In our sample, there were 2058 Bangladeshi school students, with more than half girls (63%). A total of 35.2% were 14 years old, 34.6% were 15 years old, 20% were 13 years old, 6.2% were 16 years old, 3.1% were 12 years old, 0.7% were 17 years old, and 0.3% were 11 years old. The sample was dispersed throughout all of the different grades in the school: 27.6% were in 7th grade, 10.4% in 8th, 51.3% in 9th, and 10.7% in 10th. Most participants possessed normal body weight (81%), 7.5% were overweight, 10.2% were underweight, and 1.3% had obesity (not shown).
The prevalence estimates and chi-squared p-values of the suicide-related outcomes, psychological states, and health-related behaviors stratified by sex are shown in the supplementary file, Table 1. Overall, 49.2% of the sample was sufficiently active (not shown). There were significant sex differences in the proportion of young people in each of the activity subgroups: more girls than boys were in the least active category, whereas more boys were in the somewhat active and most active categories. Within the most active group, the sub-group with less PA among boys than girls (18.9% vs. 30.3%) was selected as being sufficiently active according to WHO requirements. Just under half of the sample (48.4%) reported participation in 3 or more days in physical education classes, while one-quarter (26.7%) reported participation in 5 or more days.
Table 1
Frequency of physical activity, bicycling, physical education class and sitting activity, association with suicidal behavior for boys and girls
Characteristics | Suicidal Ideation Boy Girl OR (95% CI) | Suicidal Plan Boy Girl OR (95% CI) | Suicidal Attempt Boy Girl OR (95% CI) |
Frequency of activity (Days per week) | Least active (0–2 days) | Reference |
Somewhat active (3–5 days) | 0.8 (0.87 0.91) | 0.6 (0.64, 0.67) | 0.7 (0.77, 0.80) | 1 (1.02, 1.08) | 0.2 (0.28, 0.29) | 0.8 (0.87, 0.92) |
Most active (6–7 days) | 1.1 (1.08, 1.11) | 2.3 (2.27, 2.35) | 2.1 (2.098, 2.143) | 1.4 (1.40, 1.44) | 0.9 (0.97, 1.00) | 1.2 (1.24, 1.28) |
Walk or Bicycle to school (Days per week) | Least active (0–2 days) | Reference | |
Somewhat active (3–5 days) | 0.5 (0.58, 0.61) | 0.6 (0.62, 0.66) | 0.4 (0.46, 0.47) | 0.7 (0.74, 0.79) | 0.5 (0.58, 0.60) | 2.1 (2.02, 2.22) |
Most active (6–7 days) | 2.2 (2.25, 2.31) | 1.7 (1.70, 1.75) | 2.5 (2.52, 2.58) | 1.2 (1.25, 1.28) | 1.7 (1.77, 1.82) | 1.3 (1.36, 1.40) |
Physical Education Class Attendance (Days per week) | 3 or more days | 0.8 (0.82, 0.84) | 1.3 (1.29, 1.33) | 0.6 (0.65, 0.66) | 1.6 (1.61, 1.66) | 0.4 (0.43, 0.44) | 1.5 (1.54, 1.59) |
5 or more days | 0.6 (0.66, 0.68) | 1.4 (1.42, 1.47) | 0.3 (0.38,0.39) | 1.9 (1.88, 1.93) | 0.2 (0.24, 0.25) | 1.1 (1.10, 1.14) |
Sitting activity | 3 or more hour per day | 1.1 (1.08, 1.120 | 1.5 (1.55, 1.62) | 0.8 (0.826, 0.84) | 0.8 (0.81, 0.84) | 1.3 (1.36, 1.400) | 1.3 (1.27 1.33) |
P value = 0.000, OR = Odd Ratio, CI = Confidence Interval |
The odds ratio (OR) and the p-values for suicidal behavior-related outcomes and PA-related behaviors stratified according to sex are shown in Table 1. Overall, higher frequency of activity and "walk or bicycle to school" behaviors were associated with lower levels of suicide-related behavior. All outcomes of physical education class attendance and suicidal ideation, ideation and attempt were significantly higher OR for girls than boys. Boys had significantly higher ORs for suicidal plans and attempts in sitting activity behaviors than girls. Although girls had the higher OR for suicidal ideation when sitting for 3 or more hours per day.
