The Association Between Obesity,Metabolic Abnormality And Thyroid Function

Part Ⅰ Association between different obesity phenotypes and hypothyroidism:a study based on a longitudinal health management cohortBackground:Hypothyroidism refers to the common pathological condition of thyroid hormone deficiency and includes overt hypothyroidism and subclinical hypothyroidism.In China,the prevalence of overt hypothyroidism is 1.02%,and the prevalence of subclinical hypothyroidism increased from 3.21%in 1999 to 12.93%in 2017.Many studies have found that subclinical hypothyroidism is associated with an increased risk of cardiovascular disease events and mortality.These facts highlight the need to identify the risk factors for hypothyroidism to prevent the increase in its incidence.Obesity is also a global health problem,with growing incidence worldwide and serious adverse health effects.In recent years,the relationship between obesity and thyroid dysfunction has drawn increased attention,and it has been suggested that obesity might not be only the result but also the cause of thyroid dysfunction.A meta-analysis including 22 studies showed that the obese population had increased risks of both overt hypothyroidism and subclinical hypothyroidism.Obesity is typically associated with a constellation of metabolic abnormalities,such as dyslipidemia,hypertension,and hyperglycemia.A prospective cohort study found that participants with metabolic syndrome at baseline had an increased risk of developing subclinical hypothyroidism.Although the deleterious metabolic effects of obesity are well recognized,metabolic responses to obesity differ at the individual level.Approximately 10%to 30%of obese individuals are metabolically healthy.Additionally,there is a subgroup of individuals with abnormal metabolic parameters who are not obese.According to their obesity and metabolic statuses,individuals were classified into four different obesity phenotypes:metabolically healthy and nonobese(MHNO),metabolically healthy obese(MHO),metabolically unhealthy nonobese(MUNO),and metabolically unhealthy obese(MUO).Combining obesity with different metabolic profiles,obesity phenotypes are better predictors of cardiovascular disease and mortality than obesity per se.Different obesity phenotypes may also help us to understand whether obesity per se or coexisting metabolic abnormalities can increase the risk of hypothyroidism.Only a few previous studies have investigated whether thyroid function could identify obesity phenotypes in euthyroid subjects.However,no cohort studies have investigated the relationship between different obesity phenotypes and the development of hypothyroidism.In addition,since thyroid function and disorders often present differently in different sex and age groups,whether this relationship is modified by sex and age is also worth investigating.This study may provide new perspectives for the relationship between obesity,metabolic abnormalities and hypothyroidism,and may be helpful for identifying high-risk subjects and disease prevention.Objectives:1.To compare the incidence density of hypothyroidism in groups with different obesity phenotypes.2.To explore whether obesity phenotype is an independent risk factor for hypothyroidism.3.To explore the sex and age differences in the association between obesity phenotype and hypothyroidism.Methods:1.Study populationThe study population was derived from the health management cohort composed of individuals who underwent a comprehensive health examination in Shandong Provincial Hospital.After strict inclusion and exclusion criteria,a total of 6081 males and 2930 females who were euthyroid at baseline were included in our final analysis.The median follow-up time was 1.92(1.00-2,17)years.2.Data CollectionThe data were collected from the health management database in Shandong Provincial Hospital.Information on medical history and lifestyle was gathered by a standardized questionnaire.After an overnight fast of at least 10 hours,each subject underwent a standardized medical examination,including anthropometric and laboratory tests.3.DefinitionsObesity was defined as body mass index(BMI)≥ 25 kg/m2.Metabolically unhealthy was defined as having at least two of the following four metabolic syndrome components:(1)fasting plasma glucose(FPG)≥6.1 mmol/L;(2)systolic blood pressure(SBP)≥130 mmHg or diastolic blood pressure(DBP)≥85 mmHg;(3)triglyceride(TG)≥1.7 mmol/L;(4)high-density lipoprotein cholesterol(HDL-C)<1.0 mmol/L.According to their obesity and metabolic statuses,individuals were classified into four different obesity phenotypes:MHNO,MHO,MUNO,and MUO.MHNO was regarded as healthy phenotype in this study,while the other three phenotypes were collectively referred to as unhealthy phenotypes.Euthyroidism was defined as serum thyroid-stimulating hormone(TSH),free thyroxine(FT4),and free triiodothyronine(FT3)levels within the reference ranges.Hypothyroidism was defined as TSH>4.2 μIU/mL and FT4<12 pmol/L(overt hypothyroidism)or TSH>4.2μIU/mL and FT4 levels within the reference range(subclinical hypothyroidism).4.Statistical analysisContinuous variables were presented as the mean±standard deviation or median[25th percentile,75th percentile],Categorical variables were presented as numbers(percentages).Differences in continuous variables were compared by using one-way ANOVA or Kruskal-Wallis test.Bonferroni correction