In the present study, a total of 201 human long bones (humeri, femora, tibiae) were subject to PMI determination. Presumably, the bones had lain in the ground during the PMI until they were legally unearthed after expiration of the grave leases. They were then discovered and taken by a “bone collector.” Although the exact PMI of the respective bones was unknown, it could be narrowed down to a likely range of 20 to 100 years, based on the results of the police investigation and our own investigation, especially after evaluating the valid laws, the regulations of the cemeteries and the age of the oldest graves on the relevant parts of the cemeteries. However, it should be noted that we cannot exclude a PMI of slightly more than 100 years of individual respectively a small number of the examined bones.

The bones had then presumably been stored in the home of the “bone collector” during the period between their discovery, or removal from the respective cemeteries, and their confiscation by the police. From the authors’ point of view, this period of time, as well as the subsequent storage period of around two years at the Institute of Forensic Medicine, should not be counted towards the PMI, as it can be assumed that the decomposition process did not proceed significantly under these conditions.

The bones were assessed in regard to their macroscopic appearance, UV-fluorescence of a freshly sawn cross-section, and the chemiluminescence intensity of the luminol reaction of the bone meal obtained when setting the saw cut.

The extent of UV-fluorescence on the freshly sawn cross-section showed a weak to, at best, moderate, significant correlation with the macroscopically estimated PMI. These results fuel the disagreement arising in previous studies as to whether UV-fluorescence is an appropriate method to determine PMI, or whether it even provides valid results. Basically, for decades, UV-fluorescence has been seen as an effective way of assessing PMI, particularly in regard to the exclusion of a forensically relevant period [11, 15, 19, 21,22,23]. In light of the fact that the exact cause for UV-fluorescence of bone is still largely unknown, this circumstance seems all the more astonishing. So far, the results from various studies indicate that the protein component of the collagen in the bone substance could be responsible for the blue fluorescence in relatively fresh bone samples [24, 25], while, with increasing PMI, the change in mineral content with a simultaneous decrease in the collagen content could explain the rather yellowish to brownish fluorescence of older bone material [26]. In a more recent study by Hoke et al. [27], both the intensity and the color of bone UV-fluorescence in relation to PMI were investigated in a sample consisting of 58 cemetery bones with a known PMI between 8 and 60 years - which roughly corresponds to the PMI range of the bone collective in our present study. Because no significant correlation between the extent of UV-fluorescence and PMI was found by Hoke et al., they regarded this method of differentiating between recent and historical bones to be doubtful. On the other hand, they were able to establish a weak correlation between the color of the UV-fluorescence and the PMI. Similarly, Sterzik et al. were also able to demonstrate a significant correlation between the extent of fluorescent UV-light and 490 nm light, respectively, the color of fluorescence and the PMI [28, 29]. The correlation found by Sterzik et al. was, however, stronger than that found Hoke et al. [27].

Our investigations showed that with (macroscopically estimated) increasing PMI, there was a significant increase in the proportion of bones that showed only slight or no blue fluorescence of the freshly sawn cross-section under UV-light, and none of the bones from the PMI group 4 showed complete UV-fluorescence. In contrast, the finding that 6 of the 15 bones in PMI group 4 still showed at least patchy remnants of UV-fluorescence suggests that already a reduction in blue fluorescence may indicate a PMI beyond forensic relevance and complete absence of UV-fluorescence is perhaps not required. This finding is also consistent with the results from studies by Facchini et al. [22] and Hoke et al. [27], who were still able to detect remnants of blue UV-fluorescence in bones that were demonstrably several hundred to over 10,000 years old. These results even led Hoke et al. to conclude that it is impossible to specify an upper PMI limit for the occurrence of blue fluorescence. Furthermore, Hoke et al. discuss several factors that can significantly influence fluorescence intensity: The infiltration of Mn2+ or Fe2+ ions could, for example, reduce fluorescence; hence, the burial environment could play a decisive role in this respect. Conversely, collagen cross-linking could lead to an increase in fluorescence, and a correlation between age at death and fluorescence intensity should, therefore, also be taken into consideration.

In our collective, 6 of the 113 bones in the (shortest) PMI group 1 no longer showed UV- fluorescence and thus yielded a false-negative result. In the study by Hoke et al. [27], over 50% of all bones in the PMI range up to 60 years of age were also incorrectly determined as not forensically relevant. Consequently, a negative result (in terms of intensity) cannot, with certainty, rule out a forensically relevant PMI. In view of the legal consequences that would result from such a misjudgment, this level of uncertainty seems unacceptable.

In our investigations, the luminol test did not show a negative correlation between the degree of chemiluminescence of the luminol reaction and the macroscopically estimated PMI in our PMI group 1. Instead, it even showed a weakly positive correlation, which is quite surprising in view of previous study results. In previous studies, the luminol test has been described as a method that appears to be fundamentally capable of estimating the PMI, particularly in combination with other methods. Despite the long-established knowledge of the role of hemoglobin and its degradation products as catalysts of the luminol reaction [30], first described by Specht in 1937, it still does not seem entirely clear which components of bone samples examined with the luminol test cause chemiluminescence. In a study by Ramsthaler et al. [15], other examination methods such as the Hexagon-OBTIⓇ test and the ComburⓇ test were applied to bone samples, each of which yielded negative results for the presence of hemoglobin. Possible explanations for this finding are discussed in the aforementioned work. Irrespective of the question of which components in bone samples lead to the chemiluminescence of the luminol reaction, Ramsthaler et al.’s study yielded a weakly positive to strongly positive result in 14 of 16 samples in the critical PMI range of 10–100 years, a finding that is consistent with the results of our study. In addition, 30% of the bones with a PMI > 100 years (bones with a PMI > 1000 years included) in Ramsthaler et al.’s study yielded a positive luminol reaction. Based on other studies, the explanation for false-positive results may lie in numerous environmental factors that can cause positive chemiluminescence even though hemoglobin is no longer present in the bone [13, 31, 32]. In summary, these studies show that while the risk of a false-positive result is relevant, a negative result can, conversely, be seen as a good predictor for a PMI that is no longer forensically relevant [14,15,16, 19]. In this context, it should be noted that only a comparatively small number of bone samples with forensically relevant PMI were analyzed in the studies mentioned above. This stands in contrast to the number of bones with a forensically relevant PMI that were examined in our study. The explanation for the statistically significant correlation that was found in previous studies between the degree of chemiluminescence of the luminol method and the PMI may, therefore, be that sample collectives with sometimes very large PMI ranges, and thus also very old bone finds, were analyzed in those studies. In contrast, the present study was the first to analyze a larger number of bone samples with a more homogenous and shorter PMI range of up to 100 years with the luminol test. This circumstance may explain why no correlation between the degree of luminol chemiluminescence and PMI was found for our study collective.

However, it must be stressed that the exact PMI of the bone samples in our collective was not known and had only been estimated using macromorphological criteria. The fact that PMI estimates formed the basis for the classification of the luminol chemiluminescence intensity and the UV-fluorescence degree can be seen as a limitation of our study. Nonetheless, because of the quite well known lower and upper limit of the PMI, it can be seen as suitable for the present study, because it wasn’t the aim of the study to determine the exact PMI of the bones, but to compare the different methods.