3.1. Descriptive statistics
Table 1
Characteristics of male and female stratified by smoking intensity, CHNS 1991-20151-2
Table 1 Characteristics of male and female stratified by smoking intensity, CHNS 1991–20151−2 |
Characteristic | | Male (N = 35,907) | | Female (N = 39,441) |
| Nonsmoker (n = 14,572) | Light (n = 8,706) | Moderate (n = 10,452) | Heavy (n = 2,177) | P | | Nonsmoker (n = 38,235) | Light (n = 799) | Moderate (n = 369) | Heavy (n = 38) | P |
| |
Age(year) | | 42.7(0.1) | 42.1 (0.1) | 43.5(0.1) | 46.3 (0.2) | < 0.001 | | 42.7 (0.1) | 50.3(0.4) | 50.3(0.6) | 49.9 (0.2) | < 0.001 |
Married (%) | | 79.6(0.0) | 83.0(0.1) | 88.3(0.0) | 92.6(0.0) | < 0.001 | | 86.1(0.2) | 84.1(1.3) | 84.8(1.9) | 89.5(0.5) | 0.314 |
Education level (%) | | | | | | | | | | | | |
Primary/illiterate | | 25.7(0.4) | 31.6(0.5) | 34.8(0.5) | 39.2(1.0) | < 0.001 | | 43.3(0.3) | 67.7(1.7) | 75.0(2.3) | 50.0(2.2) | < 0.001 |
Junior | | 36.6(0.4) | 36.3(0.4) | 39.9(0.4) | 36.8(0.4) | < 0.001 | | 30.1(0.2) | 20.0(0.4) | 16.5(1.9) | 10.5(1.0) | < 0.001 |
High /above | | 36.4(0.4) | 30.9(0.5) | 24.1(0.4) | 22.2(0.9) | < 0.001 | | 24.1(0.2) | 9.0(0.1) | 5.4(0. 1) | 31.6(0. 8) | < 0.001 |
Income(1 k yuan) | | 19.1(0.5) | 15.5 (0.3) | 15.5 (0.3) | 16.7(0.3) | < 0.001 | | 12.5(0.14) | 7.8(0.14) | 6.9(0.62) | 14.1(3.6) | < 0.001 |
Employed (%) | | 74.4(0.4) | 80.7(0.4) | 83.4(0.4) | 83.9(0.4) | < 0.001 | | 65.2(0.2) | 52.4(0. 2) | 52.3(0.3) | 52.6(0.3) | < 0.001 |
PA(METs/w) | | 236.9(1.8) | 274.0 (2. 5) | 293.6(2.3) | 302.8(5.0) | < 0.001 | | 277.4(1.2) | 315.0(9.5) | 314.8(14.4) | 257.0(14.8) | < 0.001 |
Drinking (%) | | 48.4(0.01) | 72.2(0.01) | 72.2(0.01) | 74.0(0.01) | < 0.001 | | 9.7(0.2) | 26.8(0.2) | 30.9(0.2) | 34.2(0.7) | < 0.001 |
Dietary indicators | | | | | | | | | | | | |
Energy (1 k kcal) | | 2.4(0.0) | 2.4(0.0) | 2.5(0.0) | 2.6(0.0) | < 0.001 | | 2.1(0.0) | 2.1(0.0) | 2.0(0.1) | 1.9(0.0) | 0.076 |
Fat (E %)3 | | 30.4(0.1) | 29.4(0.1) | 29.9(0.1) | 30.4(0.3) | < 0.001 | | 30.3(0.1) | 28.0(0.4) | 30.1(0.7) | 29.6(2.00) | < 0.001 |
Obesity indicators | | | | | | | | | | | | |
WC (cm) | | 83.2(0.1) | 81.8(0.1) | 81.6(0.1) | 82.7(0.2) | < 0.001 | | 79.0(0.1) | 80.0(0.4) | 80.2(0.7) | 82.0(1.7) | < 0.001 |
BMI (kg/m2) | | 23.3(0.0) | 22.8(0.0) | 22.7(0.0) | 23.0(0.0) | < 0.001 | | 23.0(0.0) | 23.1(0.1) | 23.3(0.2) | 23.4(0.6) | 0.611 |
General Obesity (%) | | 8.7(0.2) | 6.1(0.1) | 5.5(0.0) | 7.1(0.0) | < 0.001 | | 8.1(0. 1) | 10.5(0. 1) | 13.8(0. 2) | 13.2(0. 6) | < 0.001 |
Abdominal Obesity (%) | | 34.6(0.41) | 28.1(0.5) | 28.7(0.4) | 32.2(0.4) | < 0.001 | | 36.3(0.0) | 41.2(0.0) | 43.1(0.0) | 55.3(0.1) | < 0.001 |
Notes: (1) Date comes from CHNS longitudinal data, from 1991 to 2015. (2) The statistics reported are the sample mean with the standard deviation in parentheses. (3) Fat (E %): percentage of energy intake from fat. |
Table 1 shows the characteristics of male and female participants stratified by daily cigarette consumption from CHNS 1991–2015. Of the 75,348 adult participants, 35,907 (47.7%) were male and 39,441(52.3%) were female. The average age of nonsmokers was about 42.7 years, with the oldest being male heavy smokers (46.3 years) and female light smokers (50.3 years). There were significant differences in the distribution of education level, working status, and individual income by cigarette consumption regardless of gender (p < 0.001). Average physical activity (PA) increased from 236.9 METs/w in male nonsmokers to 302.8 METs/w in male heavy smokers. Among females, the highest PA was 315.0 METs/w in light smokers and the least was 257.0 METs/w in heavy smokers. The average drinking