The Association Between Alcohol and Dysmenorrhea in University Students in North China CURRENT STATUS: UNDER REVIEW

Background Evidence supporting the definitive effect of alcohol consumption on dysmenorrhea has been sparse. The current study was aim to evaluate the relationship between alcohol consumption and dysmenorrhea among age-stratified female college students in northern China. Methods A total of 3692 female college students were included in this cross-sectional study. The logical regression model was performed to evaluate the association between alcohol consumption and primary dysmenorrhea. The model adjusted for confounding factors such as age and body mass index, and estimated the odds ratio (ORs) and 95% confidence interval (CIs). Results Multivariable-adjusted models showed the analysis stratified by age at menarche (AAM) revealed that the prevalence of dysmenorrhea in participants with AAM < 13 years old (67.2%) was significantly higher than that in participants with AAM ≥13 years old (61.3%) ( P = 0.003). Alcohol consumption showed a dose-response relationship with dysmenorrhea in participants with AAM ≥13 years old. Conclusion Our findings showed AAM modified the association of alcohol consumption with primary dysmenorrhea among female university students in North China.


Introduction
Dysmenorrhea is a gynecological disorder that affects nearly half of women of reproductive age [1,2] and is characterized by cramping pain in the lower abdomen that lasts for 1-3 days before or at the onset of menstruation [3]. Dysmenorrhea has important physical, behavioral, psychological and social repercussions, and is often accompanied by fatigue, headache, sweating, dizziness, diarrhea, nausea and vomiting [4]. It is worth noting that dysmenorrhea is a common gynecological disease among female college students. It is reported that the prevalence rate of dysmenorrhea among university students ranges from 46 to 93% in many countries [5][6][7][8][9][10][11] and from 44 to 56% in China [12]. This leads to poor performance in activities, and even reduce the quality of life. [13,14]. Dysmenorrhea is a gynecological disorder that affects 40−70% of women of reproductive age. The prevalence of dysmenorrhea is related to many factors, such as young AAM [11, [15][16][17], long menstrual cycle and amount of menstrual bleeding[11, 15,16], active or passive smoking and alcohol use [5,16], stress [18], family history of dysmenorrhea [5, 16,19], eating regularity [5,20], sleep hygiene [10], and body mass index [16].
Current health trends are stressing the importance of appropriate alcohol use among female university students.
To date, no studies have reported a conclusive effect of alcohol use on dysmenorrhea. The aim of the present study was to determine the prevalence of dysmenorrhea in university students, and to evaluate whether alcohol use (scored as "yes" or "no") and alcohol consumption (scored as "no", "at least once a month", or "at least once a week") was associated with dysmenorrhea in an AAMstratified population of university students in North China.

Study population
This study was based on a large-scale epidemiological study conducted in North China from May 2016 to Aug 2016. The study evaluated whether alcohol use (scored as "yes" or "no") and alcohol consumption (scored as "no", "at least once a month", or "at least once a week") were associated with dysmenorrhea in female undergraduate students after AAM-stratified.
A total of 3,837 subjects age≥18 volunteered to participate in the study and completed the questionnaire and examination. Subsequently, 20 participants were removed from statistical analysis due to incomplete data, especially incomplete data regarding dysmenorrhea. Moreover, 125 were excluded from the study owing to exclusion criteria. Finally, a total of 3692 female college students were enrolled into this study using the following criteria.
Inclusion criteria for the study were: (1) full-time female university student, and (2) nulliparous.
Exclusion criteria were: (1) diagnosis of gynecological or endocrine disease, (2) current active smokers or active smoker history, and (3) oral contraceptive user.

Questionnaire design
Each participant was asked to learn detailed information about the questionnaire before completing it. The information included the objective of the study and instructions on filling out the questionnaire.
A standard anonymous questionnaire was used for this study. The questionnaire was designed to obtain information about participant-reported outcome for dysmenorrhea, demographic characteristics, and environmental exposures. A brief definition of primary dysmenorrhea was attached to the questionnaire, which was explained as "two or more days of menstrual pain during menstrual bleeding (with respect to the baseline)". Pain was defined as abdominal or low back pain. A positive dysmenorrheal history was identified as "any menstrual pain during the previous 12 months" [21]. Participants should not have a medical history of other gynecological or endocrine disease, or oral contraceptives within the previous 12 months [22]. Passive smoking and alcohol consumption were assessed through self-reporting. Passive smoking was defined as the inhalation of second-hand smoke, which was queried as "Were you exposed to someone smoking indoors in your presence?" Alcohol use was queried as "Did you drink beer, white wine, or red wine?", and was scored as "Yes" or "No". Alcohol consumption was queried on three levels and scored as follows: "No", "At least once a month", or "At least once a week".

Statistical analysis
The characteristics of the study participants with respect to the primary dysmenorrheal status, which is coded as Yes or No, are presented in Table 1. Continuous variables are reported as mean ± standard deviation (SD), whereas categorical variables are reported as percentages. The chi-square test or t-test was used to test for differences between the two dysmenorrheal groups. Differences in the prevalence of primary dysmenorrhea with respect to alcohol consumption are presented in Table   2. The effects of alcohol consumption on primary dysmenorrhea are presented in Table 3. The effects of alcohol consumption on primary dysmenorrhea after AAM-stratified are presented in Table 4. We assessed the association of alcohol consumption with primary dysmenorrhea using logistic regression models. The logistic regression model to estimate odds ratios (ORs) and 95% confidence intervals (CIs) was adjusted for confounding factors including age, body mass index, household income, mother's education, mother's history of dysmenorrhea, passive smoking, area of residence, and AAM.

