Physical Activity Guidelines and Long-term Weight Gain
Physical Activity Guidelines and Long-term Weight Gain
Baseline characteristics of men and women by level of physical activity are presented in Table 1. Age, weight, BMI, smoking, alcohol consumption, hypertension and occupational activity varied across physical activity groups at baseline, where active participants had favourable cardiovascular profile compared with inactive participants. The mean overall increase in weight during 22 years of follow-up was 7.7 kg (CI 7.4 to 8.1) for men and 8.5 kg (CI 8.2 to 8.8) for women. The weight gain was largest between the first and second wave, with 5.3 kg (CI 5.0 to 5.6) increase for men and 6.1 kg (CI 5.8 to 6.4) increase for women.
In multiadjusted analyses for age, smoking, alcohol and diabetes status, we found a clear dose–response association (p<0.001 for trend) between changes in body weight according to physical activity levels after allowing physical activity to be updated over an 11-year period in men and women. Compared with inactive men, above recommended level of physical activity was associated with 2.1 kg (CI 1.8 to 2.4) less weight gain, whereas recommended level of physical activity was associated with 0.7 kg (CI 0.5 to 0.9) less weight gain (p for trend <0.001). The corresponding estimates of lower weight gain among women were 1.8 kg (CI 1.5 to 2.2) and 0.5 kg (CI 0.3 to 0.7) for above recommended and recommended levels, respectively (p for trend <0.001).
We assessed modifying effects of age, BMI, smoking status and occupational activity on the association between physical activity and weight change, and found significant interactions with age and BMI in men and women. However, occupational activity and smoking modified the association only in men, and we found no evidence of interaction among women (Table 2). In stratified analyses of BMI, there was a clear dose–response relationship between physical activity and weight gain only for overweight/obese men (BMI>25 kg/m) and obese women (BMI>25 kg/m), with a p for trend <0.001 for both. We observed less weight gain associated with above recommended levels, that is, >150 min of moderate-intensity activity or >60 min of vigorous activity each week, in overweight/obese (BMI>25 kg/m; p<0.001) men and obese (BMI>30 kg/m) women (Table 2). Obese men and women also had significantly lower weight gain with recommended level of physical activity, that is, 150 min moderate-intensity or 60 min of vigorous activity weekly, compared with the reference category of inactive participants (Table 2).
We also examined ORs of gaining ≥2.3 kg (meaningful weight gain) associated with physical activity updated over time. Compared with inactive participants, the ORs of meaningful weight gain in the above recommended category were 0.79 (CI 0.69 to 0.91) and 0.70 (CI 0.60 to 0.81) in men and women, respectively (Table 3). There was a significant interaction with age for the odds of meaningful weight gain across levels of physical activity for both sexes, with lower ORs of gaining meaningful weight when physically active in older participants. In men >40 years, there was significantly reduced OR of meaningful weight gain with above recommended levels of physical activity. This was also true for women 40–59 years, whereas women >60 years had reduced OR of meaningful weight gain with recommended and above recommended levels of physical activity. BMI modified the association in women (p values for interaction <0.001), and only normal weight (BMI 18.5–24.9 kg/m) and overweight (BMI 25.0–29.9 kg/m) women were less likely to gain meaningful weight associated with high levels of physical activity (Table 3). In men, all but the lowest BMI category (<18.5 kg/m) had lower ORs for meaningful weight gain associated with above recommended levels of physical activity.
In subgroup analyses of men and women who reported to maintain their physical activity over the years, we observed that men who maintained above recommended levels of physical activity had increased 5.6 kg (CI 3.4 to 7.9) in body weight, while those who remained inactive had increased 9.1 kg (CI 7.9 to 10.3). Recommended levels of physical activity were associated with 7.2 kg (CI 5.5 to 8.8) increase in men. For women, maintaining above recommended levels of physical activity was associated with an increase of 3.8 kg (CI −0.2 to 7.8), compared with inactive women who increased 9.4 kg (CI 8.1 to 10.8). Recommended levels of physical activity in women were associated with an increase of 8.6 kg (CI 6.9 to 10.3; figure 2).
(Enlarge Image)
Figure 2.
Weight at Nord-Trøndelag County Health Study (HUNT) 1, HUNT 2 and HUNT 3 in men and women who maintained the same level of physical activity during the whole follow-up period.
In separate sensitivity analyses, the results were not substantially different from the main analyses. When we adjusted for body weight at baseline, we found that the results were similar to the results obtained without this adjustment (see online supplementary table S1 http://bjsm.bmj.com/content/48/20/1489/suppl/DC1). In a cross sectional analysis of body weight and physical activity from H3, we also adjusted for sitting time in addition to the variables that were already included in the multivariable models. We observed similar dose–response inverse associations between body weight and levels of physical activity after additional adjustments of sitting time (see online supplementary table S2 http://bjsm.bmj.com/content/48/20/1489/suppl/DC1). Among men, the body weight was 4.4 kg (CI 3.4 to 5.3) lower in those who reported above recommended levels of physical activity, compared with inactive men. Similarly, among women who reported above recommended levels of physical activity, body weight was 6.7 kg (CI 5.7 to 7.7) lower compared with women who were inactive (see online supplementary table S2 http://bjsm.bmj.com/content/48/20/1489/suppl/DC1).
