Abstract
The techniques of time series analysis are applied to data on the nitrogen output of adult rats with a view to understanding the underlying mechanism governing daily fluctuations in nitrogen output. The model was fitted to the nitrogen output of rats fed at different protein levels to observe any changes in the parameters. The analysis reveals that an autoregressive model of order one was an adequate fit for the nitrogen-output data when the rats were fed at protein requirement levels, and it undergoes a change at low protein levels. In particular, at the zero protein level, i.e. during the conditions of physiological stress, the nitrogen-output series were random for all the rats. Analysis thus shows that while the autoregressive model implies the existence of a regulatory mechanism underlying a given process, the random series implies conditions of physiological stress, i.e. the breakdown of a regulatory mechanism. This exercise highlights the importance of time series models in the field of physiology.