@article{oai:repo.qst.go.jp:00047316,
 author = {Minamimoto, Takafumi and Yamada, Hiroshi and Hori, Yukiko and Suhara, Tetsuya and 南本 敬史 and 堀 由紀子 and 須原 哲也},
 issue = {4},
 journal = {Experimental Brain Research},
 month = {May},
 note = {In the process of motivation to engage in a behavior, valuation of the expected outcome is comprised of not only external variables (i.e., incentives) but also
internal variables (i.e., drive). However, the exact neural mechanism that integrates these variables for the computation of motivational value remains unclear. Besides, the signal of physiological needs, which serves as the primary
internal variable for this computation, remains to be identified. Concerning fluid rewards, the osmolality level, one of the physiological indices for the level of thirst, may be an internal variable for valuation, since an increase in
the osmolality level induces drinking behavior. Here, to examine the relationship between osmolality and the motivational value of a water reward, we repeatedly
measured the blood osmolality level, while 2 monkeys continuously performed an instrumental task until they spontaneously stopped. We found that, as the total amount of water earned increased, the osmolality level progressively decreased (i.e., the hydration level increased) in an individual-dependent manner. There was a significant negative correlation between the error rate of the task (the
proportion of trials with low motivation) and the osmolality level. We also found that the increase in the error rate with reward accumulation can be well explained by a formula describing the changes in the osmolality level. These results provide a biologically supported computational formula for the motivational value of a water reward that depends on the hydration level, enabling us to identify the neural mechanism that integrates internal and external variables.},
 pages = {609--618},
 title = {Hydration level is an internal variable for computing motivation to obtain water rewards in monkeys},
 volume = {218},
 year = {2012}
}