Tromsø Study 1,2 4
J. Svartberg, 3H. Schirmer, 4T. Wilsgaard, 5,6E. B. Mathiesen, 4I. Njølstad, M.-L. Løchen and 1,2R. Jorde
1 Tromsø Endocrine Research Group, Department of Clinical Medicine, UiT The Arctic University of Norway, 2Division of Internal Medicine, University Hospital of North Norway, 3Department of Clinical Medicine, 4Department of Community Medicine, 5Brain and Circulation Research Group, Department of Clinical Medicine, UiT The Arctic University of Norway, and 6Department of Neurology and Neurophysiology, University Hospital of North Norway, Tromsø, Norway
SUMMARY Low testosterone levels are associated with metabolic and cardiovascular disease risk factor, and have been shown to predict type 2 diabetes mellitus (T2DM), myocardial infarction (MI) and all-cause mortality. It is not known if these associations are causal or not. Recently, it has been shown that the serum testosterone levels are associated with single-nucleotide polymorphisms (SNPs), and we therefore studied the associations between one of these SNPs, rs1799941 on the Sex Hormone-Binding Globulin (SHBG) gene, and MI, T2DM, cancer and death. DNA was prepared from men who participated in the fourth survey of the Tromsø Study in 1994– 1995 and who were registered with the endpoints MI, T2DM, cancer or death and a randomly selected control group. For mortality, the observation time was set from 1994, and for the other endpoints from birth. The endpoint data were completed up to 2010–2013. Genetic analyses were successfully performed in 5309 men, of whom 1454 were registered with MI, 638 with T2DM, 1534 with cancer and in 2226 who had died. Men with the minor homozygote genotype had significantly higher levels of total testosterone (14.7%) and SHBG (24.7%) compared with men with the major homozygote genotype, whereas free testosterone levels did not differ significantly between the genotypes. The SNP rs1799941 was not significantly associated with MI, T2DM, cancer or mortality. Thus, our result does not support a causal relationship between total testosterone and SHBG and MI, T2DM, cancer or mortality, suggesting that low testosterone more likely is a marker of poor health.
INTRODUCTION The search for eternal youth has created a market for treatments that might affect the process of ageing, and testosterone is one of the hormones that have been in focus. It is well accepted that testosterone levels decline with increasing age, although the individual variation is large (Svartberg et al., 2003). Male hypogonadism is characterized by a low serum testosterone level in combination with a diversity of symptoms and signs such as reduced libido and vitality, decreased muscle mass, increased fat mass and depression. Similar symptoms in 212
Andrology, 2014, 2, 212–218
combination with subnormal testosterone levels are seen in some elderly men, and several attempts have been made to identify symptoms and corresponding testosterone levels that would define late-onset hypogonadism as a syndrome (Bhasin et al., 2010; Wu et al., 2010). However, symptoms of testosterone deficiency in elderly men are non-specific and are difficult to discriminate from symptoms of other conditions and diseases common in older men. Despite this uncertainty, the sale of testosterone has increased tremendously over the last few years (Handelsman, 2012). © 2013 American Society of Andrology and European Academy of Andrology
SNP RELATED TO TESTOSTERONE AND SHBG LEVELS
In observational studies, low testosterone levels tends to predict the development of metabolic syndrome and type 2 diabetes mellitus (T2DM) (Oh et al., 2002; Laaksonen et al., 2004; Vikan et al., 2010) and have been associated with both all-cause mortality and cardiovascular disease (CVD) mortality (Vikan et al., 2009a; Araujo et al., 2011). It is, however, not known whether these testosterone-disease associations are causal or not. It has recently been shown that the serum level of testosterone also is determined by genetic factors (Ahn et al., 2009; Nenonen et al., 2009; Skjaerpe et al., 2009; Ohlsson et al., 2011), and in the study by Ahn et al. (2009), men with the wild-type singlenucleotide polymorphism (SNP) rs1799941 in the Sex HormoneBinding Globulin (SHBG) gene, had significantly lower testosterone levels then men with the homozygous variant. One could therefore assume that men genetically predisposed for having lower testosterone levels, would be at risk for developing conditions and diseases associated with low serum testosterone levels. The Tromsø Study is a longitudinal epidemiological population-based health study. In the fourth survey in 1994–1995, blood samples for preparation of DNA were collected in approximately 27 000 subjects. The participants were followed up with registration of incident myocardial infarction (MI), T2DM, cancer and death and we therefore had the opportunity to test if a polymorphism affecting the testosterone level in men could be associated with these hard endpoints.
