Study setting and population
This cross-sectional survey was conducted during January-March 2018 in China. The inclusion criteria for the survey participants were 1) aged 30 years or older at the interview; and 2) in a state of clear consciousness. The exclusion criterion for the survey was the participants didn’t have data on osteoporosis awareness. The participants were identified and interviewed face-to-face by medical students in Jilin University (Enrolment year: 2017; Specialty: Preventive Medicine) using a structured questionnaire. The participants for this survey included the relatives, friends and neighbours of the students. All medical students were trained by related researchers together before conducting the survey. This study was approved by the Ethical Committee Board at School of Public Health, Jilin University. Each participant also provided written-informed consent to this study.
Study measures
Socio-demographics (sex, age, body weight, height, residence, education level, and family annual income), lifestyle information (smoking, alcohol use), prior fracture and prior bone mineral density test were collected using a structured questionnaire. Body mass index (BMI) was calculated as body weight (kg) divided by squaring of body height (m2). Educational level was classified as primary and below (Years 1-6), junior (Years 7-9), senior (Years 10-12) and undergraduate and postgraduate educations (Years 13 and above). Residence was classified as urban, rural and cross regions. Cross region is an overlapping area between urban and rural area that has both urban and rural.
We assessed osteoporosis awareness level using the following domains, including by definition, diagnosis, signs/symptoms, treatment, complications, prognosis, causes, risk factors, and prevention of osteoporosis; the questionnaire about osteoporosis awareness was the same as a previous published research [13]. The reliability and validity of the questionnaire was tested among 30 Chinese subjects prior to the formal survey. This questionnaire had a good internal consistency (Cronbach’s α = 0.746). Under factorial validity test, the related components explained a cumulative 60% of the variance in the awareness scores. The Kaiser-Meyer-Olkin measure of sampling adequacy and Bartlett’s test of sphericity both showed that the results were suitable for factor analysis. Besides the osteoporosis awareness questions, we also collected the sources of the participants acquiring their existing osteoporosis knowledge (e.g., newspapers and magazines, advertising leaflets, television or radio health program).
Statistical analysis
We used descriptive statistics to describe the characteristics of the study population and main variables. Continuous variables were shown as mean ± standard deviation (SD); categorical variables were shown as percentages.
Awareness scores were created by assigning a “1” to each correct response and a “0” to each incorrect or “unsure” response. The items were summed for a possible range of 0 to 29, with higher scores reflecting greater awareness. Awareness score was defined as percent of correct answer and individual domain scores were calculated - an average of the percentages of correct answers to each question.
To test the relationship between risk factors and overall awareness score we used multiple linear regression models. It was used to assess the association between overall awareness score and risk factors, with for all covariates such as sex, age, body mass index, residence, educational level, family annual income, prior bone mineral density test, prior fracture, smoking, and alcohol use. Covariates are selected based on their significant in the univariate test.
All statistical analyses were performed by using SPSS software (version: 25.0; SPSS Inc, Chicago, IL).