Since 1985, only four studies have been published that present data on attachment loss in populations. The purpose of this study was to present the incidence of attachment loss over an 18‐month period in a representative sample of community‐dwelling older adults. In addition, the utility of multivariate prediction models to distinguish between people who will and will not experience disease progression was explored. The Piedmont 65 + Dental Study of the Elderly is a longitudinal investigation of a random sub‐sample of over 1,000 community‐dwelling people over the age of 65 in five contiguous North Carolina counties. The five dentists who conducted the in‐home examinations and interviews at baseline also participated at 18 months and examined the same subjects. The overall attrition rate for blacks was 27% and the rate for whites was 23%. At 18‐months, the baseline attachment level scores for those subjects who were lost from the study were not significantly different from the scores of those who remained in the study. A change in attachment loss of 3 mm or more over the 18 months was set as a conservative estimate of actual change taking place. Blacks were significantly more likely to experience attachment loss irrespective of the number of sites involved (P <.05). Approximately half the people had at least one losing site and about one‐third of the blacks and one‐fourth of the whites had at least two losing sites. A total of 24% of blacks and 16% of whites had 3 or more losing sites. The multivariate etiologic model for blacks implicates the presence of P. gingivalis and P. intermedia in concentrations greater than 2%. In addition, people who indicate that they do not floss regularly, that their memory is getting worse, and had dental visits more than 3 years ago were at higher risk of attachment loss over the 18‐month period. The etiologic model for whites indicates that people who had P. gingivalis present in concentrations greater than 2% were at higher risk for attachment loss than people who did not have P. gingivalis; that people who sought medical care in the last 6 months were at higher risk; that people who felt depressed were at higher risk; and that people who smoke cigarettes regularly were at higher risk. J Periodontol 1994; 65:316–323.

This study investigated salivary levels of Streptococcus mutans, lactobacilli, and caries experience in a random sample of 448 black and 362 white older dentate adults living in North Carolina. Significant proportions of the participants had stimulated salivary flow rates less than 1.0 mL/min, salivary buffering capacity less than 4.0, S. mutans levels of 10(5) cfu/mL or more in stimulated whole saliva, or lactobacilli levels of 10(5) cfu/mL or more. Each of these factor levels could be considered, on a clinical basis, to increase caries risk. In general, people with higher levels of S. mutans or lactobacilli had more untreated coronal and root caries, but not greater total caries experience.