Given these points, I used “Multivariable analysis”. Multivariable analysis is a statistical procedure to assess the obtained data correctly by summarizing the characteristics of the obtained data on the mutually-correlated multivariates (plural characteristic features) and collating them in accordance with the objectives. In such occasions, the systemic factorial data including the data on a trace amount of materials in blood was input to a computer because the forerunner researchers’ studies indicated association of the systemic factorial data with saliva secretion functions (Figure 9). As the soft-ware for analysis, we employed SAS (Statistical Analytical System). Then, these data were subjected to computer analysis to prepare the equation models (General linear multiple regression model) comprising the highly correlated factors to both secretion amount of saliva and protein concentrations in saliva. As the next step, these equation models were used to perform various investigations. For example, suppose that any teeth-related factors remain within these equation models, it is conceivable that teeth play an important role in affecting saliva secretion. If such an influence is so limited, they are eliminated during the course of computer analysis.Upon performing this analytical study, veteran internists of the Health Science Center of Kyushu University and specialists on statistics working for Department of mathematics, Faculty of Science of Kyushu University were kindly asked to cooperate to avoid detracting its medical and statistical scientific features. For calculation, super-computer installed in the Computer Center of Kyushu University was used; however, it took about 1 year only for conducting calculation. As the results, we have obtained scientifically extremely reliable equation models concerning salivary secretion amount (Figure 10).Although some steps concomitantly using the “Principal component analysis” were performed as the preliminary steps so as to decide directly the equation models about secretion amount of saliva and protein contents, these steps were omitted from the provision of this book because these steps are too difficult for the general publics to understand and furthermore, they are only the process to obtain the final equation models. However, if you are interested in reading these parts, please refer to my thesis with the title of “Influences of teeth on autonomic nervous system” (Reference 59) carried in “The NIPPON Dental Review” No. 584, a medical journal specified for dental specialists.Now looking at this equation model which includes teeth as the explanatory valuable, it is obvious that teeth play an important role in secretion of saliva. Incidentally, the selected teeth were maxillary first molar, second molar, first and second bicuspid and submaxillary second molar. Figure 11 shows the equation model indicating protein concentration in saliva. This equation demonstrates that as shown in this figure, multiple correlation coefficient, contribution ratio and p value show excellent levels, and this extremely reliable equation model like those for secretion of saliva is now available, indicating that teeth are included as an important factor. The selected teeth were maxillary canine, submaxillary first bicuspid, submaxillary second bicuspid and submaxillary first molar.According to these results, it is obvious that teeth are greatly contributory to secretion of saliva and protein concentration in saliva. For your easy confirmation of compatibility of the equation models to the actually observed values, Figure 12 is illustrated. Tendency of each value is beautifully consistent while it is anticipated that the equation models obtained by multivariable analysis remarkably reflect the actually observed values. Based on the above-stated facts, multivariable analysis based on background factors proved the correlation between teeth and salivary fluid. Existence of teeth influenced secretion of saliva directly. Qualitative and quantitative remarkable decreases in secretion of saliva which were noted during the course of the study on the vivid elderly persons were deeply associated with loss of teeth.
■Figure 9
Systemic parameters used for the Multivariable analysis
Systemic factors comprising 22 items were selected by cooperation of internal physicians.
1. Gender
2. Age
3. Body weight
4. Pulse rate
5. Blood pressure
6. RBC
7. Hb
8. Hematocrit (Blood bio-Chemical test data)
9. Adrenaline
10. Noradrenaline
11. Blood glucose level
12. Cholesterol
13. Triglyceride
14. Total protein
15. β-lipoprotein
16. HDL-cholesterol
17. Albumin
18. Albumin/globulin ratio
19. Ca
20. Na
21. Mg
22. K
■Figure 10
Equation model for salivary secretion volume
The equation models (General linear multiple regression model) showed that the “Contribution ratio” indicative of power of explanation and reliability was as high as 0.828. The multiple correlation coefficient indicating the compatibility between the actually determined values and the calculated values from the equation model was as high as 0.91; furthermore, the p value indicative of the statistical significance was as small as 0.0005. All of them suggest that the scientific credibility is extremely high. The fact that the dental factors are selected as the explanatory variables on the right-hand side means that teeth affect definitely the salivary secretion amount. Difference of signals such as (+) and (-) at the top of the factors according to teeth is to be noted. This indicates that kinds of teeth exert different influences.Salivary secretion volume=-7.62
+0.03 X Submaxillary second molar
+0.06 X Maxillary first bicuspid
-0.11 X Maxillary second bicuspid
+0.19 X Maxillary first molar
-0.15 X Maxillary second molar
+1.49 X Total protein amount
-2.59 X Albumin amount
+4.23 X Albumin/globulin ratio
+0.91 X Mg amount
+0.005 X Blood glucose level
-0.21 X K amount
+0.016 X Ht value
-0.043 X salivary protein concentration
Multiple correlation coefficient=0.91
Contribution ratio =0.828
P value<0.0005
■Figure 11
Equation model for salivary protein concentrations
In this equation model concerning the protein concentrations,the contribution ratio, the multiple correlation coefficient and p value were favorable in the similar manners as those of the salivary secretion amount, indicating higher scientific credibility. These findings suggest that teeth provide definitive influences on the salivary protein concentrations.Salivary protein concentrations=-9.57
+1.06 X Maxillary canine
+1.97 X Submaxillary first bicuspid
-1.48 X Submaxillary second bicuspid
-1.13 X Submaxillary first molar
+0.03 X Blood glucose level
+0.0059 X Adrenaline amount
+4.08 X Mg amount
-2.19 X salivary secretion amount
Multiple correlation coefficient=0.89
Contribution ratio =0.79
P value<0.0001
■Figure 12
The values calculated by the equation model were well coincident with the actually determined values.
In both the salivary secretion amount and the protein concentrations, the values calculated by the equation model were well coincident with the actually determined values. This proves that the credibility of the obtained equation model was scientifically high.
Subscribe to:
Post Comments (Atom)
No comments:
Post a Comment