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

Study design and population

National Health and Nutrition Examination Survey (NHANES) comprises complex-survey samples of a nationally representative population of noninstitutionalized U.S. civilians. NHANES study was a large prospective cohort in which participants were linked to mortality outcomes [18]. The data in NHANES were collected in each 2-year cycle since 1999. A total of 25,131 participants 18 of age or older who were eligible for MMA determination were included, because serum MMA was only measured in participants recruited in 5 study cycles (1999–2000, 2001–2002, 2003–2004, 2011–2012, and 2013–2014). After excluding participants who were pregnant (n = 886) or had missing covariates of diabetes status (n = 6), possible type 1 diabetes (aged less than 20 years with only insulin for antidiabetic treatment, n = 45)[19], or missing data for mortality status (n = 30), 24,164 adults were included for analysis.

We categorized diabetes status into four groups: previously diagnosed type 2 diabetes (i.e., diagnosis by doctors and/or the use of antidiabetic agents, n = 2535), undiagnosed diabetes (i.e., no previous diagnosis, no use of antidiabetic medication, and plasma HbA1c levels of ≥ 6.5% and/or fasting plasma glucose (FPG) of ≥ 126 mg/dL, n = 776), prediabetes (i.e., no previous diagnosis, no use of antidiabetic medication, and plasma HbA1c levels in the ranges of 5.7–6.5% and/or FPG in the ranges of 100–126 mg/dL, n = 5406), no diabetes (i.e., no previous diagnosis, no use of antidiabetic medication, HbA1c < 5.7% and FPG < 100 mg/dL, n = 15,447).

Measurement of serum methylmalonic acid

As mentioned in our previous reports [18, 19], MMA levels were measured in venous blood using gas chromatography-mass spectrophotometry (GC/MS) in participants during 1999–2004 and measured by liquid chromatography-mass spectrophotometry (LC-MS/MS) in participants during 2011–2014. High-performance GC/MS (Model 6890 GC system and Model 5973 mass selective detector, Hewlett-Packard, San Fernando, CA) was applied for MMA determination in NHANES 1999–2004. In brief, 275 µL specimens with internal standard solution supplemented by isotope-labelled methyl-d3-malonic acid (d3MMA) were extracted and subsequently derivatized with cyclohexanol to produce a dicyclohexyl ester. The chromatography was performed on a DB-5MS capillary GC column (0.25 mm x 30 m, 0.25-µm, J&W Scientific, Folsom, CA) within 15 min. The effluent part from the GC column was monitored with a mass-selective detector by the selected ion monitoring process. MMA levels were quantitated using the peak area ratios of MMA and the internal standard, isotope-labelled d3MMA. LC-MS/MS method was established for MMA detection in NHANES 2011–2014 using Thermo-Electron HPLC System and Thermo-Electron TSQ triple quadrupole mass spectrometer (Thermo Scientific, West Palm Beach, FL.). Compared to GC/MS procedures, LC-MS/MS method required less sample volume (75ul) and a shorter run time (6 min). MMA is extracted from 75 µL of sample along with an added internal standard (d3-MMA) via liquid-liquid extraction method with tert-butylmethylether. Then, the extracted organic acid is derivatized with butanol to form a dibutylester. The Hypersil Gold C18 column (2.1 mm × 50 mm, 1.9 μm particle size, Thermo Fisher Scientific) was selected because it allowed an excellent separation between MMA and succinic acid with relatively short chromatography run time and narrow analyte peaks. The butanol is evaporated under vacuum and the derivatized sample is reconstituted in acetonitrile/water (v/v 50/50). MMA was chromatographically separated from succinic acid using isocratic mobile phase (0.1% acetic acid: methanol 40:60 (v/v), 0.4 mL/min, 35 °C) within 6 min (retention time 3.47 min for SA and 4.25 min for MMA). Multiple reaction monitoring (MRM) was performed in positive electrospray ionization mode, with two transitions each for MMA (m/z 231 → 119 and 175.1) and for d3-MMA (m/z 234.1 → 122.1 and 178.1). Both methods showed excellent correlation and agreement [20]. MMA in the range of 50–2000 nmol/L has a favourable linear pattern with detectable signal intensity. Samples with MMA levels of more than 400 nmol/L were repeatedly analyzed to eliminate the detection error. The total coefficient of variation ranged from 4 to 10%, and the recovery rate was 94.0–96.0% in each study cycle. Detailed protocols and procedures have been described elsewhere [19].

Study outcomes

Participants in NHANES were linked to the National Death Index to identify vital status. The publicly used mortality data were recorded from baseline until death or December 31, 2015. The detailed protocols of the linkage and analysis guidelines for mortality data have been published by the National Center for Health Statistics [18]. Mortality attributed to heart disease was identified according to the International Classification of Diseases, 10th Revision (ICD-10), including I00-I09, I11, I13, and I20-I51.