was applied to all multiple comparisons.For categorical variables,the chi-square test or Fisher’s exact test was used for group comparisons.The incidence density of hypothyroidism was calculated using the number of incident cases of hypothyroidism and person-years of follow-up in each group.A Mid-P exact test was applied for to calculate 95%confidence interval(C1)and compare incidence densities between groups.The generalized estimation equation(GEE)method was used to investigate the associations between different metabolic obesity phenotypes and the development of hypothyroidism,which is an extension of the generalized linear model to allow for analysis of repeated measurements or other correlated observations.Sensitivity analyses were conducted by excluding partipants with positive thyroperoxidase antibodies(TPOAb)at baseline.The odds ratio(OR)and 95%CI for different obesity phenotypes in association with hypothyroidism were reported.A two-tailedp-value<0.05 was considered statistically significant.All statistical analyses were performed in RStudio(version 1.0.143)with R(version 4.0.0).Results:1.Characteristics of different obesity phenotypes by sex at baseline and during follow-upBased on obesity phenotype at baseline,in both males and females,the MHNO group was the largest group of the four obesity phenotype groups(MHNO:72.49%vs.32.91%,p<0.001).The proportions of MHO,MUNO,and MUO were higher in males than in females(MHO:32.71%vs.15.56%,MUNO:9.06%vs.4.78%,MUO:25.32%vs.7.17%,all p<0.001).Among 6081 males,4087(67.21%)had consistent obesity phenotypes,and 1994(32.79%)had changes in obesity phenotypes during follow-up.Among the 2930 women,2452(83.69%)had consistent obesity phenotypes,and 478(16.31%)had changes in obesity phenotypes during follow-up.2.Baseline characteristics of males and females based on obesity phenotypesThe average age of males was 50.53±14.11 years.The average age of females was 46.56±14.24 years.Based on obesity phenotypes at baseline,individuals with the MUNO and MUO phenotypes had significantly higher TG,SBP,DBP,and FPG,while individuals with the MHO and MUO phenotypes had higher BMI in both males and females(all p<0.05).Baseline serum thyroid function(TSH levels and FT4 levels)were comparable between the four groups in both males and females.3.The non-MHNO group had a higher incidence density of hypothyroidism than the MHNO groupThe incidence densities of hypothyroidism in males with consistent obesity phenotypes during follow-up were 12.19(8.62-16.76),15.87(11.39-21.56),14.52(6.74-27.57)and 19.88(14.06-27.34)per 1000 person-years in the MHNO,MHO,MUNO,and MUO groups,respectively.Based on the changes in obesity phenotypes during follow-up,the incidence densities of hypothyroidism were 22.35(13.86-34.26),9.67(4.71-17.74),and 17.91(14.93-21.32)per 1,000 person-years in the MHNO to non-MHNO(MHO,MUNO or MUO),non-MHNO to MHNO and non-MHNO to non-MHNO groups,respectively.Compared with the MHNO group,the MUO,MHNO to non-MHNO,and non-MHNO to non-MHNO groups had significantly higher incidence densities(all p<0.05).In contrast to males,although a trend exists,there was no significant difference between the non-MHNO groups and the MHNO group in females.In summary,males with unhealthy phenotypes had a significantly higher incidence density of hypothyroidism than the MHNO group.4.MHO,MUNO,and MUO phenotypes were independent risk factors for developing hypothyroidism compared with the MHNO phenotype in malesThe results of GEE analyses showed that in males,the MHO,MUNO,and MUO phenotypes were all risk factors for developing hypothyroidism compared with the MHNO phenotype.In the multivariable-adjusted model,participants with the MHO phenotype had a 1.53-fold increased odds of developing hypothyroidism(OR=1.53,95%CI=1.05-2.21,p=0.026),while participants with the MUNO and MUO phenotypes had 1.81-fold(OR=1.81,95%CI=1.08-3.02,p=0.023)and 1.85-fold(OR=1.85,95%CI=1.24-2.77,p=0.003)increased odds of developing hypothyroidism,respectively.Contrasting the findings among male participants,non-MHNO phenotypes were not independent risk factors for the development of hypothyroidism in female participants.Sensitivity analyses showed that findings remained unchanged.In summary,the MHO,MUNO,and MUO phenotypes were independent risk factors for the development of hypothyroidism compared with the MHNO phenotype in males,while such an association was not found in females.5.Age modifies the relationship between obesity phenotypes and hypothyroidism in malesAmong younger males,the MHO and MUO phenotypes were associated with a higher risk of developing hypothyroidism than the MHNO phenotype.After adjusting for potential confounding factors,participants with the MHO and MUO phenotypes had 1.69-fold(OR=1.69,95%CI=1.03-2.76,p=0.036)and 1.95-fold(OR=1.95,95%CI=1.14-3.35,p=0.015)increased odds of developing hypothyroidism,respectively.In contrast to younger males,only the MUNO phenotype was an independent risk factor of developing hypothyroidism among older males(OR=2.28,95%CI=1.16-4.47,p=0.017).In addition,older males with the MUO phenotype also showed a trend toward an increased risk of developing hypothyroidism(OR=1.77,95%CI=0.96-3.25,p=0.067).We also performed subgroup analyses in females according to age.Nevertheless,non-MHNO phenotypes were not independent risk factors for developing hypothyroidism in either younger or older females.Sensitivity