rate of current smokers was significantly higher than that of nonsmokers regardless of gender (p < 0.001). Moreover, as smoking intensity increased, rate of drinking rose significantly (p < 0.001). Male heavy smokers had the highest dietary energy intake (2600 kcal) and percentage of dietary energy intake from fat (30.4%), while female nonsmokers had the highest percentage of dietary energy intake from fat (30.3%). The BMI, WC, general obesity, and abdominal obesity rates of male nonsmokers were significantly larger than those of male current smokers, whereas the opposite held true for females. In addition, it is important to note that, among current smokers, male heavy smokers had larger BMI (23.0 kg/m2), WC (82.7 cm), general obesity (7.1%), and abdominal obesity (32.2%) than other smokers (p < 0.001).
Figure 1 shows the shift in distribution of smoking status in Chinese adults by gender. Among males, nonsmokers and heavy smokers increased from 31.0% and 5.5% in 1991 to 47.4% and 6.8% in 2015, respectively (p < 0.001). However, male moderate smokers and light smokers decreased from 32.7% and 30.8% in 1991 to 23.8% and 22.0% in 2015, respectively (p < 0.001). In general, the trend of the smoking intensity distribution in women was consistent with that in men. Specifically, female nonsmokers and heavy smokers increased from 96.0% and 0.05% in 1991 to 98.3% and 0.1% in 2015, respectively (p < 0.001). Female moderate smokers and light smokers decreased from 3.30% and 0.65% in 1991 to 1.10% and 0.50% in 2015, respectively (p < 0.001).
Figure 2 shows the shift in the distribution of BMI and WC by smoking status among different genders. It is noteworthy that BMI and WC are increasing significantly both in current smokers and in nonsmokers regardless of gender (p < 0.001). However, among males, both the BMI and WC of current smokers are higher than those of nonsmokers, which is opposite to the distributions among females (p < 0.001).
3.2. Longitudinal relationship of smoking behaviors and body weight outcomes (BMI and WC) in Chinese males and females from 1991 to 2015
Figure 3 shows that the prevalence of general obesity and abdominal obesity increased significantly in both current smokers and nonsmokers in the provinces of mainland China surveyed by CHNS from 1991 to 2015 (p < 0.001). In addition, among current smokers with different smoking intensities, the prevalence of general obesity and abdominal obesity in males is significantly higher than in females (p < 0.001).
The results of the multilevel mixed-effects linear regression modeling used to examine the effects of smoking behaviors on BMI and WC are presented in Table 2. In general, compared with nonsmokers, male current smokers showed significant negative effects on BMI and WC. After controlling for confounding factors, the net effect on male light smokers, moderate smokers, and heavy smokers was a significant decrease of 0.40 kg/m2(β: −0.40, 95% CI: − 0.48, − 0.31), 0.51 kg/m2 (β: −0.51, 95% CI: −0.60, − 0.43), and 0.29 kg/m2 (β: −0.29, 95% CI: −0.43, − 0.14) in BMI, respectively. Female light smokers and moderate smokers had net effects of significant decrements of 0.61 kg/m2 (β: −0.61, 95% CI: −0.86, − 0.35) and 0.66 kg/m2 (β: −0.66, 95% CI: −1.06, − 0.26) in BMI, respectively. The reductions in cigarette consumption from moderate smokers and light smokers to nonsmokers were linked with significant WC decreases of 0.87 cm (β: −0.87, 95% CI: −1.14, − 0.58) and 0.86 cm (β: −0.87, 95% CI: −1.12, − 0.59) in males and 1.58 cm (β: −1.58, 95% CI: −2.94, − 0.22) and 1.37 cm (β: −1.37, 95% CI: −2.13, − 0.59) in females, respectively.
