@ARTICLE\{IMM2007-04908,
    author       = "T. L. Nielsen and C. Hagen and K. Wraae and K. Brixen and P. H. Petersen and E. Haug and R. Larsen and M. Andersen",
    title        = "Visceral and Subcutaneous Adipose Tissue Assessed by Magnetic Resonance Imaging in Relation to Circulating Androgens, {SHBG,} and {LH} in Young Men",
    year         = "2007",
    month        = "apr",
    journal      = "Journal of Clinical Endocrinology and Metabolism",
    volume       = "",
    editor       = "",
    number       = "",
    publisher    = "",
    url          = "http://www2.compute.dtu.dk/pubdb/pubs/4908-full.html",
    abstract     = "Context: No large studies have examined the relation between circulating androgen levels and regional, abdominal adiposity in young men using magnetic resonance imaging (MRI).
Objective: To study the role of visceral and subcutaneous adipose tissue ({VAT} and {SAT}) on circulating androgens and to examine the impact of obesity on androgen reference-intervals. Design, Setting, and Participants: Population-based study of 783 Danish, 20-29 year-old men. Ninety-eight men were ruled out of the healthy reference-population by predefined criteria related with hypogonadism. Total, central, and lower extremity fat mass ({TFM,} {CFM,} and {LEFM}) were assessed in all men by {DXA} and {MRI} was performed in 406 men. Main Outcome Measures: Total, bioavailable, and free testosterone ({TT,} {BT,} and {FT}), androstenedione (\&\#8710;4AD), dihydrotestosterone (DHT), estradiol (E2),  2/TT-ratio, sex hormone-binding globulin (SHBG), and luteinizing hormone (LH). Results: Significant, inverse, linear relationships were observed between {TT,} {BT,} {FT,} {DHT,} {SHBG} and all {DXA} measures, whereas a positive correlation was found for the E2/TT-ratio. No associations were found for \&\#8710;4AD. Independent, inverse relations between {CFM} and {TT} (p<10-9), {BT} (p<0.005), {FT} (p<0.005), {DHT} (p<0.05), and {SHBG} (p<10-18) as well as a positive relation between {CFM} and E2/{TT} (p<0.001) were found in multiple regression analyses. {LEFM} was positively associated with {SHBG} (p<0.005). No other outcomes were independently related with {LEFM}. Multiple regression analyses of the {MRI} data showed that {VAT} was independently, inversely related with {BT} (p=0.002) and {FT} (p=0.003), whereas {SAT} was not. {SHBG} was independently, inversely related to {SAT} (p<10-6). Both {VAT} and {SAT} were inversely associated with {TT} (p=0.001 and p=0.005) and {DHT} (p=0.024 and p=0.018), but only {VAT} remained significantly related with {TT} (p=0.003) and {DHT} (p=0.049) when adjusted for {SHBG}. {LH} increased with increased {VAT} (p=0.004), but dropped suddenly in the most viscerally obese (p=0.006). The reference-intervals for {TT} (8.5-29.3 vs. 12.5-37.6 nmol/l), {BT} (6.1-16.9 vs. 7.6-20.7 nmol/l), and {DHT} (0.63-2.5 vs. 0.85-3.2 nmol/l) were significantly displaced in obese men compared to non-obese men. 22.9\% of the obese men and 15.3\% of the ruled out men had {TT} levels below 12.5 nmol/l.
Conclusion: {BT} and {FT} decline linearly with increasing {VAT} in young men. {TT} and {DHT} are inversely related with both {VAT} and {SAT,} but the relation with {SAT} may be secondary to decreased {SHBG} levels. {CFM} assessed by {DXA} is not a useful measure of visceral adipose tissue. {LH} data suggest that Leydig cell function is reduced in moderate visceral obesity, while excessive visceral obesity may resemble a state of hypogonadotropic hypogonadism. The reference-intervals for {TT,} {BT,} {FT,} and {DHT} were displaced in obese men. Referenceintervals for these androgens should be established in healthy, non-obese men.
Trial Registration: ClinicalTrials.gov: NCT00150163"
}