Table 2 presents the characteristics of the sample and the sex-adjusted OR of psychosocial behavior by specified characteristics. Both male and female students who were most active per day over the last week compared to those who were not were 2 times and 1.5 times (OR 2.1, OR 1.4), respectively, less likely to have no close friend. In addition, male students who were most active in the last 7 days compared to those who were least active were almost three times (OR 2.9) less likely to suffer from anxiety. Females who did not sit for three hours or more were 63% and 58% less likely to be bullied and feel lonely, respectively, than those who did. Students who did not attend 5 or more days in PE class compared to those who did have a 37% (male) and 22% (female) lower risk of being bullied.
Table 2
Associations between frequency of physical activity, bicycling, physical education class and sitting activity, psychosocial behavior for boys and girls
Characteristics | Anxiety Boy Girl OR (95% CI) | Bully Boy Girl OR (95% CI) | Loneliness Boy Girl OR (95% CI) | No close friend Boy Girl OR (95% CI) |
Frequency of activity (Days per week) | Least active (0–2 days) | Reference |
Somewhat active (3–5 days) | 0.3 (0.34, 0.35) | 0.4 (0.44, 0.47) | 0.6 (0.66, 0.67) | 0.3 (0.38, 0.40) | 0.3 (0.37 0.39) | 0.4 (0.41, 0.43) | 11.7 (11.01,12.45) | 0.9 (0.94, 0.98) |
Most active (6–7 days) | 2.9 (2.87, 2.97) | 0.9 (0.89, 0.93) | 1.2 (1.20, 1.22) | 0.8 (0.88, 0.90) | 0.72 (0.73, 0.75) | 1.067 (1.052, 1.083) | 2.140 (2.113, 2.167) | 1.4 (1.39, 1.43) |
Walk or bicycle to school (Days per week) | Least active (0–2 days) | Reference |
Somewhat active (3–5 days) | .4 (0.422, 0.43) | 0.5 (0.50, 0.53) | 0.6 (0.61, 0.63) | 0.5 (0.54, 0.56) | 0.4 (0.43, 0.44) | 0.3 (0.36, 0.37) | 0.8 (0.79, 0.83) | 0.9 (0.92, 0.97) |
Most active (6–7 days) | 2.5 (2.52, 2.620 | 1 (1.00, 1.04) | 0.8 (0.86, 0.88) | 0.8 (0.87, 0.88) | 1.5 (1.56, 1.59) | 1 (1.07, 1.10) | 0.5 (0.52, 0.54) | 1.2 (1.26, 1.29) |
PE attendance (Days per week) | 3 or more days | 0.9 (0.93, 0.96) | 1.4 (1.43, 1.48) | 0.9 (0.93, 0.94) | 1.7 (1.77, 1.80) | 1 (1.01, 1.03) | 1.1 (1.13, 1.16) | 0.9 (0.90, 0.93) | 0.6 (0.662, 0.67) |
5 or more days | 0.8 (0.86, 0.90) | 1.7 (1.75, 1.81) | 1.3 (1.36, 1.38) | 1.2 (1.20, 1.23) | 1 (1.04, 1.069) | 1 (1.05, 1.09) | 1.1 (1.16, 1.19) | 0.5 (0.53, 0.55) |
Sitting activity | 3 or more hour per day | 0.7 (0.70, 0.73) | 2 (1.98, 2.07) | 0.9 (0.95, 0.97) | 1.6 (1.61, 1.66) | 1 (1.04, 1.06) | 1.5 (1.55, 1.61) | 1 (1.06, 1.09) | 1.76 (1.736, 1.78) |
*P value = 0.787, ** P value = 0.004, P value = 0.000, OR = Odd Ratio, CI = Confidence Interval |
Adolescents with risky-health behaviors, including ever having sexual intercourse (girls vs. boys, OR 3 vs. 0.7), ever abusing drugs (boys vs. girls, OR 13.5 vs. 7.4), and smoking (girls vs. boys, OR 2.8 vs. 0.0), had a greater likelihood of being in the most active group in terms of frequency of activity. Girls than boys students who were most active in terms of frequency of activity, walking or bicycling to school groups, attended 3 or hours days in PE class including setting activity had a lower chance of engaging in sexual intercourse. While boys had significantly higher OR for most active group in frequency of activity, walking or bicycling to school groups, attended 3 or hour per days in PE class, setting activity and drug abuse than girls. In addition, boys had a higher OR for both outcomes of PE class attendance and smoking, as illustrated in Table 3.
