We present a form of parametric survival analysis that incorporates exact, interval-censored, and right-censored times to deciduous tooth emergence. The method is an extension of common cross-sectional procedures such as logit and probit analysis, so that data arising from mixed longitudinal and cross-sectional studies can be properly combined. We extended the method to incorporate and estimate a proportion of agenic teeth. While we concentrate on deciduous tooth emergence, the method is relevant to studies of permanent tooth emergence and other developmental events.
Deciduous tooth emergence data were analyzed from four longitudinal studies. The samples are 1,271 rural Guatemalan children examined every three months up to age two and every six months thereafter as part of the INCAP study; 397 rural Bangladeshi children examined monthly to age one and quarterly thereafter as part of the Meheran Growth and Development Study; 468 rural Indonesian children examined monthly as part of the Ngaglik study; and 114 urban Japanese children examined monthly in studies from 1910 and 1920. Although all four studies were longitudinal, many observations from the Guatemala and Bangladesh studies were effectively cross-sectionally observed. Three different parametric forms were used to model the eruption process: a normal distribution, a lognormal distribution, and a lognormal distribution with age shifted to shortly after conception. All three distributions produced reliable estimates of central tendencies, but the shifted lognormal distribution produced the best overall estimates of shape (variance) parameters. Estimates of emergence were compared to other studies that used similar methods. Japanese children showed relatively fast emergence times for all teeth. Bangladeshi and Javanese children showed emergence times that were slower than are found in most previous studies.
Estimates of agenesis were not significantly different from zero for most teeth. One or two central incisors showed significant agenesis that ranged from 0.1 to 0.8% in three of the samples; even so, failure to model the agenic proportion did not seriously bias the estimates.