On the Addition of a Variable to a Previously Optimized System for the Measurement of the Nitrate-Nitrite Reductase Activity in Salivary Sediment

¹ Dental Section, Oklahoma Medical Research Institute, Oklahoma City, Oklahoma

In this study an attempt was made to add a fourth variable to a previously optimized system, dependent upon three independent variables. Two techniques, the method of steepest ascent and the one-factor-at-a-time procedure, were utilized. The one-factor-at-a-time method was used initially for an estimation of the optimal level of the fourth variable. It was used again when it became apparent that the temperature effect was overriding the effect of substrate concentration in the system. The method of steepest ascent was utilized for the interim investigations. The results of this study provided the most practical conditions of temperature (37° C.), pH (7.38), electron donor (1.024 mg. yeast extract/ml) and substrate concentration (18 µg. NO^-₃/ml) under which salivary nitrate-nitrite reductase activity can be measured in a clinical study.

In addition this investigation illustrates how a purely mechanical application of the method of steepest ascent can lead to the development of erroneous conclusions. A number of circumstances such as, when the system being optimized is very delicately balanced, when the optimum region is a broad plateau, when one independent variable causes a more striking effect than the others, when in the final analysis of the central composite design a non-negative coefficient for a quadratic term occurs in the fitted model, and inattention to an observation that is in conflict with the prediction equation, and/or the experimenter's expectations are some of the circumstances that can occasion (or signal) such (pending) difficulties.