Background and Purpose: Aortic calcification is a major risk factor for cardiovascular disease (CVD) related deaths. More than 50% of patients die without clinical symptoms. This is a need to risk-stratify asymptomatic individuals with sub-clinical atherosclerosis, for identification of highest risk patients to initiate intervention. The use of non-invasive diagnostic tools for identification of advanced atherosclerotic plaques by x-ray analysis is receiving increased attention.  By automated analysis of Morphological Atherosclerotic Calcification Distribution (MACD) in lumbar x-rays, we investigated the relation between mortality and biological aspects of plaque dynamics such as of number, size, morphology and distribution in the lumbar aorta of postmenopausal women. We compared the MACD index to the Framingham AC24 score and correlated those to baseline demographic and biochemical risk factors.  
Methods: 308 women aged 48 to 76 were followed for 8.3±0.3 years and cardiovascular, cancer and all-cause deaths were recorded. Several aortic calcification markers were quantified: number, morphology and distribution from outlines of the calcified plaques in lumbar X-rays at baseline. These markers were compared to SCORE card, Framingham score, and the Aortic Calcification Severity score - AC24, and correlated to baseline demographics and all standard traditional risk-factors.  
Results: AC24 adjusted by age, waist circumference, and triglyceride levels predicted mortality in postmenopausal women (CVD p=0.03, All-cause p=0.006). The SCORE card and the Framingham score resulted in mortality odds ratios (OR) of 5.0 and 5.2 - defining high risk as =6 and =18, respectively. All scores based on the calcification geometry provided highly significant predictions. The MACD index provided an OR of 20 which was significantly higher than the AC24 index OR of 5.0 and any other single or multivariate metabolic/physical marker, alone or in combination. The MACD index, compared to that of AC24, was significantly correlated to baseline glucose levels (p<0.01 R=0.4), and better correlated to triglyceride levels.  
Conclusions: The newly established MACD-index that encompasses information on plaque morphology, distribution and growth potential provides additional biological information emphasizing that smaller plaques with a spread elongated morphology have a larger growth potential and thereby subsequent rupture potential. This results in a superior identification of CVD related death with an OR of 20.  This increased biological information of the MACD index correlated strongly to glucose levels suggesting that this assessment method entails information of pre-diabetic status that is an important risk factor of plaque development, rupture and subsequent death.