The results of this study revealed that within this population, more than a quarter of women, up to two years after menarche, experienced primary dysmenorrhea, approximately 10% reported non-cyclical pelvic pain, and almost 4% suffered from dyspareunia. All chronic pain conditions were independently associated with one another. Considering that primary dysmenorrhea occurs temporally before the others, although it cannot be definitively confirmed in this study's design, it is plausible to hypothesize about a potential causal or facilitating effect of primary dysmenorrhea on other chronic pains, a notion already discussed in the literature (18).
The prevalence of primary dysmenorrhea identified falls within the range reported worldwide, but interestingly, it is nearly double the rate we previously identified in the non-indigenous population living in an urban area in the capital of Ecuador (50). In that population, the observed rate of hormonal contraceptive use was considered low, at around 25%, while in this indigenous population, the usage rate is only 5.5%, ranging from 3% among those with primary dysmenorrhea to 6% among those without this condition. We believe that this difference can be due to cultural and religious reasons. Considering that these medications are associated with a significant improvement in dysmenorrhea symptoms (51), the low frequency of usage, in our view, may be a crucial factor contributing to the increased reporting of menstrual pain by these women.
On the other hand, there is a lower prevalence of primary dysmenorrhea among indigenous women without a history of interbreeding in the family. Although our study cannot deeply discuss this difference, we can propose at least two hypotheses. The first is that there may be a specific racial and/or genomic characteristic of this population. However, this is purely speculative, as genomic data on indigenous populations are still limited (52). The second hypothesis we consider is socio-cultural significance. Menstruation in the indigenous community that maintains its deep-rooted beliefs is often characterized as "private women's business." It is sometimes seen as a sign of impurity. Stigma, secrecy, and shame associated with discussing menstruation can reduce symptom reporting among indigenous women with more conservative cultural values and taboos (53). This may make dysmenorrhea less likely to be reported in this group.
Another point that captures our attention is the association of primary dysmenorrhea with systemic arterial hypertension, although the confidence interval of the prevalence ratio has included the null value. The link with adverse cardiovascular events has already been identified by other researchers (20, 54), and this may perhaps be attributed to a systemic inflammatory status observed in these women, which, however, requires more detailed evaluation (55). Studies have shown an association between cardiovascular events not only with a higher amount of body fat but also with its distribution (56, 57). The "pear" shape, with a more homogeneous distribution of fat tissue primarily on the hips, has been associated with this (58).
In parallel, the relationship between BMI and dysmenorrhea is controversial. Longitudinal studies with large cohorts have shown evidence of a U-shaped relationship between these conditions (59), suggesting a more significant connection with body constitution than weight itself. Finally, a large British cohort has demonstrated significant associations between body shape and inflammatory and metabolic biomarkers (60).
In our study, the prevalence of non-cyclical pelvic pain was similar to that observed in Latin American countries such as Brazil, where it's close to 10% (61, 62), and in urban communities of Ecuador, where it is 8.9% (50). It was positively associated with various other painful conditions, reinforcing the link between the condition and nociplastia (63), and perhaps reflecting the clinical expression of central sensitization that commonly occurs in this group of patients (64). These findings are also supported by the apparent protective effect identified in the practice of physical exercise and non-cyclical pelvic pain. Recent literature has shown that physical exercise can strengthen the modulation promoted by the central nervous system (65, 66), reducing pain sensitization (67), and there is a direct inverse relationship between measures of physical activity and chronic pain levels (68).
The observed prevalence of dyspareunia in this population was significantly lower than that previously reported in the urban population of Ecuador (50). The exact reason for this difference remains uncertain. Regardless of the low prevalence of dyspareunia in the studied population, what is equally remarkable is the fact that virtually all the women who reported pain during sexual intercourse did not discontinue intercourse for that reason. Taken together, we believe that this finding could be attributed to the patriarchal structure of indigenous society, where "male" attitudes prevail, exacerbating the social and biological vulnerability to which indigenous women are historically subjected. They have often been tied to familial and communal roles, hindering their ability to express their desires and preferences (http://repositorio.utn.edu.ec/handle/123456789/6165) (69).
Despite the absence of affirmative responses regarding the presence of intrauterine infections, 66.3% of the women reported frequent and concurrent urinary symptoms, which could potentially be linked to infectious processes secondary to Neisseria gonorrhoeae and Chlamydia trachomatis, causative bacterial agents of urethritis and pelvic inflammatory disease (70). An underdiagnosis of pelvic inflammatory disease might also explain the association with previous pregnancy losses, as there is a connection between intrauterine infection and abortions (71). The findings concerning urinary discomfort contrast with those related to intrauterine infections, which exhibit a notably low prevalence.
We did not observe any association between pain and psychological symptoms, smoking, alcoholism, or violence, as we had observed in other studies conducted in Ecuador and Brazil. One possible justification for this could be the low reported prevalence of these conditions in this specific community, which is close to or less than 1% for each.
Strengthens and Limitations
Our study's strength lies in the inclusion of a large and representative cohort of the Kichwa indigenous population. Furthermore, it was conducted with meticulous methodological rigor, granting it robust inferential power. However, there are certain limitations associated with both the characteristics of the population and the analysis itself.
Ethnic self-identification, defined as "the right of every person to freely and voluntarily decide whether or not to belong to a nationality or people" may have also altered the population distribution of the women studied, as each individual can choose to identify themselves accurately or erroneously with a particular nationality or people, even if they do not genuinely belong.
The results of this study may also be subject to biases primarily related to the indigenous and Andean worldviews. These worldviews, shaped by beliefs, values, and knowledge systems, play a pivotal role in the social life of these human groups and define their cultural identity. It is crucial to emphasize the social and biological vulnerability of indigenous women, who unfortunately remain entrenched in economic, social, and cultural inequalities, where patriarchal social structures persist. Even though all our interviewers were women representing the local indigenous community, which facilitated the feasibility of the research, some topics are still considered taboo, particularly concerning sexual activity, illicit substance use, tobacco, alcohol, psychological symptoms, and violence. We believe this might have influenced the identification of independently associated factors and, in some way, hindered the formulation of education and healthcare policies tailored to indigenous women, especially with regard to dyspareunia.
Regarding the analysis, certain aspects are inherent to logistic regression models. For our study, the backward stepwise approach proved to be the most suitable model for the data. It may be influenced by the relationship between the number of candidate variables and the sample size, but this was not a concern given our relatively large cohort. Including all variables would add significant complexity and could potentially compromise the model's generalizability. To mitigate this, we conducted a correlogram and aimed to avoid including highly correlated variables in the simulated models. Nonetheless, this was done judiciously to prevent the premature exclusion of relevant variables.
It is also challenging to ensure that all potential combinations of predictors have been tested. The significance of the p-value does not always equate to clinical relevance, making the interpretation of the effect (in this case, odds ratio) crucial. Moreover, it is impossible to establish a causal relationship between the outcomes and associated variables, which is a limitation inherent to cross-sectional studies.
On the other hand, backward elimination allows for the advantage of initially considering the effects of all variables simultaneously, which is especially important in cases of potential collinearity, as mentioned earlier. Other factors that balance its limitations include its ease of application, objectivity, reproducibility, interpretability, and the enhanced generalization achieved by reducing the number of predictor variables.