The association between obesity, cardiometabolic disease biomarkers, and innate immunity-related inflammation in Canadian adults

Figures & data

Table 1 Population characteristics

Characteristic	n1	% or mean (lclm, uclm)2
Sex3
Male	861	49.2 (48.8, 49.5)
Female	944	50.8 (50.5, 51.2)
Age (years)3	1805	44.3 (44.0, 44.6)
Age group3
18–35	506	32.5 (30.5, 34.4)
>35–65	980	56.3 (54.1, 58.5)
>65	319	11.3 (10.0, 12.6)
Ethnicity3
Caucasian	1535	84.3 (71.8, 91.9)
Asian4	144	11.4 (6.1, 20.1)
Other5	83
Physical activity index3
Active	392	19.9 (16.1, 24.3)
Moderate	464	24.7 (21.1, 28.8)
Inactive	949	55.4 (48.5, 62.2)
Smoking activity3
Ever	892	46.9 (43.2, 50.5)
Never	908	53.1 (49.5, 56.8)
Waist circumference (cm)
Males	861	94.6 (93.2, 95.9)
Females	930	86.7 (84.7, 88.8)
BMI (kg/m^2)	1795	26.9 (26.4, 27.4)
Systolic blood pressure (mmHg)	1804	111.5 (110.2, 112.7)
Diastolic blood pressure (mmHg)	1804	70.9 (70.1, 71.8)
Taking cardiac medication3	420	18.3 (15.9, 21.0)
Has diabetes3	86	4.0 (3.3, 4.9)
Metabolic markers
Apolipoprotein A1 (g/L)	1805	1.4 (1.4, 1.5)
Apolipoprotein B (g/L)	1801	0.9 (0.9, 0.9)
LDL-C (mmol/L)	1798	3.0 (3.0, 3.1)
HDL-C (mmol/L)	1805	1.3 (1.3, 1.4)
Total cholesterol (mmol/L)	1805	4.9 (4.8, 5.0)
Total:HDL-C ratio	1805	3.9 (3.8, 4.0)
Triglycerides (mmol/L)	1803	1.3 (1.3, 1.4)
HbA1c (%)	1760	5.6 (5.5, 5.7)
Inflammatory markers
C-reactive protein (mg/L)	1669	2.3 (2.1, 2.5)
Fibrinogen (g/L)	1752	0.0297 (0.0283–0.0309)
Homocysteine (μmol/L)	1793	7.8 (7.5, 8.2)
Other
Vitamin D (nmol/L)	1784	67.4 (65.1, 69.8)

Notes:

1 Unweighted frequency;

2 weighted measure;

3 self-reported data;

4 coefficient of variation = 27% denoting a marginal estimate with a high sampling variability;

5 coefficient of variation = 34% denoting an unacceptable estimate. Ethnicityspecific estimates for the Other group are not presented as per Statistics Canada recommendations.

Abbreviations: lclm, lower confidence limit; uclm, upper confidence limit; LDL-C, low-density lipoprotein cholesterol; HDL-C, high-density lipoprotein cholesterol; Total:HDL-C ratio, total cholesterol/HDL ratio; HbA_1c, glycosylated hemoglobin.

Figure 1 Relationship between metabolic markers and BMI by BMI group.

Notes: Unweighted and untransformed values for biomarkers and BMI were plotted on a single graph with green open circles (o) representing subjects who fall in the normal weight range, orange plus signs (+) representing those that fall in the overweight range, and red exes (x) representing those who fall in the obese range. The top 1% of biomarker values for HbA_1c were removed from the graph to exclude potential outliers. Unadjusted linear regression lines within each BMI group were plotted to highlight trends in the data (shown in blue). Gray dotted lines are shown vertically on the graph to represent the cut-points between normal weight, overweight, and obese. Regression coefficients (β), standard errors, and P-values were calculated using weighted linear regression adjusted for age, sex, smoking status, physical activity, and ethnicity. BMI, triglycerides, and HbA_1c were log-transformed to improve normality in linear regression models.
Abbreviations: BMI, body mass index; HbA_1c, glycosylated hemoglobin; LDL-C, low-density lipoprotein cholesterol; HDL-C, high-density lipoprotein cholesterol; Total:HDL-C ratio, total cholesterol/HDL ratio; SE, standard error.

