Skip to main content

Table 3 Results of structural equation models with change in HbA1c as outcome

From: Associations between local descriptive norms for overweight/obesity and insufficient fruit intake, individual-level diet, and 10-year change in body mass index and glycosylated haemoglobin in an Australian cohort

LDN: Overweight/obesity N  = 1908 Unadjusted models Adjusted models a
Fruit serves (log transformed) Estimate 95% CI P value Estimate 95% CI P value
 ΔHbA1c on LDN (overweight/obesity) 0.008 0.003 to 0.012 0.001 0.009 0.004 to 0.013 0.000
 ΔHbA1c on fruit intake(log) −0.004 − 0.010 to 0.002 0.241 − 0.004 − 0.012 to 0.003 0.259
 Fruit intake(log) on LDN (overweight/obesity) − 0.025 − 0.047 to − 0.002 0.030 − 0.025 − 0.047 to − 0.002 0.030
 Indirect effect *100 0.009 − 0.009 to 0.027 0.316 0.009 − 0.010 to 0.028 0.347
 Total effect *100 0.795 0.352 to 1.238 0.000 0.881 0.423 to 1.340 0.000
 Model fit AIC 6485.203 BICadj 6518.479 AIC 6063.327 BICadj 6129.877
Vegetable serves (log transformed)
 ΔHbA1c on LDN (overweight/obesity) 0.008 0.003 to 0.012 0.001 0.009 0.004 to 0.013 0.000
 ΔHbA1c on vegetable intake(log) −0.003 − 0.010 to 0.005 0.456 − 0.001 − 0.009 to 0.007 0.727
 Vegetable intake(log) on LDN (overweight/obesity) −0.012 −0.034 to 0.009 0.266 −0.012 − 0.034 to 0.009 0.262
 Indirect effect *100 0.003 −0.007 to 0.014 0.525 0.002 −0.009 to 0.012 0.737
 Total effect *100 0.793 0.346 to 1.239 0.001 0.873 0.414 to 1.332 0.000
 Model fit AIC 6595.252 BICadj 6628.527 AIC 6176.544 BICadj 6243.095
Sugary drink intake (0 v > 0)
 ΔHbA1c on LDN (overweight/obesity) 0.008 0.003 to 0.012 0.001 0.009 0.004 to 0.013 0.000
 ΔHbA1c on sugary drink intake 0.001 −0.006 to 0.008 0.731 −0.001 −0.008 to 0.006 0.857
 Sugary drink intake on LDN (overweight/obesity) 0.028 0.004 to 0.052 0.025 0.028 0.004 to 0.053 0.023
 Indirect effect *100 0.003 −0.015 to 0.021 0.725 −0.002 −0.022 to 0.019 0.860
 Total effect *100 0.791 0.346 to 1.237 0.000 0.871 0.414 to 1.329 0.000
 Model fit AIC 7118.570 BICadj 7151.845 AIC 6715.416 BICadj 6781.966
LDN: Insufficient fruit intake N  = 1966 Unadjusted models Adjusted models a
Fruit serves (log transformed) Estimate 95% CI P value Estimate 95% CI P value
 ΔHbA1c on LDN (insufficient fruit intake) 0.006 0.002 to 0.011 0.005 0.007 0.002 to 0.011 0.005
 ΔHbA1c on fruit intake(log) −0.004 − 0.011 to 0.002 0.145 − 0.004 − 0.011 to 0.002 0.188
 Fruit intake(log)on LDN (insufficient fruit intake) −0.029 − 0.052 to − 0.007 0.010 −0.030 − 0.052 to − 0.007 0.010
 Indirect effect *100 0.013 −0.008 to 0.034 0.218 0.013 −0.009 to 0.035 0.246
 Total effect *100 0.650 0.203 to 1.097 0.004 0.668 0.214 to 1.123 0.004
 Model fit AIC 6703.708 BICadj 6737.402 AIC 6258.888 BICadj 6326.276
Vegetable serves (log transformed)
 ΔHbA1c on LDN (insufficient fruit intake) 0.006 0.002 to 0.011 0.005 0.007 0.002 to 0.011 0.005
 ΔHbA1c on vegetable intake(log) −0.005 − 0.012 to 0.003 0.228 − 0.003 − 0.012 to 0.005 0.392
 Vegetable intake(log) on LDN (insufficient fruit intake) −0.006 − 0.034 to 0.021 0.658 − 0.006 − 0.034 to 0.021 0.652
 Indirect effect *100 0.003 −0.0110to 0.015 0.654 0.002 −0.008 to 0.012 0.659
 Total effect *100 0.650 0.201 to 1.098 0.005 0.664 0.209 to 1.119 0.004
 Model fit AIC 6836.997 BICadj 6870.692 AIC 6395.732 BICadj 6463.120
Sugary drink intake (0 v > 0)
 ΔHbA1c on LDN (insufficient fruit intake) 0.006 0.002 to 0.0101 0.005 0.007 0.002 to 0.011 0.004
 ΔHbA1c on sugary drink intake 0.002 −0.004 to 0.009 0.473 0.001 −0.006 to 0.008 0.861
 Sugary drink intake on LDN (insufficient fruit intake) 0.025 0.002 to 0.047 0.033 0.025 0.002 to 0.048 0.032
 Indirect effect *100 0.006 −0.011 to 0.022 0.481 0.002 −0.016 to 0.019 0.860
 Total effect *100 0.647 0.199 to 1.096 0.005 0.658 0.201 to 1.114 0.005
 Model fit AIC 7365.307 BICadj 7399.001 AIC 6940.353 BICadj 7007.741
  1. Numbers in bold are all statistically significant with p values
  2. aAdjusted for area-level income and individual-level age, sex, employment status, education, marital status, and smoking status; Abbreviations: AIC Akaiki’s Information Criterion, BIC adj sample size adjusted Bayesian Information Criterion, CI confidence interval, HbA 1c glycosylated haemoglobin, LDN local descriptive norm