Most studies indicate that there are differences in lip patterns between males and females. However, nine studies demonstrated the opposite. After a detailed analysis, it was found that there is no specific factor that explains why nine studies found no differences between sexes and the others found differences between sexes, apart from methodological flaws related to the sampling method, which may have conditioned the result achieved by each study. At the same time, several factors may explain the divergence of results within the group of studies that found sex differences and within the group of studies that found no differences.

One of the factors that may have influenced the result of each study concerns the difficulties in analysis arising from the age heterogeneity of the samples. In several studies, the authors included participants from a wide age range [8, 17, 33, 36, 41, 47,48,49], disregarding age stratification. This procedure may induce errors in the print analysis, because depending on the individual’s age and, therefore, the developmental stage of the lips, the labial lines may show some blurring, making it difficult to correctly analyze or differentiate the sex of the individual. According to Mamandras, between 8 and 18 years old, the lips vary constantly in length and thickness, reaching their maturity in late adolescence. In females, the upper lip reaches maturity at 14 years old and the lower lip at 16 years old, whereas in males, both the lower and upper lips reach the end of development at 18 years old [50]. Around 30 years old, the signs of aging begin to appear around the mouth, but the lips maintain their tone [51]. From 40 years old and with advancing age, wrinkles grow in the skin adjacent to the lips, the intercomissural distance increases, and lip height decreases. These age-related effects do not change the type of lip furrows [52], but the change in the natural lip anatomy can make it difficult to identify the lip pattern and, subsequently, lead to errors in registration. Therefore, it is possible that up to 18 years old and from 40 years old onwards, there is difficulty in analyzing lip prints, caused by the growth and aging of the lips, respectively. It seems that the 18 to 40 age group is the most effective in analyzing and identifying the individual’s lip pattern. The study by Randhawa et al. precisely proved this, since in the sample distributed over three age groups, the greatest accuracy of cheiloscopy in sex estimation was obtained in the group of individuals aged between 21 and 40 years old [34]. Similarly, in the study by Multani et al., the highest percentage of correct identifications was achieved in the age group of 36 to 45 years old. In individuals above 45 years old, there was greater difficulty in correctly identifying the sex [44]. Chaudhari et al. mentioned lip changes due to age as the possible reason for the results obtained [49].

There is a notorious tendency of researchers to justify the inexistence of sex differences with the age heterogeneity of the sample. However, from the results of the present review, this reason may not be enough to explain the results since several studies included individuals outside the age range of 18 to 40 years old and were able to find differences between sexes. Therefore, it is intended to clarify that difficulties in analysis due to lip growth and aging, and subsequent error in recording, occur regardless of whether there may be differences between sexes. However, to avoid their influence on the results, it would be more appropriate for the studies to be carried out in age-stratified samples.

On the other hand, it is possible to see, in several studies, marked sample differences between the proportion of male and female participants concerning the target population [7, 36, 53,54,55,56], which may again influence the results. In these cases, the sample size of one of the sexes may have been insufficient and may have conditioned obtaining a different predominant pattern than would have been identified if the sample had been representative.

Another factor that may have influenced the results is the use of a sample consisting of several population groups [29, 45, 53]. In the study by Peeran et al., for example, the inability to differentiate the various ancestral patterns among the sampled population is even presented by the authors as the main limitation. The southern Libyan population, the region where the study was conducted, is composed of individuals of various ancestries [53], which may have been a decisive factor in the lip print analysis since there are studies that showed that the lip pattern varies between different population groups [12, 42]. Thus, if this hypothesis is proven, it would be more reliable for the analysis of sex differences to be performed among people of the same ancestry or population group.

Thus, the difficulty in the analysis caused by age heterogeneity, the lack of sample representativeness, and the diversity of ancestry or population groups may have influenced the results of different studies.

