This cross-sectional study was based on secondary data from panoramic digital examinations of patients 50 years and older. The age of 50 was selected, as it has been reported that age-related loss of bone was more prominent in women after the age of 50 compared to those below 50.[16] This study was approved by the Ethical Committee of the Faculty of Dentistry, Universitas Indonesia.
Panoramic radiographs from January 2017 to December 2018 were chosen. An initial 439 digital panoramic radiographs were collected from the radiology medical records of the study hospital. After evaluating the quality of the radiographs in the area to be interpreted, eight radiographs were found not to be in the panoramic standard head position. Incomplete variable data and radiographs that showed cysts involving the jaw, benign or malignant tumors, suspected or self-reported systemic conditions, and jaw fractures that affected the numbers of remaining teeth were not included in the study. For subjects who underwent panoramic radiographic examinations more than once during the 2017–2018 period, only the final radiographic examination was used if it met the inclusion criteria. A total of 428 subjects of panoramic radiographs were measured in the study. All subjects were Indonesian.
The numbers of remaining teeth (including the third molars) were evaluated on panoramic radiographs using the following categories: healthy, carious, or treated. Root tips indicated for extraction were not included as the remaining teeth. Measurement of the MIC was determined by observing the mandible distally from the mental foramen on both sides. Subjects were divided into three groups (C1–C3), according to the Klemetti method[7] (Figure 1). C1 (normal cortex) indicated that the endosteal margin of the cortex was even and sharp on both sides, C2 (mildly to moderately eroded cortex) indicated that the endosteal margin showed semi-lunar defects (lacunar resorption) or appeared to form endosteal cortical residues on one or both sides, and C3 (severely eroded cortex) indicated that the cortical residues were clearly porous on one or both sides.
All digital radiographs were taken with the Veraviewepocs 2D® (J. Morita, Kyoto, Japan). The Veraviewepocs 2D® orthopantomograms were taken with a high-resolution Charge-Coupled Device (CCD) sensor (32-bit microprocessor) (J. Morita, Kyoto, Japan). The panoramic exposure was made at ten milliampere-seconds (mAs) for 12–15 s at 70–80 kVp. Radiographs were taken by the same radiographer. One well-trained examiner with ten years’ experience determined the MIC classification. To calculate the intra- and inter-examiner agreement, a dentist was trained to measure MIC classification. After a period of calibration, 100 panoramic radiographs were chosen randomly and reanalyzed within a period of one week. The kappa agreement for intra- and inter-agreement was 0.88 and 0.80, respectively. Based on the guidelines of the interpretation of kappa statistics performed by Landis and Koch,[17] the following categories were considered: 0.00 (poor), 0.00–0.20 (slight), 0.21–0.40 (fair), 0.41–0.60 (moderate), 0.61–0.80 (substantial) and 0.81–1.00 (almost perfect). The kappa value of intra- and inter agreements was in the category of almost perfect and substantial, respectively.
Statistical analysis
The results of the MIC classification were divided into two groups. Group 1 comprised MIC class 1 and class 2 which are still considered normal. Group 2 consisted of MIC class 3 which is considered at high risk of osteoporosis. The homogeneity of the collected data was assessed, and a non-normal distribution was observed. Mann-Whitney nonparametric tests were used to verify the relationship between MIC classification and the number of remaining teeth. Finally, multiple linear regression analysis was used to assess the relationship between MIC classification and the number of remaining teeth after controlling for confounding factors. The number of remaining teeth was used as a dependent variable. MIC classification (C1 and C2:0; C3:1), age (group 1: 50–64 years old; group 2: 65–88 years old), gender (female:0; male:1) used as an independent variable. The level of significance was set at P < 0.05 for these tests. Statistical analysis was performed using SPSS software for Windows, version 20.0 (SPSS, Inc, Chicago, IL, USA).