Sociobehavioral aspects of periodontal disease
W. MURRAY THOMSON, AUBREY SHEIHAM & A. JOHN SPENCER
Periodontal disease (periodontitis) is one of the most common chronic diseases suffered by adults. It is bacterially mediated inflammation that extends deep into the tissues which support the teeth, causing loss of connective tissue and alveolar bone (52). If left unchecked in susceptible individuals, it can result in the loosening and eventual loss of teeth. After dental caries, it is the leading cause of tooth loss among adults in developed countries (15, 50), and it has been shown to be associated with poorer oral-healthrelated quality of life (29). The source of the bacteria is dental plaque, the highly complex biofilm that forms on the tooth surfaces and which is disrupted by oral self-care measures such as toothbrushing and the use of dental floss.
Periodontitis has been shown to have impacts on sufferers day-to-day lives (12, 26). Periodontal treatment has been shown to ameliorate those but not to eliminate them completely (44). Thus, periodontitis is not just a threat to the dentition; it affects oralhealth-related quality of life.
Epidemiological investigating and reporting on periodontitis
The collection and reporting of epidemiological data on periodontitis is challenging (2, 6). The most important and meaningful parameter to measure in epidemiological investigations of periodontitis is periodontal attachment loss. The recording of periodontal attachment loss at six sites per tooth on all teeth within the mouth is considered to be current best practice in periodontal epidemiology (27). This involves a maximum of 168 sites per individual, as third molars are usually excluded because variations in their clinical presentation tend to unduly influence attachment-loss estimates.
In many studies, time or resource constraints preclude full-mouth or six-site recording, and partial recording protocols are used (such as half-mouth recording or three sites per tooth, or both); these result inevitably in the underestimation of attachment-loss occurrence, the degree of which is dependent upon how partial the partial recording protocol is (22, 65, 69). Nevertheless, there is some evidence that such underestimation does not unduly affect analyses which seek to identify risk indicators or risk factors (69).
The second consideration is what is actually being measured at those sites during the clinical examination. The customary approach is to measure and record separately: (i) the gingival recession, which is the distance from the cemento– enamel junction to the free gingival margin, with this being recorded as negative if there is gingival enlargement; and (ii) the pocket probing depth, which is the distance from the free gingival margin to the bottom of the pocket or gingival sulcus (27).
During subsequent data analysis, the attachment loss for each site is computed as the sum of those two measurements. Once that has been determined for each site, the derivation of person-level estimates is undertaken, as the unit of analysis for any investigations of risk factors must be the individual rather than the site (7).
In order of their usefulness for epidemiological studies, the three person-level measures are those of prevalence, extent and severity (47). The prevalence of attachment loss is expressed as the proportion of persons with one or more sites with attachment loss above a designated threshold value (such as 4 or 5 mm). It is usually reported as a percentage. The extent of attachment loss is the mean percentage of sites with attachment loss exceeding a designated threshold value (usually, but not always, the same 54
Periodontology 2000, Vol. 60, 2012, 54–63
Printed in Singapore. All rights reserved 2012 John Wiley & Sons A/S
PERIODONTOLOGY 2000 value used for the prevalence estimate). The severity of attachment loss is the mean attachment loss across all measured sites (or across all measured sites meeting a minimum threshold attachment loss value such as ‡2 mm). The problem with using severity as a measure is that the averaging of site-level measurements tends to have the effect of obscuring or flattening out important differences between exposure groups, with the result that investigations of associations with risk factors tend to be biased towards the null hypothesis. Papers presenting the findings of studies on the analytical epidemiology of periodontitis most frequently report estimates for the prevalence of the condition, with the extent of the condition being the next most frequently used estimate.
When examining the occurrence of periodontitis across populations or population subgroups, case definitions based on the person-level measures are used to reduce the information to an easily managed level. The use of a case definition for moderate or severe periodontitis brings oral epidemiology into line with much of medical epidemiology, where the presence ⁄ absence dichotomy for a disease or condition has driven most analytic approaches (63).
There are a number of case definitions for periodontitis. These case definitions are conditional statements that take the following format: if an individual has at least a specified minimum number of sites of attachment loss and at least a specified minimum number of sites with pocket probing depth, then an individual is defined as a case of periodontitis. Case definitions vary in both the thresholds used for attachment loss and pocket probing depth and the minimum number of affected sites. There are also definitions for different periodontitis severity, which are used to distinguish moderate and severe cases. Using these case definitions with population survey data leads to population prevalence estimates for periodontitis.
Such estimates can be compared across countries or different subgroups of any population. As periodontitis is age- and sex-related, it is important to standardize – either directly or indirectly – the age and sex distributions so that comparisons are valid and not confounded.