Participants who attended the Menopause Clinic at Shanghai Sixth People’s Hospital took part in this cross-sectional study. Han Chinese women between the ages of 40 and 67 who were going through menopause were recruited. Exclusion criteria were (1) with rhinitis; (2) have severe internal illnesses and / or conditions such as myocardial infarction, stroke, and cancer; (3) current smoking (at least once a week for the past 6 months); (4) excessive alcohol consumption (at least one pack per month for the past 6 months); (5) have thyroid disease; (6) with tubercle and cachexia; (7) missing data. Ultimately, 715 participants were recruited into this study.
The basic socio-demographic information was obtained from a questionnaire through a face-to-face interview that was previously used  (see supplementary file 1); The variables included age, marital status, employment status, educational level, monthly income, menopause, menopause, history of chronic diseases (e.g. high blood pressure, diabetes mellitus, rhinitis, other diseases) and lifestyle (e.g. smoking, alcohol consumption) recorded. Guided tour through the stages of reproductive aging workshop (STRAW + 10) Participants were divided into three different menopausal subgroups, namely menopausal group (consecutive irregularities for more than 7 days of the menstrual cycle), early postmenopausal group (missed menstruation for 12 months – 5 years) and late postmenopausal group (missed menstruation.) For ≥5 years). Hypertension was defined by a prior diagnosis from the questionnaire or by the criteria set out in the Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation and Treatment of High Blood Pressure (JNC7, ). While diabetes mellitus was identified with FPG ≥ 7 mmol / L or was receiving diabetes treatment according to WHO criteria .
Assessment of the frequency of snoring
Participants were asked by the question to rate the frequency of sleep-snoring that was previously applied [18, 19]. “Have you snored in the past 4 weeks? And if so, how often per week? ”And the possible answers were“ never ”,“ rarely ”,“ occasionally ”and“ regularly ”, correspondingly“ never ”,“ <1 night per week ”,“ 1–2 nights per week ” "And" ≥ 3 nights per week "(see additional file 1).
Anthropometric and laboratory tests
We measured and recorded the weight and height of the participants. Body mass index (BMI) was calculated by dividing your weight in kilograms by the square of your height in meters. We measured the blood pressure of all participants on the right arm three times in a row after 5 minutes of sitting (systolic blood pressure (SBP), diastolic blood pressure (DBP)). Blood samples were taken to determine the serum concentration of triglyceride (TG), cholesterol (TC), high-density lipoprotein (HDL), low-density lipoprotein (LDL) and fasting blood sugar (FBG) after an overnight fast.
We measured body composition by BIA (TBF-418B analyzer; TANITA) of lean mass (LM), fat mass (FM) and lean mass (FFM), and each segment included upper / lower limbs and trunk. At the same time, we also recorded the basal metabolic rate (BMR) . The well-trained staff led the participants to remove heavy clothing, socks and shoes and keep the hand electrodes barefoot in contact with the electrodes on the foot pads . Fat mass (total and each segment) and lean mass (total and each segment) were given in dichotomized form, with a cutoff of the highest quartile as the higher (comparison of the highest with the lower two tertiles). We defined ≥17.11 kg, ≥1.41 kg and ≥ 9.11 kg as higher total fat mass, higher fat mass of the upper limbs and higher fat mass of the trunk, respectively.
All statistical analyzes were performed by SPSS 22.0 (IBM Corporation, Armonk, NY, USA). The data were tested for normal distribution using the Kruskal WallisH test. Levene’s test of homogeneity of variance was also performed. Variables were represented as mean ± standard deviation (SD) if they showed normal distributions, while medians (interquartile range) or values (%). One-way ANOVA (normal distributions), the Kruskal Wallis H-test (skewed continuous variables), and the χ2-test (categorical variables) were performed to compare the differences between the four groups. Snoring was analyzed as a categorical variable with never as a reference group. The relationship between body composition and frequency of snoring was calculated by multiple logistic regression analysis. The covariates included TG, TC, HDL, LDL, FBG, SBP, DBP, age, marital status, employment status, educational level, monthly income, menopausal age, menopausal status, high blood pressure, diabetes mellitus. A bilateral p <0.05 was considered significant.