Table 3
Associations between frequency of physical activity, bicycling physical education class and sitting activity, sexual intercourse, drug abuse and smoking for boys and girls
Characteristics | Sexual Intercourse Boy Girl OR (95% CI) | Drug Abuse Boy Girl OR (95% CI) | Smoking Boy Girl OR (95% CI) |
Frequency of activity (Days per week) | Least active (0–2 days) | Reference |
Somewhat active (3–5 days) | 0.3 (0.30, 0.31) | 0.8 (0.80, 0.85) | 2.7 (2.67, 2.79) | 10.7 (10.34, 11.15) | 1.2 (1.14, 1.3) | 0.6 (0.64, 0.67) |
Most active (6–7 days) | 0.7 (0.72, 0.73) | 3.060 (2.993, 3.128) | 13.5 (13.32,13.87) | 7.4 (7.03, 7.82) | 0.0 (0.09, 0.10) | 2.8 (2.81, 2.90) |
Walk or Bicycle to school (Days per week) | Least active (0–2 days) | Reference |
Somewhat active (3–5 days) | 0.8 (0.789, 0.81) | 0.5 (0.52, 0.55) | 1.3 (1.30, 1.36) | | 3.9 (3.68, 4.180 | 1.3 (1.325, 1.413) |
Most active (6–7 days) | 0.7 (0.74, 0.76) | 1.63 (1.60, 1.67) | 4.2 (4.15, 4.29) | 1.6 (1.59, 1.67) | 0.2 (0.21, 0.22) | 1.5 (1.55, 1.60) |
Physical Education Class Attendance (Days per week) | 3 or more days | 0.4 (0.42, 0.42) | 2 (2.05, 2.13) | 1.4 (1.41, 1.46) | 0.3 (0.36, 0.389) | 3.6 (3.60, 3.70) | 0.9 (0.96, 0.99) |
5 or more days | 0.6 (0.61, 0.631) | 0.9 (0.88, 0.92) | 0.2 (0.27, 0.28) | 0.4 (0.39,0.42) | 4.2 (4.18, 4.28) | 1.3 (1.36, 1.40) |
Sitting activity | 3 or more hour per day | 0.9 (0.95 0.97) | 1.9 (1.94, 2.04) | 3.6 (3.64, 3.75) | 0.3 (0.295, 0.33) | 0.7 (0.75, 0.78) | 0.8 (0.84, 0.88) |
Tobacco product = Tobacco other than cigarettes, P value = 0.000, OR = Odd Ratio, CI = Confidence Interval |
The frequency of physical and sitting activity, as well as the use of tobacco products, were significantly higher in girls than in boys. There were large differences in the adolescent lifestyle in terms of frequency of activity. Male adolescent alcohol abuse was lower with more exercise. Most activity groups were associated with lower levels of alcohol abuse in female adolescents who walk or use bicycles to school. For both girls and boys, subgroup comparisons revealed significant differences in the possibility of being physically abused between the somewhat active and most active subgroups and the least active subgroups, with the somewhat active and most active group having lower levels of the possibility of being physically abused. Non-significant differences in OR were observed for girls who did not attend 5 hours or more in PE class and were physically abused compared with those who did. In the currently tobacco products user and physically abused adolescent categories, girls had a significantly higher OR than boys students in the sitting 3 or more hours per day group, as shown in Table 4.