analyses showed that findings remained unchanged.Thus,age modifies the relationship between obesity phenotypes and hypothyroidism in males.Conclusion:1.Males with unhealthy phenotypes had a significantly higher incidence density of hypothyroidism,and females with unhealthy phenotypes showed a trend of increased incidence density of hypothyroidism.2.There are sex differences in the association between obesity phenotypes and the development of hypothyroidism.Unhealthy phenotypes were independent risk factors for the development of hypothyroidism in males,while no association between obesity phenotypes and hypothyroidism was observed among females.3.There are age differences in the association between obesity phenotypes and the development of hypothyroidism.In younger males,excessive adiposity per se was an independent risk factor for hypothyroidism,while in older males,metabolic abnormalities were associated with an increased risk of developing hypothyroidism.4.Both obesity and metabolic abnormalities were associated with an increased risk of hypothyroidism in males.Part Ⅱ Statin Use and Benefits of Thyroid Function:A Retrospective Cohort StudyBackground:Hypothyroidism is a common pathological condition of thyroid hormone deficiency,including overt hypothyroidism(OH)and subclinical hypothyroidism(SCH).In China,the prevalence of SCH has significantly increased from 3.21%in 1999 to 12.93%in 2017.The most frequent cause of SCH is Hashimoto’s thyroiditis in iodine-sufficient areas.However,risk factors contributing to the increasing prevalence of SCH remain unclear.In part one,we have explored the relationship between different obesity phenotypes and hypothyroidism,and found that both obesity and metabolic abnormalities were associated with an increased risk of hypothyroidism.In recent years,studies have revealed the emerging role of the disturbance of lipid metabolism in the development of hypothyroidism.Our previous prospective observational study found that high baseline total cholesterol(TC)level was a risk factor of progression to OH in patients with SCH,which suggested that cholesterol may influence thyroid function.It is known that SCH is associated with an increased risk of cardiovascular disease,and cholesterol is a key element in the development of cardiovascular disease.If cholesterol-lowering therapy can benefit thyroid function,we can not only find a possible way to relieve the disease burden of SCH,but also provide additional evidence that cardiovascular mortality and morbidity can be reduced by the proper control of cholesterol levels.Statins are widely used due to their ability to lower cholesterol in clinical practice.Besides,statins also have pleiotropic actions such as anti-inflammatory and immunomodulatory properties.Only a few studies have investigated the effects of statins on thyroid function,and the results were inconsistent.A reason for these findings could be due to small sample sizes that were limited to hospital-based patients only.It still remains inconclusive whether statin use is associated with improved thyroid function in the general population.We believe that it is worth clarifying this relationship,as well as investigating whether this is mediated by the cholesterol-lowering function of statins.This study could provide additional evidence for the etiological research and clinical practice of hypothyroidism.Objectives:1.To assess the association between statin use and thyroid function2.To explore the role of the cholesterol-lowering effect in the association between statin use and thyroid functionMaterials and Methods:1.Study design and participantsThis study involves retrospective analyses of the population derived from the community-based REACTION study,which was a prospective observational cohort study in China investigating the epidemiology of metabolic diseases in residents aged 40 years or older.In this study,data were obtained from participants who enrolled in Ningyang County,Shandong Province between April 2011 and July 2017.We included 5146 participants who had more than one visit during the study period and assessed for eligibility.After exclusion,201 participants who had statin therapy during the follow-up period were defined as the statin group.Considering non-random treatment allocation and potential confounding covariates,we used baseline sex-,age-,TC-and thyroid function-matched participants without lipid-lowering therapy as the control group(1:1 match).Ultimately,201 participants in the statin group and 201 participants in the control group were included in the final analysis.The median follow-up time of the study population was 2.72(1.22)years.2.Data collectionTrained investigators obtained information on demographic characteristics,medical history(including statin use),and other essential information from a well-established questionnaire through a face-to-face interview.Blood samples were collected between 0800 h and 1000 h after at least 10-hour fasting.The intraassay and interassay coefficients of variation were always below 5%for all of the above parameters.Above data were collected at baseline and each follow-up.3.Main exposure and study outcomesThe exposure was statin use during follow-up.The primary outcome was thyroid status at the end of the follow-up.Euthyroidism was defined as serum thyroid-stimulating hormone(TSH),free thyroxine(FT4),and free triiodothyronine(FT3)levels