Figure 2 Relationship between BMI and biomarkers of inflammation and plasma vitamin D by BMI groups.

Notes: Unweighted and untransformed values for biomarkers and BMI were plotted on a single graph with green open circles (o) representing subjects who fall in the normal weight range, orange plus signs (+) representing those that fall in the overweight range, and red exes (x) representing those who fall in the obese range. The top 1% of biomarker values for 25(OH)D were removed from the graphs to exclude potential outliers. Unadjusted linear regression lines within each BMI group were plotted to highlight trends in the data (shown in blue). Gray dotted lines are shown vertically on the graph to represent the cut-points between normal weight, overweight, and obese. Regression coefficients (β), standard errors, and P-values were calculated using weighted linear regression adjusted for age, sex, smoking status, physical activity, and ethnicity. BMI, CRP, fibrinogen, homocysteine, and 25(OH)D were log-transformed to improve normality in linear regression models.
Abbreviations: BMI, body mass index; 25(OH)D, 25-hydroxyvitamin D; CRP, C-reactive protein; SE, standard error.

Table 2 Adjusted mean levels of biomarkers by BMI groups

Biomarker	Normal weight			Overweight			Obese			Overall population
	n2	Mean3	SE3	n2	Mean3	SE3	n2	Mean3	SE3	β ± SE	P Trend 4
Apolipoprotein A1 (g/L)	646	1.49	0.02	663	1.40	0.02	415	1.34	0.02	−0.31 ± 0.05	<0.0001
Apolipoprotein B (g/L)	643	0.87	0.02	663	0.92	0.03	414	1.00	0.03	0.29 ± 0.04	<0.0001
LDL-C (mmol/L)	643	2.82	0.10	663	3.01	0.10	411	3.33	0.13	1.18 ± 0.17	<0.0001
HDL-C (mmol/L)	646	1.48	0.04	663	1.30	0.02	415	1.19	0.03	−0.61 ± 0.06	<0.0001
Total cholesterol (mmol/L)	646	4.86	0.10	663	4.92	0.09	415	5.10	0.11	0.64 ± 0.14	0.0009
Total:HDL-C ratio	646	3.47	0.10	663	4.01	0.09	415	4.51	0.14	2.29 ± 0.21	<0.0001
Triglycerides (mmol/L)1	645	0.96	1.04	662	1.19	1.05	415	1.56	1.05	1.02 ± 0.06	<0.0001
HbA1c (%)1	634	5.66	1.01	644	5.65	1.02	403	5.95	1.02	0.11 ± 0.01	<0.0001
C-reactive protein (mg/L)1	567	0.65	1.15	640	1.26	1.16	392	2.28	1.16	2.80 ± 0.19	<0.0001
Fibrinogen (g/L)1	623	0.03	1.02	650	0.03	1.02	404	0.03	1.02	0.32 ± 0.03	<0.0001
Homocysteine (μmol/L)1	645	7.41	1.05	659	7.39	1.05	410	7.99	1.05	0.12 ± 0.05	0.045
25(OH)D (nmol/L)1	644	55.366	1.04	657	56.426	1.03	410	47.656	1.04	−0.25 ± 0.06	<0.0001

Notes:

1 Biomarker was log-transformed for linear regression analyses producing P-value for trend results;

2 unweighted frequency for the number of subjects used in each analysis to produces mean, SE, and P-values;

3 adjusted (age, sex, smoking status, physical activity, ethnicity and season of clinic visit, for vitamin D only), untransformed, weighted means and SE;

4 P-value for the relationship between log-transformed BMI and biomarker adjusted for age, sex, smoking status, physical activity, and ethnicity (and season of clinic visit, for vitamin D only);

5 P-value for linear regression coefficient no longer significant after adjustment for cardiac drug use (P = 0.1) or when excluding subjects with diabetes (P = 0.6);

6 vitamin D means only adjusted for age, sex, physical activity, ethnicity, and season of clinic visit in order to maintain 11 degrees of freedom in the model.

Abbreviations: BMI, body mass index; LDL-C, low-density lipoprotein cholesterol; HDL-C, high-density lipoprotein cholesterol; Total:HDL-C ratio, total cholesterol/HDL ratio; HbA_1c, glycosylated hemoglobin; 25(OH)D, 25-hydroxyvitamin D; SE, standard error.