Table 4
Associations between frequency of physical activity, bicycling physical education class and sitting activity, alcohol and physically abuse for boys and girls
Characteristics | Tobacco Products Boy Girl OR (95% CI) | Alcohol Abuse Boy Girl OR (95% CI) | Physically Abuse Boy Girl OR (95% CI) |
Frequency of activity (Days per week) | Least active (0–2 days) | | | Reference | |
Somewhat active (3–5 days) | 1.4 (1.41, 1.46) | 2.1 (2.10, 2.21) | 0.3 (0.31, 0.32) | 0.1 (0.17, 0.19) | 1.3 (1.34, 1.37) | 0.5 (0.513, 0.53) |
Most active (6–7 days) | 0.5 (0.50, 0.52) | 1.1 (1.15, 1.19) | 4.2 (4.16, 4.35) | 0.9 (0.86, 0.96) | 3.3 (3.329, 3.37) | 0.7 (0.78, 0.79) |
Walk or Bicycle to school (Days per week) | Least active (0–2 days) | | | Reference | |
Somewhat active (3–5 days) | 0.7 (0.69, 0.73) | 0.8 (0.80, 0.86) | 0.1 (0.193, 0.20) | 0.3 (0.31, 0.35) | 1.5 (1.54, 1.58) | 1.9 (1.92, 1.99) |
Most active (6–7 days) | 0.8 (0.83, 0.85) | 0.9 (0.91, 0.94) | 0.9 (0.95, 0.99) | 7.8 (7.32, 8.42) | 2.1 (2.11, 2.13) | 1.2 (1.28, 1.30) |
Physical Education Class Attendance (Days per week) | 3 or more days | 1.7 (1.68, 1.728) | 1 (1.0 1.03) | 0.5 (0.58, 0.60) | 2.1 (2.05, 2.24) | 0.9 (0.90, 0.91) | 1* (0.99, 1.00) |
5 or more days | 1.2 (1.24, 1.27) | 1.1 (1.12, 1.16) | 0.3 (0.31, 0.33) | 0.3 (0.33, 0.38) | 0.9 (0.90, 0.91) | 0.98** (0.98, 0.99) |
Sitting Activity | Sitting 3 or more hour per day | 1.3 (1.28, 1.32) | 12.3 (11.60, 13.24) | 6.5 (6.47, 6.71) | | 0.8 (0.83, 0.84) | 1.2 (1.22, 1.25) |
*P value = 0.628, ** P value = 0.006, P value = 0.000, OR = Odd Ratio, CI = Confidence Interval |
Figures 1 and 2 depict the sex-stratified distributions of the three and two clusters for girls and boys, respectively, including the significant predictors. The silhouette measurements for girls were good, while the measurements for boys were fair. Physical abuse indicators were the most important predictors of girls' scores 1. For boys, they were mostly subjected to bullying by others. All suicidal and psychosocial behaviors were recognized as key predictors among girls and boys.
Table 5 described the estimates of the relative contributions of suicidal, psychosocial, and risky-health determinants with the frequency of PA behavior, based on gender. The initial models only included risky health behaviors for girls and psychosocial factors for boys. The R2 values for the frequency of PA increased by 0.006 and 0.020, respectively, in models that independently contained psychosocial indicators for girls and risky-health indicators for boys. The final model included all three groups of variables, and the final R2 values for this model were 0.014 for girls and 0.051 for boys.
Table 5
Hierarchical regression model for frequency of physical activity
Sex | Model | R Square | P Value |
Female | Risky-health (Lifestyle) | 0.018 | 0.00 |
Risky-health (Lifestyle) + Psychosocial | 0.020 | 0.00 |
Risky-health (Lifestyle) + Psychosocial + Suicidal behaviors | 0.026 | 0.00 |
Male | Psychosocial | 0.018 | 0.00 |
Psychosocial + Risky-health (Lifestyle) | 0.036 | 0.00 |
Risky-health (Lifestyle) + Psychosocial + Suicidal behaviors | 0.046 | 0.00 |
*For boys: Psychosocial = Bully, No close friend, Loneliness & anxiety; Risky-health (Lifestyle) = Physically abuse, Smoking & Tobacco other than cigarette; Suicidal behaviors = Suicidal plane, ideation and attempt. *For girls: Risky-health (Lifestyle) = Physically abuse, Smoking & Tobacco other than cigarette; Psychosocial = Bully, No close friend, Loneliness & anxiety; Suicidal behaviors = Suicidal plane, ideation and attempt. |