Regular article Sex difference in the relationship between salivary testosterone and inter-temporal choice Hirokazu Doi, Shota Nishitani, Kazuyuki Shinohara ⁎ Graduate School of Biomedical Sciences, Nagasaki University, Nagasaki, Japan abstract article info Article history: Received 19 May 2014 Revised 11 December 2014 Accepted 12 December 2014 Available online 19 December 2014 Keywords: Inter-temporal choice Discounting Testosterone Sex difference Reward Humans often prefer a small immediate reward to large reward in the future. This myopic tendency in inter- temporal choice is termed delay discounting, and has been the focus of intensive research in the past decades. Recent studies indicate that the neural regions underlying delay discounting are influenced by the gonadal ste- roids. However, the specific relationship between the testosterone levels and delay discounting is unclear at this point, especially in females. The present study investigated the relationship between salivary testosterone concentrations and discounting rates in delay- and probability-discounting tasks with healthy males and females. The results revealed a positive correlation between testosterone concentrations and delay-discounting rates in females and a negative correla- tion in males. Testosterone concentrations were unrelated to probability-discounting rates. Although causal ef- fects of testosterone cannot be certain in this correlational study, if testosterone directly influenced this behavior, observed sex differences in delay discounting may be evidence of a curvilinear effect of testosterone. Alternatively, the findings may reflect inverse pattern of responsiveness to testosterone between male and fe- male neural systems, or basic sex-difference in the neural mechanism underlying delay-discounting independent of testosterone itself. © 2014 Elsevier Inc. All rights reserved. Introduction Humans share a universal tendency to value an immediate reward more highly than a larger reward in the future (Green et al., 2004; Jimura et al., 2009). That is, the subjective evaluation of a reward is discounted according to the delay in reward delivery; the longer the delay in reward delivery, the larger becomes the degree of discounting. This seemingly irrational tendency, often observed during inter- temporal choice, is termed “delay discounting” (DD). For example, peo- ple frequently succumb to strong urges to eat high-calorie, fat-saturated foods for immediate pleasure (immediate reward), ignoring the benefit of fitness and health in the future (delayed reward). A lack of premedi- tation regarding future consequences and weak self-control as observed in such everyday situations are considered to be fundamental elements of addiction (Marsch and Bickel, 2001). As such, the mechanisms under- lying DD in inter-temporal choice have been a topic of intensive research during the past few decades (Mischel, 1961; Kirby and Maraković, 1996; Shamosh et al., 2008). In a typical DD experiment, two alternatives, i.e. a smaller immediate reward and a larger delayed reward, are presented to participants. DD rate, the steepness of discounting of future reward, is estimated on the basis of participant's responses to different pairings of delay and reward. Importantly, previous studies have shown that individual differences in DD rate predict behaviors outside the labo- ratory. For example, a lower DD rate has been linked to a higher ac- ademic achievement, indicating that DD rate can predict levels of effort and perseverance devoted to attaining an improved future outcome (Kirby et al., 2002, 2005). Likewise, previous studies have revealed links between large DD rates and dysfunctional behaviors, like pathological gambling and substance abuse (Kirby and Petry, 2004; Green and Myerson, 2004). Previous studies have identified a number of factors, which poten- tially exert influences on DD rate, such as working memory capacity (Hinson et al., 2003), intelligence (de Wit et al., 2007), reward sensitiv- ity (Appelhans et al., 2011) and mood (Imhoff et al., 2014; Rounds et al., 2007; Worthy et al., 2014). In addition to these, many existing studies assumed, implicitly or explicitly, the link between DD rate and impulsiv- ity (Evenden and Ryan, 1996; Kirby et al., 1999), although there is some controversy over the validity of such assumption (Reynolds et al., 2006). “Impulsivity” is a multifaceted phenomenon that includes discrete impulsivity-related behaviors/personalities (Whiteside and Lynam, 2001; Evenden, 1999; Swann et al., 2002). Among the four components of impulsivity classified by Whiteside and Lynam (2001), the lack of premeditation is closely related, conceptually speaking, to DD. Likewise, Figner et al. (2010) have revealed that suppression of the dorsolateral prefrontal cortex leads to an increased DD rate, which indicates a rela- tionship between DD and low self-control (see also, Evenden and Hormones and Behavior 69 (2015) 50–58 ⁎ Corresponding author at: 1-12-4 Sakamoto-cho, Nagasaki City, Nagasaki 852-8523, Japan. Fax: +81 95 819 7036. E-mail address: kazuyuki@nagasaki-u.ac.jp (K. Shinohara). http://dx.doi.org/10.1016/j.yhbeh.2014.12.005 0018-506X/© 2014 Elsevier Inc. All rights reserved. Contents lists available at ScienceDirect Hormones and Behavior journal homepage: www.elsevier.com/locate/yhbeh