Considering the group of studies that found differences between sexes and the group of studies that did not find the same results, it is possible to verify that within each group, and even between groups, the results differed, especially in the predominant lip pattern in each sex. These discrepancies can be explained by several factors, especially by the geographic origin of the sample. As mentioned earlier, there seems to be an evidence that lip pattern varies between individuals from different population groups. Thus, if this is proven, it is to be expected that the predominant pattern in males and females may vary depending on the population group to which they belong.

The sampling methodology also explains the different results obtained, starting with the sample size, which ranged from 20 to 1399 participants, and only one study selected participants randomly [17]. It is natural that in a sample of 20 individuals, the same predominant patterns may not be found as in a sample of a thousand individuals, for example, especially when dealing with convenience samples, where people are not randomly selected. In fact, given the variability of the factors under study, selecting the appropriate sample size is an important step to make an inference of an outcome to the population. The sample size should be calculated in advance, considering the population size and the frequency of events, to ensure that the results obtained from the sample can be generalized to the population [57]. Understandably, it is limiting to conduct studies with representative samples of populations as large as those studied by the included articles. However, it is important to consider the detriment of conducting studies of reduced power or the risk of obtaining biased results.

Another factor that can explain the differences in the predominant pattern of each sex among the various studies presented is the area of the lip considered to classify the labial grooves, which varied between the studies. With such diverse analysis areas and knowing that the type of lines and their frequency vary throughout the labial mucosa [58], it is possible that such a factor explains the discrepancies observed between the studies. So, while some investigators studied the numerical superiority of the type of lines present on the whole lip, segmented or not, others only studied the numerical superiority of the type of lines found on a restricted lip portion. Thus, it is to be expected that when analyzing the whole lip print, one will find a particular type of pattern that may not be the same as that found when analyzing only a smaller portion (Fig. 6).

Representative illustration of the influence of the lip area under study in determining the predominant pattern. (LL, lower lip)

The same is true when the whole lip is analyzed and segmented in different ways. That is, analyzing the lip divided into four parts, for example, produces different results from those achieved when analyzing the lip divided into six, as exemplified in Fig. 7.

Representative illustration of the influence of the segmentation of the lip area under study in determining the predominant pattern

Another parameter that can make a difference is that some authors determine the predominant pattern according to the pattern that prevailed in the total segments of each sex, while others determine the predominant pattern of each sex according to the pattern that prevailed. For example, in the study by Ramakrishnan et al., a sample of 100 individuals was analyzed. Lip prints were divided into four quadrants, and in the results, the authors report that 33% of the total quadrants of the males showed type I [17]. This calculation, for the total number of quadrants, gives a different result from that which would be found if the determination of the lip pattern were made considering the predominant pattern of each of the individuals of the sex under analysis. Figure 8 provides an example. In a sample of 10 males, the lip prints are divided into four quadrants, and the analysis is done according to the numerical superiority of the type of lines observed in each quadrant. Next, the lip pattern of the print under analysis is determined, considering the pattern that dominated in each of the four quadrants. The conclusion is reached that types I' and II are the predominant patterns in males. However, when considering the totality of the quadrants, the pattern that prevails in a larger number of segments is type V. In 11 of the 40 quadrants, type V dominates. This leads to the conclusion that type V is the predominant pattern in males. As can be seen, depending on how one intends to calculate the predominant pattern, the result may vary.

Representative illustration of lip print analysis by quadrants. Within each rectangle is shown the predominant pattern of the lip print represented above

During the analysis of lip prints, although most authors used the full S&T classification, it was also possible to identify some studies that applied the modified classification [33, 36, 41, 47, 53, 59]. The change in classification, mostly by adding or omitting some types of lines, results in data that would differ from those found if the original classification was applied. This further contributes to the discrepancy in results between studies [25, 40, 41, 53].