3.3. The effects of smoking behaviors on general obesity and abdominal obesity in Chinese males and females from 1991 to 2015
The odds ratios of the multilevel mixed-effects logistic regression model used to examine the effects of smoking behaviors on general obesity and abdominal obesity risks are summarized in Table 3. After controlling for confounding factors, the odds ratio (95% CIs) for general obesity in males were 0.72 (0.65–0.81) for light smokers and 0.68 (0.61–0.76) for moderate smokers compared with nonsmokers. Similarly, the odds ratio (95% CIs) for abdominal obesity in males were 0.80 (0.76–0.86) for light smokers and 0.81 (0.76–0.87) for moderate smokers compared with nonsmokers. Moreover, the odds ratio (95% CIs) for abdominal obesity in females were 0.83 (0.71–0.98) for light smokers and 0.73 (0.58–0.91) for moderate smokers compared with nonsmokers. However, there was no significant effect of smoking intensity on general obesity among females
Table 2 Regression coefficients of BMI and WC according to smoking intensities among Chinese male and female, CHNS 1991–20151−4 |
Variables | | Male | | Female |
| Nonsmoker (n = 14,572) | Light (n = 8,706) | Moderate (n = 10.452) | Heavy (n = 2.177) | P trend | | Nonsmoker (n = 38,235) | Light (n = 799) | Moderate (n = 369) | Heavy (n = 38) | P trend |
| |
BMI(kg/m2) | | | | | | | | | | | | |
Model 1 | | Ref | -0.29*** (-0.38,-0.21) | -0.38*** (-0.46,-0.30) | -0.08** (-0.22,-0.07) | ༜0.001 | | Ref | -0.01 (-0.26, 0.24) | -0.03 (-0.43, 0.37) | -0.11 (-0.16,0.94 | 0.923 |
Model 2 | | Ref | -0.32*** (-0.40,-0.28) | -0.45*** (-0.53,-0.37) | -0.23*** (-0.38,-0.09) | ༜0.001 | | Ref | -0.59*** (-0.83,-0.33) | -0.63*** (-1.04,-0.23) | -0.52 (-1.59,0.54) | ༜0.001 |
Model 3 | | Ref | -0.39*** (-0.48,-0.32) | -0.50*** (-0.59,-0.42) | -0.28*** (-0.42,-0.13) | ༜0.001 | | Ref | -0.60*** (-0.85,-0.34) | -0.64*** (-1.04,-0.24) | -0.54 (-1.60,-0.51) | ༜0.001 |
Model 4 | | Ref | -0.40*** (-0.48,-0.31) | -0.51*** (-0.60,-0.43) | -0.29*** (-0.43,-0.14) | ༜0.001 | | Ref | -0.61*** (-0.86,-0.35) | -0.66*** (-1.06,-0.26) | -0.53 (-1.58,-0.53) | ༜0.001 |
| | | | | | | | | | | | |
WC(cm) | | | | | | | | | | | | |
Model 1 | | Ref | -0.58*** (-0.86,-0.30) | -0.53*** (-0.79,-0.27) | -0.36 (-0.46,-0.30) | 0.054 | | Ref | 1.03*** (0.26,1.81) | 0.82 (-0.54,2.19) | 2.02 (-1.03,5.08) | ༜0.001 |
Model 2 | | Ref | -0.62*** (-0.89,-0.34) | -0.69*** (-0.95,-0.42) | -0.07 (-0.53,0.39) | ༜0.001 | | Ref | -1.38*** (-2.15,-0.61) | -1.63*** (-2.99,-0.28) | 0.23 (-2.85,3.30) | ༜0.001 |
Model 3 | | Ref | -0.86*** (-1.14,-0.59) | -0.84*** (-1.10,-0.57) | -0.19 (-0.65,0.28) | ༜0.001 | | Ref | -1.33*** (-2.10,-0.55) | -1.54*** (-2.90,-0.18) | 0.33 (-2.73,3.40) | ༜0.001 |
Model 4 | | Ref | -0.87*** (-1.14,-0.58) | -0.86*** (-1.12,-0.59) | -0.19 (-0.66,0.27) | ༜0.001 | | Ref | -1.37*** (-2.13,-0.59) | -1.58*** (-2.94,-0.22) | 0.36 (-2.70,3.41) | ༜0.001 |
