Participants and procedures
Participants were residents of Shinagawa Ward in the Tokyo metropolitan area. A total of 4617 participants studying at 17 elementary schools and their parents were enrolled. All elementary schools were similar in school–home commuting time, school starting/finishing time, and After School Care Programs without parents (ASCP) under the rules of the Shinagawa Ward school committee. All participants and parents obtained informed consent after receiving a complete description of the purpose and procedures of this study; reassurances regarding confidentiality were also provided. A total of 3006 eligible children aged 6–12 and their parents returned the questionnaire regarding sleep habits, sleep problems, and lifestyle 2 weeks after distribution by mail, giving a response rate of 65%. The present survey was conducted in late June to July 2007 to avoid the rainy season, midsummer, and winter, as climatic or other factors can affect behavior patterns.
A standardized questionnaire on sleep habits, sleep problems, and lifestyle was used in the survey.11,12 Parents were instructed to provide information regarding the usual daily lives of their children over the previous month.
The frequency of sleep bruxism was assessed by the following item: “Child grinds teeth during sleep”. Parents answered according to a four-point ordinal scale for frequency: “frequently”, “sometimes”, “occasionally”, and “never”.
Bedtime and rising time on school nights and non-school nights were used to assess sleep habits. We calculated time in bed (TIB), delay time in bedtime and rising time on weekends, and weekend sleep extensions.
Sleep problems without sleep bruxism.
The frequency of sleep problems without sleep bruxism (other sleep problems) such as sleep start, snoring, and difficulty arising were assessed by the following questions, respectively: “Child suddenly jerks his/her body, legs, or arms”, “Child snores during sleep”, and “Child can get out of bed when it is time to get up.” The frequencies of sleep starts and snoring were assessed with a four-point ordinal scale: “frequently”, “sometimes”, “occasionally”, and “never”. Difficulty arising was assessed by the following responses: “Child can get up easily”, “Child can get up with a little effort”, “Child can get up with considerable effort”, or “Child cannot get up no matter how hard he or she tries”.
Lifestyle, rearing environment, and demographic data.
We also assessed variables related to lifestyle characteristics, including the following: “Skipping breakfast” (no/occasionally/sometimes); “snack after dinner” (no/sometimes/everyday); “cram school or lessons” (no/sometimes/everyday); “television viewing, computer game playing, or Internet use (TV/computer game/Internet)”; (no/<1 h/1–2 h/<3 h/≥3 h); “using ASCP” (no/yes); “employment status of mother” (no job/part-time/flexible work-time/full-time); “frequency of mother or father returning home after 23.00 hours” (<2–3 times per month/≥1–3 times per week).
We excluded 406 participants who had the following disorders under treatment: abnormal blood pressure, cardiac disorder, digestive disorder, breathing disorder, urological diseases, hepatic disease, disorder of the nervous system, disease of the skin, sleep disturbances, autonomic dystonia, and headaches. In addition, participants were also excluded if we had no information regarding age, gender, body height, or weight. A total of 2600 participants had valid data for the present investigation.
The six grades were summarized into three categories in consideration of the eruption stage of permanent teeth: 1st/2nd grades (grades 1/2), 3rd/4th grades (grades 3/4), and 5th/6th grades (grades 5/6). During grades 1/2 the eruption stage of anterior permanent teeth occurs, during grades 3/4 the mixed dentition stage occurs, and during grades 5/6 children are right before the permanent dentition stage.10 Each category had a sufficient number of individuals for statistical testing. Children were further classified as severe bruxism group (s-Bx) and non-severe bruxism group (ns-Bx), according to the frequency of sleep bruxism by parental reports. Using these two groups as dependent factors, we performed one-way anova or the student t-test for continuous data on sleep habits and body mass index (BMI). Categorical variables, including gender, the frequency of sleep problems, and lifestyle, were analyzed by Pearson χ2-test or Fisher's exact tests, after the binary conversion of variables.
In order to locate the factors independently associated with the frequency of sleep bruxism, we performed a logistic regression analysis (stepwise forward selection). Independent factors included age and gender, bedtime, and TIB on school nights as representative markers for sleep habits. The frequency of skipping breakfast and snacking after dinner were selected as dietary habits related to regularity of lifestyle. Attending cram school or lessons, TV/computer game/Internet, and using ASCP were selected to be representative of behavior patterns after school. Employment status of the mother and the frequency of a parent returning home after 23.00 hours were selected to be representative of the rearing environment.
Before the analyses, we converted all categorical variables into binary factors: gender (0 = female, 1 = male), frequency of skipping breakfast (0 = no, 1 = occasionally/sometimes), frequency of snacking after dinner (0 = no/sometimes, 1 = everyday), attending cram school or lessons (0 = no/sometimes, 1 = everyday), TV/computer game/Internet (0 ≤ 3 h per day, 1 =≥3 h per day), using ASCP (0 = no, 1 = yes), employment status of mother (0 = no job/part-time/free, 1 = full-time), and frequency of a parent returning home after 23.00 hours (0 ≤ 2–3 times per month, 1 =≥1–3 times per week). Other sleep problems were converted into binary factors using an ordinal scale: sleep starts (high frequency group = never/occasionally, low frequency group = sometimes/often), snoring (high frequency group = never/occasionally, low frequency group = sometimes/often), and difficulty arising (facility group =“Child can get up easily”, “Child can get up with a little effort”; difficulty group =“Child can get up with considerable effort”, “Child cannot get up no matter how hard he/she tries”). Sleep habits such as TIB/bedtime on school nights and other sleep problems such as sleep starts, snoring, and difficulty arising have been adjusted by this logistic model, because they were confounding factors.
The adjusted odds ratio (OR) and 95% confidence intervals (CI) of each significant factor are presented to show the association. We conducted the statistical analysis using SPSS 16.0 software (SPSS, Tokyo, Japan). The level of statistical significance was set at P < 0.05.