Regarding the sex estimation classification proposed by Vahanwala et al. [24], its correct use by the studies raises doubts, since, according to the classification authors, it should be applied to the analysis of the four quadrants of the lip, and not to different or smaller areas, as was done in three studies [34, 42, 44]. Now, using a classification developed to be applied to all four quadrants of the lip, in completely different zones, will produce invalid results. To clarify whether this sex estimation classification could be applied to different areas of the lip, other than the four quadrants, the lead author of the classification was contacted, in October 2021. However, as of the date of submission of this paper, no response has been obtained.

The heterogeneity of the techniques used to collect and analyze lip prints may also have been determinant in obtaining different results, since the way these techniques are performed may affect the correct reading of the prints and, consequently, influence the results of each study [7, 17, 18, 32, 48, 60].

Regarding collection, the method employed is a fundamental step to ensure print quality. Recording the lip print is a technique-sensitive task and, therefore, depending on how the print is collected, its quality may vary. Thus, choosing the most appropriate method is essential to ensure the success of the analysis. Costa and Caldas [61] tested methods 1 to 4 and found that the application of lipstick, without rubbing the lips, followed by transfer to cellophane tape (method 1) is the method that provides the best lip print reading. This was the collection method most used by the articles included in the systematic review. A 2010 study comparing, among other collection techniques, methods 2 and 3, showed that the latter is the most appropriate because of the good quality of the print, low technical difficulty, and speed of the procedure [62]. Regardless of the advantages they may present, the main limitation reported by studies using conventional methods is the amount of lipstick applied that, in excess, can decrease the print quality [54]. Evidence shows that prints taken with a thinner layer of lipstick have better quality [63].

The pressure applied during collection and the direction can also alter the appearance of the prints [58] and consequently affect correct identification. Human lips are naturally mobile [18], and therefore, the pattern of lip wrinkles depends on how the muscle relaxes to produce the print [5]. When the muscle relaxes, the mouth remains closed, and a lip print with well-defined lines is produced. On the contrary, in the open-mouth position, the lines are less perceptible and therefore more difficult to classify [64]. The lips’ mobility explains why the same person can produce lip prints with different appearances according to the pressure and direction applied. To overcome this limitation and to avoid errors in classification, some researchers have used the photography technique to record lip prints instead of the traditional lipstick and paper recording method [18, 31]. From a practical point of view, photography has a great advantage in that it is more convenient for the subject since he does not need to apply lipstick, as would be the case with conventional methods. Suspects of a crime often resist the collection of their biometric data. Having to apply lipstick to them and take their lip print by conventional methods would be a difficult task, and worse if more than one print must be taken to ensure the best quality. Furthermore, taking the print using cellophane tape may be painful for the individual as it may cause small lesions on the lips [40]. Therefore, several authors suggest photography as the most appropriate method for taking lip prints [31, 65]. To do so, it is very important to create good lighting conditions, as mismatched shadow and light areas may influence the quality of the images [8].

Regarding analysis, the direct method was the most used, essentially because it is very practical and simple. But, if on the one hand, it is easy to use, on the other hand, it may not offer the best visualization of the prints. In latent lip prints, for example, the use of alternate light sources, like blue or green light, may allow for better results [66]. Similarly, other techniques involving filters, polarized filters, contrast, and other imaging software techniques have also offered good print visualization [67]. As for the indirect method, image-editing software allows for improving the visualization of prints by adjusting brightness, color, contrast, or enlarging details [48, 68]. In this way, the same lip print, with some imperceptible or overlapping details, may see its quality improved with the use of image editing software, while the simple use of the magnifying lens would not allow it. Thus, the analysis method applied may also influence the results since better visualization of the prints will certainly lead to an increase in correct analyses. Other imaging software, such as Adobe 7.0, has already been used by other authors, who reported better visualization, ease in identification, and recording of the lip print pattern [69].

Despite the discrepancy of results and most studies indicating the existence of differences between sexes, namely those better ranked as to the risk of bias, it is important to retain that there is no specific pattern for each sex, which makes the effectiveness of reconstructive analysis for sex estimation relative and of questionable practical utility.