| Summary: | 博士 === 國立臺灣大學 === 流行病學與預防醫學研究所 === 99 === Background: Several previous studies have investigated the association between androgenetic alopecia (AGA) and factors related to metabolic syndrome (MetS), which is known to increase the risk of type 2 diabetes mellitus and cardiovascular disease. However, the results of these studies have been inconsistent and most of them are based on prevalent survey rather than repeated surveys that preclude one from elucidating multi-step progression of AGA and also pinpointing whether the role of MetS plays in onset or progressive stage of the natural history of AGA. Furthermore, for a community-based study on these issues, the misclassifications of AGA status are usually an important concern. Therefore, studies on AGA progression and its association with MetS after considering misclassifications with an appropriate statistical method are worthy of being investigated.
Objective: (1) To elucidate if there is an association between MetS and AGA after adjustment for potential confounders. (2) To estimate the progression rates of AGA and investigate its association with MetS with or without considering misclassification of AGA status.
Methods: Population-based cross-sectional surveys were conducted in a Taiwanese community. Norwood and Ludwig classifications were used to assess the degree of hair loss in men and women. Information on components of MetS along with other possible risk factors was collected. (1) A total of 740 men aged 40 to 91 years participated in the survey between April and June 2005. The data were used to analyze the association between AGA and MetS. A logistic regression model was employed to assess the associations between MetS or each possible risk factor and the risk of moderate or severe AGA. (2) A total of 4,633 women and 2,362 men aged 30 to 95 years participated in the survey in 2005 and a total of 25,118 women and 16,884 men aged 30 to 102 years participated in the survey in 2010. A total of 899 women and 584 men aged 36 to 94 years participated in both surveys and they are used to estimate the incidence rates of AGA. Then, all of these data were utilized in estimation of AGA progression and its association with MetS. A multi-step Markov model was utilized for analysis. A Bayesian approach with Markov Chain Monte Carlo (MCMC) method for estimation of these parameters with correction for misclassifications was also used.
Results: (1) A statistically significant association was found between AGA and the presence of the MetS (Odds ratio (OR) = 1.67, 95% CI: 1.01, 2.74) as well as between AGA and the number of fulfilled MetS components (OR= 1.21, 95% CI: 1.03, 1.42) after controlling for age, family history of AGA, and smoking status. Among MetS components, high-density lipoprotein (HDL) (OR= 2.36, 95% CI: 1.41, 3.95, p= 0.001) was revealed as the most important factor associated with AGA. (2) After 5-year follow-up, 89/745 (12.0%) women and 58/369 (15.7%) men developed AGA which leads to the incidence rates of 2.4% and 3.1% per person-year in women and men, respectively. In AGA progression, MetS was significantly associated with progression of AGA in the second-step transition (Hazard ratio (HR) =1.16, 95%CI: 1.04, 1.30) but lacking of statistically significant association with the first-step transition (HR=1.03, 95%CI: 0.98, 1.08) after adjusting for age, sex, and family history. Some individual components of MetS were found significantly associated with progression of AGA. These factors included lower serum HDL level in the first-step transition from normal to mild or moderate AGA (HR=1.07, 95%CI: 1.01, 1.13), fasting glucose or DM in both transitions (HR=1.06, 95%CI: 1.01, 1.11; HR=1.20, 95%CI: 1.08, 1.33), and hypertension in both transitions (HR=1.04, 95%CI: 1.01, 1.08; HR=1.16, 95%CI: 1.06, 1.26). These estimates are consistent after considering misclassifications which revealed estimates away from the null and larger standard errors.
Conclusions: A statistically significant association was found between AGA and the presence of the MetS as well as between AGA and the number of fulfilled MetS components. With regard to progression of AGA, MetS was associated with the transition rate from mild or moderate to severe state of AGA which was revealed in both with or without considerations of misclassifications of AGA status. These results demonstrated a significant association between MetS and AGA. Identification of the MetS in moderate or severe AGA patients might be necessary for early recognition that would lead to early intervention to reduce the risk or complications of cardiovascular disease and type 2 diabetes mellitus later in life. In this study, we could also evaluate the extents of misclassifications for AGA and obtain the estimates for associations between MetS and AGA progressions after correcting for these measurements errors. These estimates could be used for predictions of transition probabilities which are useful for risk stratifications in population-based intervention studies on AGA.
|
|---|