Data description
The average age of menarche in our study was 13.42 ± 1.30 years old. Participants with dysmenorrhea tended to have lower menarche age (Table 1). Alcohol use ("Yes" or "No") and alcohol consumption ("No", "At least once a month", or "At least once a week") were significantly associated with dysmenorrhea (Table 2).

Logistic regression models
However, these differences were not significant after adjusting for confounding variables ( Table 3).
The association between alcohol use and the prevalence of dysmenorrhea was consistently stronger among participants with AAM ³13 years old than among participants with AAM <13 years old, with point estimates of 1.06-1.88 and 0.55-1.42, respectively. Alcohol consumption had a dose-response relationship with dysmenorrhea. The adjusted ORs for participants with AAM ³13 years who consumed alcohol at least once a month and at least once a week were 1.29 (CIs, 0.94-1.78) and 1.92 (CIs, 1.07-3.45), respectively, compared with those who did not consume alcohol. AAM modified the associations of alcohol use and alcohol consumption with primary dysmenorrhea (P value for interaction = 0.017) (Table 4).

Discussion
The results of this study can be divided into two levels. First, without sample stratification, we founded that there was no association between alcohol consumption and dysmenorrhea among university students in North China. However, when the samples are stratified based on the AAM, we founded that participants with AAM ≥ 13 years old had higher prevalence of dysmenorrhea, and the prevalence rate of dysmenorrhea showed a dose-response association with alcohol consumption.
Our study is one of the few efforts to investigate the association between alcohol consumption and dysmenorrhea risk among female college students. Dysmenorrhea was reported to be associated with a variety of factors, such as mother's history of passive smoking[5] and dysmenorrhea [19].
Nevertheless, few studies have explored the relationship between alcohol consumption and dysmenorrhea. The results of the unstratified multivariate analyses we studied were consistent with previous reports by several scholars, Sznajder et al. [12] reported that alcohol use was not associated   [15][16][17]. Our results showed that the age among participants with dysmenorrhea was significantly younger than that of participants without dysmenorrhea (P < 0.001). It is suggested that the younger the AAM, the greater the risk of dysmenorrhea, which is consistent with the results of Harlow et al. [35] Their study found that the earlier the AAM, the higher the incidence, duration and severity of dysmenorrhea. Their study found that the earlier the AAM, the higher the incidence and the longer the duration, and the more severe the pain.
According to studies, the effect of stratification of AAM on the association between alcohol consumption and dysmenorrhea may due to the following reasons. First, the association between younger than 13 years old of AAM and dysmenorrhea may be due to higher levels of estrogen caused by hormone patterns in the early stages of sexual maturity [36]. Second, Participants with AAM < 13 years old might reduce their alcohol use and consumption, or avoid alcohol use, due to their higher frequency of dysmenorrhea. Third, for participants older than 13 years old of AAM, the function of the adrenal and hypothalamic pituitary-gonadal axis (HPG) may mature later. In this case, alcohol consumption might disrupt the unstable maturation process, which may lead to many of the physical and hormonal changes and even dysmenorrhea.
The study had some limitations. First, the study was based on a cross-sectional sample that lacked a longitudinal trace, which might lead to selection bias. Second, alcohol use, alcohol consumption, and dysmenorrhea were evaluated by means of a retrospective questionnaire, which might be subject to recall bias. Accurate responses from the study participants were crucial for the study validity. Third, we do not have enough samples to examine the effects of alcohol consumption and alcohol consumption on dysmenorrhea in female college students younger than 13 years of age. Fourth, several underlying factors that could affect the magnitude of dysmenorrhea were not sufficiently investigated, including age at first alcohol use, disease which could cause dysmenorrhea, lifestyle, drug abuse, exercise, and genotypic variation. Future studies including these additional factors are needed. Finally, the questionnaire did not query about the occurrence of dysmenorrhea before the start of alcohol use, which may possibly have a confounding effect on the association of alcohol use and alcohol consumption with dysmenorrhea, due to the participants with dysmenorrhea may tend to avoid using alcohol or reduce alcohol consumption.

Conclusion
In summary, our study revealed that alcohol consumption showed a dose-response relationship with dysmenorrhea in participants with AAM ≥13 years old. AAM-stratified clearly shows for the first time that there is a significant association between alcohol consumption and dysmenorrhea. Therefore, AAM may be a critical factor for modifying alcohol consumption and dysmenorrhea in university students in North China. Future research is needed to confirm our findings and to elucidate the underlying mechanisms.

Consent to publish
We have obtained consent to publish from the participant to report individual patient data.

Competing interests
The authors declare that they have no competing interests.  All values are percent of total subjects; other values are mean ± SD.  a Logistic regression model was adjusted for age, body mass index, household income, mother's education, mother's history of dysmenorrhea, passive smoking, area of residence, and age at menarche. Table 4. Crude and adjusted odds ratio (OR) and 95% confidence interval (CI) of dysmenorrhea association with alcohol use and alcohol consumption stratified by menarche age a Logistic regression model was adjusted for age, body mass index, household income, mother's