MATERIALS AND METHODS The Tromsø Study The Tromsø Study, conducted by UiT The Arctic University of Norway in cooperation with the National Health Screening Service, is a longitudinal population-based multipurpose study focusing on lifestyle-related diseases. The fourth survey was performed in 1994–1995, the fifth in 2001–2002 and the sixth in 2007–2008; 27 158, 8130 and 12 984 subjects attended respectively (Thelle et al., 1976; Jacobsen et al., 2012). Definition endpoints The definition of endpoints has been described thoroughly in a previous publication (Jorde et al., 2012). Shortly, possible cases of T2DM and MI were identified through self-reported diabetes and/or MI in questionnaires in the fourth, fifth and sixth surveys of the Tromsø Study and by linkage of the fourth survey participant list to the University Hospital of North Norway digital discharge diagnosis registry. The hospital medical record was then retrieved for case validation. In addition, a systematic manual and electronic search on all participants registered with cardiovascular diagnose codes were performed. The T2DM and MI endpoints were included till the end of 2011, but completely updated till the end of 2010. Information on cancer incidence and cancer location was retrieved from the Cancer Registry of Norway updated till the end of 2010. Information on death was obtained from the Causes of Death Registry, and information on moving out of the Tromsø area and emigration from Norway was obtained from the Norwegian Registry of Vital Statistics updated till 18 January 2013. Selection of study cohort The selection of the study cohort has been described previously (Jorde et al., 2012). Briefly, limited funding did not allow © 2013 American Society of Andrology and European Academy of Andrology
ANDROLOGY DNA preparation and genetic analyses of the entire Tromsø Study cohort and we therefore, in 2010, decided on a case-cohort design. With this approach, the same ‘control cohort’, randomly selected from the entire cohort who attended the fourth survey in 1994–1995, could be used as ‘control cohort’ for all the different endpoints (Kulathinal et al., 2007). A total of 9528 subjects were selected for participation in 2011. The final data set included 9471 subjects (4500 men) and among these 4175 subjects were in the control cohort. In 2013 the data set was updated and we also added SNP data on an additional 2509 subjects (869 men) from the Tromsø Study fourth survey cohort. With this extended data set, we now had SNP data including all subjects participating in the second phase of the fourth survey in 1994–1995 (Jacobsen et al., 2012). We did not have relevant information on the SNP rs1799941 to perform an accurate power calculation, and therefore set the base values at the conventionally used significance level of 5% at 80% power, with minor allele frequencies of 15% and 7% in cases and controls respectively. The sample size needed was then calculated to 239 cases and 239 controls. The inclusion of more than 5000 men in our study would clearly give more than sufficient power to detect differences in interest. Measurements At the survey in 1994–1995, the participants filled in questionnaires on medical history, and lifestyle factors. Blood pressure, height and weight, serum total cholesterol and triglycerides were analysed as previously described (Mathiesen et al., 2001). Analyses of testosterone and SHBG have been described previously (Svartberg et al., 2003), but shortly sera from the second visit in 1994–1995 were stored at 70 °C, and after a median storage time of 6.5 years, thawed in 2001 and analysed for sex hormones and SHBG using an automated clinical chemistry analyzer (Immulite 2000; Diagnostic Products, Los Angeles, CA, USA). Free testosterone values were calculated from total testosterone and SHBG using the Vermeulen formula (Vermeulen et al., 1999). Whole blood was collected for preparation of blood clots at the first visit (1994–1995) that were later stored at the HUNT/ NTNU Biobank in Levanger, Mid-Norway where the DNA was also prepared using a manual isolation method based on Clotted Lysate Preparation Protocol for 8 and 16 samples on the Autopure LS Instrument from Gentra (Gentra Systems Inc. MN, USA) using reagents from Qiagen (Qiagen NV, Venlo, The Netherlands). Based on the study by Ahn et al. (2009), where 874 SNPs in 37 candidate genes in the steroid hormone pathway were examined in relation to circulating levels of testosterone and SHBG in Caucasian men, one SNP, rs1799941, in the SHBG gene was reported to be most strongly associated with both testosterone and SHBG levels and was selected for analysis in this study. After we had performed our SNP analyse, information about other SNPs affecting the testosterone levels have been published (Ohlsson et al., 2011). The SNP rs1799941 was directly genotyped by KBioscience (http://www.kbioscience.co.uk) using KASP (KBioSience AlleleSpecific Polymorphism) SNP genotyping system, a competitive allele-specific polymerase chain reaction, which has previously been described in detail (Jorde et al., 2012). The call rate for the rs1799941 was 98.3%. Andrology, 2014, 2, 212–218
213
ANDROLOGY
J. Svartberg et al.
Statistics The relation between SNP genotypes and the endpoints MI, T2DM, cancer and mortality were evaluated in Cox regression analyses with age and body mass index (BMI) as covariates. For MI, the risk factors systolic blood pressure, serum cholesterol and smoking status were included. For mortality, the observation time was set from 1994, and for the other endpoints from birth. Cut-off for the observation period was by the end of 2010 for cancer, by the end of 2011 for MI and T2DM and 18 January 2013 for mortality. The pre-defined control cohort was used as controls for the men with a specific endpoint (cases). As this control cohort was randomly selected from the entire cohort, it also included a considerable number of men with one or more endpoints. When analysing a specific endpoint, men in the control cohort with that specific endpoint were moved to the case group (which only included men with that specific endpoint). Therefore, the size of the control cohort varied depending on endpoint in question. Distribution of the continuous variables total and free testosterone and SHBG, BMI, blood pressure and lipids and HbA1c were evaluated for skewness and kurtosis and by visual inspection of histograms and found normal except for triglycerides and HbA1c which were normalized by log transformation. Trends across the genotypes were evaluated with linear regression with age and BMI as covariates. The genotype frequency was examined for compliance with Hardy–Weinberg equilibrium using chi-squared analysis (Rodriguez et al., 2009). The data are shown as mean SD. All tests were performed two-sided, and a p-value