Results
Baseline characteristics of men and women by level of physical activity are presented in Table 1. Age, weight, BMI, smoking, alcohol consumption, hypertension and occupational activity varied across physical activity groups at baseline, where active participants had favourable cardiovascular profile compared with inactive participants. The mean overall increase in weight during 22 years of follow-up was 7.7 kg (CI 7.4 to 8.1) for men and 8.5 kg (CI 8.2 to 8.8) for women. The weight gain was largest between the first and second wave, with 5.3 kg (CI 5.0 to 5.6) increase for men and 6.1 kg (CI 5.8 to 6.4) increase for women.
In multiadjusted analyses for age, smoking, alcohol and diabetes status, we found a clear dose–response association (p<0.001 for trend) between changes in body weight according to physical activity levels after allowing physical activity to be updated over an 11-year period in men and women. Compared with inactive men, above recommended level of physical activity was associated with 2.1 kg (CI 1.8 to 2.4) less weight gain, whereas recommended level of physical activity was associated with 0.7 kg (CI 0.5 to 0.9) less weight gain (p for trend <0.001). The corresponding estimates of lower weight gain among women were 1.8 kg (CI 1.5 to 2.2) and 0.5 kg (CI 0.3 to 0.7) for above recommended and recommended levels, respectively (p for trend <0.001).
We assessed modifying effects of age, BMI, smoking status and occupational activity on the association between physical activity and weight change, and found significant interactions with age and BMI in men and women. However, occupational activity and smoking modified the association only in men, and we found no evidence of interaction among women (Table 2). In stratified analyses of BMI, there was a clear dose–response relationship between physical activity and weight gain only for overweight/obese men (BMI>25 kg/m) and obese women (BMI>25 kg/m), with a p for trend <0.001 for both. We observed less weight gain associated with above recommended levels, that is, >150 min of moderate-intensity activity or >60 min of vigorous activity each week, in overweight/obese (BMI>25 kg/m; p<0.001) men and obese (BMI>30 kg/m) women (Table 2). Obese men and women also had significantly lower weight gain with recommended level of physical activity, that is, 150 min moderate-intensity or 60 min of vigorous activity weekly, compared with the reference category of inactive participants (Table 2).
We also examined ORs of gaining ≥2.3 kg (meaningful weight gain) associated with physical activity updated over time. Compared with inactive participants, the ORs of meaningful weight gain in the above recommended category were 0.79 (CI 0.69 to 0.91) and 0.70 (CI 0.60 to 0.81) in men and women, respectively (Table 3). There was a significant interaction with age for the odds of meaningful weight gain across levels of physical activity for both sexes, with lower ORs of gaining meaningful weight when physically active in older participants. In men >40 years, there was significantly reduced OR of meaningful weight gain with above recommended levels of physical activity. This was also true for women 40–59 years, whereas women >60 years had reduced OR of meaningful weight gain with recommended and above recommended levels of physical activity. BMI modified the association in women (p values for interaction <0.001), and only normal weight (BMI 18.5–24.9 kg/m) and overweight (BMI 25.0–29.9 kg/m) women were less likely to gain meaningful weight associated with high levels of physical activity (Table 3). In men, all but the lowest BMI category (<18.5 kg/m) had lower ORs for meaningful weight gain associated with above recommended levels of physical activity.
In subgroup analyses of men and women who reported to maintain their physical activity over the years, we observed that men who maintained above recommended levels of physical activity had increased 5.6 kg (CI 3.4 to 7.9) in body weight, while those who remained inactive had increased 9.1 kg (CI 7.9 to 10.3). Recommended levels of physical activity were associated with 7.2 kg (CI 5.5 to 8.8) increase in men. For women, maintaining above recommended levels of physical activity was associated with an increase of 3.8 kg (CI −0.2 to 7.8), compared with inactive women who increased 9.4 kg (CI 8.1 to 10.8). Recommended levels of physical activity in women were associated with an increase of 8.6 kg (CI 6.9 to 10.3; figure 2).
(Enlarge Image)
Figure 2.
Weight at Nord-Trøndelag County Health Study (HUNT) 1, HUNT 2 and HUNT 3 in men and women who maintained the same level of physical activity during the whole follow-up period.
In separate sensitivity analyses, the results were not substantially different from the main analyses. When we adjusted for body weight at baseline, we found that the results were similar to the results obtained without this adjustment (see online supplementary table S1 http://bjsm.bmj.com/content/48/20/1489/suppl/DC1). In a cross sectional analysis of body weight and physical activity from H3, we also adjusted for sitting time in addition to the variables that were already included in the multivariable models. We observed similar dose–response inverse associations between body weight and levels of physical activity after additional adjustments of sitting time (see online supplementary table S2 http://bjsm.bmj.com/content/48/20/1489/suppl/DC1). Among men, the body weight was 4.4 kg (CI 3.4 to 5.3) lower in those who reported above recommended levels of physical activity, compared with inactive men. Similarly, among women who reported above recommended levels of physical activity, body weight was 6.7 kg (CI 5.7 to 7.7) lower compared with women who were inactive (see online supplementary table S2 http://bjsm.bmj.com/content/48/20/1489/suppl/DC1).
Source...