Table 5 Relative risk of hypothyroidism (including clinical and subclinical hypothyroidism) and subclinical hypothyroidism-only according to quartiles of serum magnesium determined using multiple logistic regression analyses.

	Model 1^a		Model 2^b		Model 3^c		Model 4^d
	OR (95% CI)	p	OR (95% CI)	p	OR (95% CI)	p	OR (95% CI)	p
Hypothyroidism
Serum magnesium		0.000		0.000		0.000		0.000
≤0.55	4.482 (2.438–8.239)	0.000	4.544 (2.468–8.369)	0.000	4.841 (2.584–9.070)	0.000	4.785 (2.582–8.866)	0.000
0.551–0.85	0.982 (0.627–1.535)	0.935	0.999 (0.636–1.570)	0.997	1.024 (0.647–1.621)	0.918	0.995 (0.634–1.559)	0.981
0.851–1.15	1.00		1.00		1.00		1.00
>1.15	0.954 (0.561–1.623)	0.863	0.916 (0.536–1.565)	0.747	0.847 (0.491–1.461)	0.551	0.973 (0.572–1.657)	0.920
Subclinical hypothyroidism
Serum magnesium		0.000		0.000		0.000		0.000
≤0.55	4.517 (2.382–8.567)	0.000	4.531 (2.382–8.617)	0.000	4.971 (2.557–9.666)	0.000	4.654 (2.435–8.896)	0.000
0.551–0.85	0.885 (0.540–1.451)	0.629	0.919 (0.558–1.512)	0.739	0.951 (0.572–1.580)	0.846	0.890 (0.543–1.461)	0.646
0.851–1.15	1.00		1.00		1.00		1.00
>1.15	1.066 (0.608–1.872)	0.823	1.038 (0.589–1.831)	0.897	0.903 (0.505–1.613)	0.730	1.087 (0.619–1.910)	0.772

^aModel 1: adjusted for age, sex, anti-thyroid peroxidase antibody, anti-thyroid globulin antibody, and serum iodine concentration.
^bModel 2: additionally adjusted for smoking status, and body mass index.
^cModel 3: adjusted for all covariates in model 2 as well as income and education.
^dModel 4: adjusted for all covariates in model 1, but age was used as a classification variable according to youth, middle age, and old age (as shown in Table 1).
Regression analyses using the forward method, backward method, and all arguments simultaneously were performed and the results were similar.
OR: odds ratio; CI: confidence interval.

Quick links

Search