Other variables

General characteristics at baseline, including age, sex, race/ethnicity (non-Hispanic White, non-Hispanic Black, Hispanic, or others), and smoking status (never, former, current) were reported by participants using standardized questionnaires [19]. Adequate physical activity was defined as at least 150–300 min/week of moderate-intensity, 75–150/week min of vigorous physical activity, or equivalent combination of moderate and vigorous physical activity. Body mass index (BMI) was calculated as weight (in kilograms) divided by height (in meters squared). Systolic/diastolic blood pressure was calculated as the mean of three eligible values. Hypertension was defined as the use of antihypertensive therapy or average blood pressure ≥ 140/90 mmHg. Chronic kidney disease (CKD) was defined as an estimated glomerular filtration rate < 60 mL/min/1.73 m2 using the Chronic Kidney Disease Epidemiology Collaboration formula. Cardiovascular disease (CVD) included self-reported coronary heart disease, heart failure, or stroke. Cancer was defined as self-reported cancer or a malignant tumour diagnosed by a physician. Data concerning the duration of diabetes, antidiabetic medications, and peripheral complications in persons with diabetes (e.g., foot ulcer/sore, meroparesthesia, and retinopathy) were extracted from diabetes questionnaires. Menopause was defined as amenorrhoea longer than 12 months attributed to menopause/hysterectomy.

All assessments of biospecimens were performed in the central laboratory. Total cholesterol (TC), high-density lipoprotein cholesterol (HDL-C), HbA1c, plasma glucose, and serum creatinine levels were measured according to validated laboratory methods during each study cycle. C-peptide (n = 6,726), insulin (n = 11,606), serum total testosterone (TT, n = 11,232), estradiol (n = 6,309), and sex hormone-binding globulin (SHBG, n = 5,797) levels were only measured in a subset. Free testosterone levels were estimated using the empirical free testosterone formula, as previously reported [21]. Serum cobalamin (vitamin B12) levels were measured using the commercial radioassay kit (Bio-Rad Laboratories, 1993) from 1999 to 2004 and using the automated electrochemiluminescence immunoassay (Roche, Elecsys 170) from 2011 to 2014. Both assays were calibrated to correct for changes across years according to NHANES recommendations [19].

Statistical analysis

According to the analytical guidelines of the NHANES study, clustering, strata, and sample weights were recorded to account for the complex sampling design and nationally representative estimates, unless otherwise noted [18]. Variables are expressed as weighted means (standard error, SE) and percentages. At the baseline, sex-specific characteristics were described by diabetes status. The generalized linear model was used to estimate the means of blood metabolic indicators after adjustment for age, sex, race/ethnicity, BMI, smoking, hypertension, and eGFR. The correlation coefficients between serum sex steroid hormones (total testosterone, free testosterone, estradiol, and SHBG) and MMA were estimated using the Spearman method. The correlation analysis between MMA and sex hormone profiles was repeated after adjustment for menopause status in females.

Sex-specific mortality rates across diabetes groups were estimated based on the Poisson distribution method and presented as events per 1000 person-years of follow-up. Cox proportional hazard regression analysis was used to estimate hazard ratios (HRs) and 95% confidence intervals (CIs) for the mortality risk per doubling of serum MMA (each unit increase of log2 transformed values of MMA), which were stratified by diabetes status and sex. The proportional hazards assumption was tested and fulfilled. In the multivariate analyses, model 1 was adjusted for age, and model 2 was further adjusted for race/ethnicity, smoking status, physical activity, BMI, hypertension, cancer, CVD, the ratio of TC/HDL-C, eGFR, plasma HbA1c, and cobalamin. The significance of the interaction between MMA levels and sex for mortality risk was assessed via the weighted Wald test.

Several sensitivity analyses were conducted. First, to examine the dose-response relationship between serum MMA concentrations and mortality risk by diabetes status and sex, a restricted cubic spline with four knots (5th, 35th, 65th, and 95th ) was used based on the unweighted Cox regression model with multivariate adjustment as mentioned above. Second, the analysis for adults with diagnosed diabetes was additionally adjusted for diabetes-related variables, including the duration of diabetes, diabetic complications (including diabetic limb ulcer/sore, retinopathy, and peripheral neuropathy), UACR (mg/g), metformin use, ACEI/ARB, and blood-lipid lowering drugs. We further reanalyzed participants with eligible sex hormone profiles to assess whether sex hormones modified the relationship between MMA and mortality risk in males and females, and menopausal status was considered for female participants. All tests with a 2-sided p value less than 0.05 were considered statistically significant using Stata (version 12).

留言 (0)

沒有登入
gif