within the reference ranges.SCH was defined as TSH>4.2 μIU/mL and FT4 levels within a reference range.OH was defined as TSH>4.2 μIU/mL and FT4 levels<12 pmol/L.The secondary outcome was serum TSH levels at the end of follow-up.4.Statistical analysisContinuous variables were expressed as means±standard deviations or medians(interquartile ranges).Categorical variables were summarized as numbers(percentage).Differences of continuous variables between different groups were compared by using Student’s t-test or the Mann-Whitney test.Differences of categorical variables were tested by using the chi-square test.For the comparison of measurements at baseline and follow-up,the paired t-test and Wilcoxon matched-pair signed-rank test were employed.We used logistic regression models to assess the relationship between statin use and thyroid status.We further used linear regression models to examine the relationship between statin use and log-transformed serum TSH levels at the end of follow-up.The potential confounders that may affect thyroid function were adjusted in both the multivariate logistic regression models and the linear models.In addition,a mediation analysis was performed to examine whether the association of statin use with TSH levels was mediated by TC changes.All statistical analyses were performed using SPSS version 24.0 for Windows.Results:1.Baseline characteristics of subjects in the statin group and the control groupAge,sex,TC,FT4,TSH,and thyroid antibodies were well matched between the two groups.Low-density lipoprotein cholesterol(LDL-C),fasting plasma glucose(FPG),alanine aminotransferase(ALT),aspartate aminotransferase(AST),and follow-up time were comparable between the statin group and the control group.FT3,body mass index(BMI),systolic blood pressure(SBP),and diastolic blood pressure(DBP)were lower in the control group,and estimated glomerular filtration rate(eGFR)was lower in the statin group(all p<0.05).There were more people accompanied with cardiovascular disease,hypertension,and diabetes mellitus in the statin group compared with the control group(all p<0.05).2.Changes in TC levels and thyroid function in the statin group and the control groupIn the control group,serum TC levels increased from 5.73 ± 1.23 mmol/L to 6.03 ±1.13 mmol/L(p<0.001),while in the statin group,serum TC levels reduced from 5.83± 1.55 mmol/L to 4.95 ± 1.05 mmol/L(p<0.001).At the end of follow-up,serum TC levels were significantly lower in the statin group than those in the control group(p<0.001),and the percentage of normal thyroid function in the statin group was higher than that in the control group(88.06%vs.76.12%,p=0.002).Accordingly,the serum TSH levels at the end of follow-up were significantly lower in the statin group compared with the control group(2.32 μIU/mL vs.2.61 μIU/mL,p=0.007).The subgroup analyses showed similar results.Subjects in the statin group had better outcomes of thyroid function than the control group in both euthyroid and SCH subjects.3.Statin use was independently associated with normal thyroid functionTo evaluate the relationship between statin use and the outcomes of thyroid function and if it was independent of confounding factors,we utilized the logistic regression models.Subjects with statin use had higher odds of normal thyroid function at the end of follow-up compared with the control group(OR=2.314,95%CI=1.354-3.953,p=0.002).The results were similar after adjusting for the confounders(OR=2.510,95%CI=1.380-4.564,p=0.003).The results of subgroup analyses suggested that statin use was independently associated with a better chance of normal thyroid function at the end of follow-up in both euthyroid and SCH participants at baseline.4.Statin use was independently associated with a decrease in TSH levels at the end of follow-upAs TSH is a more sensitive marker of thyroid dysfunction,we further performed linear regression analysis to examine the relationship between statin use and serum TSH levels at the end of follow-up.After adjustment for confounders,statin use was independently and negatively associated with the log-transformed TSH levels at the end of follow-up[standardized coefficients(Beta)=-0.099,p=0.010].Similar results were achieved in the subgroup analysis of euthyroid subjects at baseline(Beta=-0.092,p=0.035).These results demonstrated that statin use was associated with a decrease in TSH levels,even in euthyroid subjects.5.TC changes serve as a mediator of the association between statin use and TSH levelsTo explore if TC changes mediated the relationship between statin use and TSH levels at the end of follow-up,we performed a mediation analysis.After adjustment for confounders,the total effect of statin use on log-transformed TSH levels at the end of follow-up was significant(βTol=-0.061,p=0.002).After adjustment for TC changes,the direct effect of statin use on log-transformed TSH levels at the end of follow-up was no longer significant(p=0.201),and showed a significant indirect effect(βInd=-0.038,p=0.001).These results indicated a mediation effect of TC changes in the relationship between statin use and TSH levels at the end of follow-up.A similar mediation effect was also found in euthyroid subjects at baseline.Conclusion:1.Statin use was independently associated with better outcomes of thyroid function.2.Statins may benefit thyroid function through their cholesterol-lowering effects.