Abstract
Objectives
Malnutrition challenges human health worldwide. There is no measurement for the cause-effect relation between 'nutrient intake' and 'sign, symptoms' of malnutrition that made corresponding diagnosis elusive, less with precision. This constrained human fight against malnutrition with precision.
Calorie plays important role in malnutrition. In statistics, it's hard to conduct a quantitative analysis with 'total calorie' and any other calorie-non-zero nutrients included in one statistical model due to a strong condition of 'independent and identically distributed (iid)', required by Central Limitation Theorem (CLT). Based on CLT, the common linear Regression method works for estimation, prediction or error reduction. The linear regression method is restricted in malnutrition study because of the (iid) condition.
Methods
The condition (iid) has been weakened by a breakthrough, published in a dissertation entitled "Parameter Estimation and Interpretation in Spatial Autoregression Models (SAM-p)', MSU library (1998). SAM-p resolved a decades long difficult problem in Social Network Analysis, created 'p' a new measurement in statistics, the correlation coefficient between a vector and a matrix.
Results
Mal-N-p is a specified version of SAM-p for malnutrition study. Here W(food), a recorded food intake matrix by a subject (an individual or a group of people) over a period can be transformed into W(nutrient) a recorded nutrient intake. An appropriate 'dietary reference intake (DRI)', a vector V(DRI) is selected for the subject. Based on guidelines of SAM-p working on a maximum likelihood function, an estimate of p between (0, + 1) can be obtained, as the correlation coefficient between W(nutrient) and V(DRI). The higher the p is to 1, the subject's diet behavior is 'closer' to the selected DRI and is 'better'. The estimate p is a nutrient-oriented new measurement in malnutrition study.
Conclusions
The mal-N-p algorithm can help to make DGAC science-based stronger. With a visible estimate p, mal-N-p provides encouragement to the public to self-evaluate their own diet behavior.
Funding Sources
None.