Sexual dimorphism in mitochondrial dysfunction and diabetes mellitus: evidence from a population-based cohort study

Table 3 Sex-specific all-cause mortality risk associated with serum MMA according to diabetes status

		Deaths/ person-years ^a	Mortality rate ^b (95%CI)	Rate Difference	Age-Adjusted HR (95%CI)	p value	Multivariable-adjusted HR (95%CI)	p value	p-interaction
No diabetes
	Females	726/73,992	7.32 (6.70–8.02)	0.62 (0.35–0.90)	1.38 (1.24–1.54)	< 0.001	1.29 (1.13–1.47)	< 0.001	0.782
	Males	862/69,463	7.94 (7.28–8.68)	0.62 (0.35–0.90)	1.39 (1.27–1.53)	< 0.001	1.34 (1.19–1.52)	< 0.001	0.782
Prediabetes
	Females	382/18,222	17.57 (15.56–19.90)	1.44 (0.65–2.21)	1.51 (1.28–1.79)	< 0.001	1.49 (1.26–1.78)	< 0.001	0.108
	Males	528/20,060	19.01 (17.01–21.28)	1.44 (0.65–2.21)	1.39 (1.21–1.61)	< 0.001	1.30 (1.08–1.58)	0.007	0.108
Undiagnosed diabetes
	Females	64/2,518	19.66 (14.52–27.02)	5.78 (3.61–7.95)	2.33 (1.69–3.21)	< 0.001	1.50 (1.01–2.23)	0.045	0.186
	Males	102/3,148	25.44 (19.77–33.08)	5.78 (3.61–7.95)	1.84 (1.45–2.34)	< 0.001	1.66 (1.12–2.51)	0.015	0.186
Diagnosed diabetes
	Females	309/8,720	32.38 (28.11–37.43)	9.77 (8.33–11.21)	1.26 (1.07–1.48)	0.007	1.19 (1.04–1.37)	0.014	0.028
	Males	402/8,484	42.15 (37.14–47.97)	9.77 (8.33–11.21)	1.61 (1.42–1.83)	< 0.001	1.58 (1.36–1.86)	< 0.001	0.028

^a unweighted, ^b weighted death rate per 1000 person-year. HR, hazard ratio; CI, confidence interval. Multivariable-adjusted model was adjusted for age (continuous), race/ethnicity (non-Hispanic white, non-Hispanic black, Hispanic-Mexican and other), smoking status (never, ever, and current), Adequate physical activity(no/yes), BMI (normal, overweight, and obesity), hypertension (no/yes), cancer (no/yes), CVD (no/yes), TC/HDL-C ratio (continuous), eGFR (continuous), HbA1c (continuous), and serum cobalamin (continuous). Mortality rate was calculated according to the Poisson distribution. HR (95%CI) and p-value were calculated by weighted Cox proportional hazards regression. The interaction between MMA and sex was assessed by survey-weighted Wald test

ISSN: 1758-5996