Note: (1) The statistics reported are the marginal effects of independent variables with 95% confidence interval in parenthesis. (2) Model 1 was the crude unadjusted model; Model 2 adjusted for age, working status, marriage status, individual income, education level; Model 3 additionally adjusted for physical activity, drinking status; model 4 further adjusted for dietary energy intake and percentage of energy intake from fat. (3) BMI: Body mass index, WC: waist circumstance. (4) ***, **and*indicates statistical significance at the 1% 5% and 10% level, respectively. |
Table 3 Odds ratio of general obesity and abdominal obesity across smoking intensities among Chinese male and female, CHNS 1991–2015 1−4 |
Variables | | Male | | Female |
| Nonsmoker (n = 14,572) | Light (n = 8,706) | Moderate (n = 10,452) | Heavy (n = 2,177) | P trend | | Nonsmoker (n = 38,235) | Light (n = 799) | Moderate (n = 369) | Heavy (n = 38) | P trend |
| |
General obesity | | | | | | | | | | |
Model 1 | | 1.00 Ref | 0.76*** (0.69, 0.85) | 0.72*** (0.65,0.80) | 0.96 (0.81,1.15) | ༜0.001 | | 1.00 Ref | 1.31 (1.03, 1.66) | 1.61 (1.19, 2.19) | 1.26 (0.48,3.28) | 0.755 |
Model 2 | | 1.00 Ref | 0.76*** (0.69, 0.85) | 0.70*** (0.63,0.78) | 0.93 (0.78,1.11) | ༜0.001 | | 1.00 Ref | 0.87 (0.69, 1.11) | 1.03 (0.75,1.40) | 0.92 (0.35,2.44) | 0.710 |
Model 3 | | 1.00 Ref | 0.73*** (0.65, 0.82) | 0.69*** (0.61,0.76) | 0.91 (0.75,1.09) | ༜0.001 | | 1.00 Ref | 0.87 (0.68, 1.11) | 1.02 (0.75,1.40) | 0.91 (0.34,2.43) | 0.655 |
Model 4 | | 1.00 Ref | 0.72*** (0.65, 0.81) | 0.68*** (0.61,0.76) | 0.90 (0.75,1.09) | ༜0.001 | | 1.00 Ref | 0.87 (0.68,1.12) | 1.01 (0.74,1.39) | 0.93 (0.35,2.46) | 0.663 |
| | | | | | | | | | | | |
Abdominal obesity | | | | | | | | | | |
Model 1 | | 1.00 Ref | 0.84*** (0.79,0.89) | 0.85*** (0.80,0.90) | 1.03 (0.93,1.14) | ༜0.001 | | 1.00 Ref | 1.39** (1.19,1.62) | 1.22 (0.98,1.52) | 1.71 (0.87,3.36) | ༜0.001 |
Model 2 | | 1.00 Ref | 0.84*** (0.79,0.90) | 0.84*** (0.79,0.89) | 0.98 (0.88,1.09) | ༜0.001 | | 1.00 Ref | 0.83** (0.71,0.98) | 0.73*** (0.58, 0.91) | 1.21 (0.60,2.44) | ༜0.001 |
Model 3 | | 1.00 Ref | 0.81*** (0.76, 0.86) | 0.82*** (0.77,0.87) | 0.96 (0.86,1.07) | ༜0.001 | | 1.00 Ref | 0.84** (0.72, 0.99) | 0.74*** (0.59, 0.92) | 1.23 (0.61,2.48) | ༜0.001 |
Model 4 | | 1.00 Ref | 0.80*** (0.76,0.86) | 0.81*** (0.76,0.87) | 0.95 (0.86,1.07) | ༜0.001 | | 1.00 Ref | 0.83** (0.71,0.98) | 0.73*** (0.58, 0.91) | 1.24 (0.62,2.48) | ༜0.001 |
Note: (1) Date comes from CHNS longitudinal data, from 1991 to 2015. (2) The statistics reported are the odds ratio with 95% confidence interval in parenthesis. (3) Model 1 was the crude unadjusted model; Model 2 adjusted for age, working status, marriage status, individual income, education level; Model 3 additionally adjusted for physical activity, drinking status; model 4 further adjusted for dietary energy intake and percentage of energy intake from fat. (4) ***, **and*indicates statistical significance at the 1% 5% and 10% level, respectively. |