[Comparison of health-care coverage between children with congenital hypothyroidism and healthy children in Iran]

Background and Aim: provision of timely healthcare for children under 5 years of age specially children with congenital hypothyroidism and also regular monitoring of growth through measuring weight, height and head circumference have been of great importance. The aim of this study was to make a comparison between healthy children and children with congenital hypothyroidism in regard to health-care coverage in Iran

Materials and Methods: this study was conducted in five provinces in different geographical areas of Iran in 2015. 240 children with congenital hypothyroidism and 240 healthy children were entered into the study. Children with congenital hypothyroidism were selected randomly from the national registere database. The children in the control group were matched with the children with hypothyroidism in regard to age, gender and place of residence. Anthropometric indices including weight, height and head circumference of the children of less than 3 years of age were recorded as the health care indices. STATA software version 13 was used for data analysis

Results: the mean percentage of measurement and record of weight, height and head circumference in children below 3 years of age with congenital hypothyroidism were 82.1, 78.7 and 55.4%, respectively. The mean percentage of measurement and record of weight, height and head circumference in healthy children [control group] below 3 years of age were 82.2%, 80.9 and 55.1%, respectively. There was a statistically significant difference between the two groups in relation to the percentage of height measurement [P =0.02]

Conclusion: according to the results of this study, percentage of measurement and record of the anthropometric indices in Iranian children of less than 3 years of age is low. More attention to the measurement and record of these physical growth indices in children of less than 3 years of age specially children with congenital hypothyroidism who need regular follow up, is essential

[Evaluation of factors associated with time to sputum smear conversion in smear-positive pulmonary TB patients in Golestan province, Iran [2009-14]]

Background and Objective: Tuberculosis caused by Mycobacterium tuberculosis is one of the deadliest infectious diseases in the world. The amount of the bacilli in tuberculosis will reduce rapidly by starting effective antibiotic treatment and the remained bacilli in the sputum will be an important scale to respond to treatment. This study was conducted to evaluate the examin factors associated with the conversion of positive smear to negative one in tubercular-pulmonary patients in Golestan province, in northern Iran

Methods: This retrospective cohort study was carried out on 2093 patients with smear positive pulmonary tuberculosis registered in Golestan province, northern Iran from March 2009-14 referred to health centers. The outcome of this study was to determine the time of changing to negative of the first smear during the treatment and evaluating its relationship with demographic variables, the density of bacilli in the smear, culture, chest x-ray, diabetes and HIV

Results: 67.5% of patients among 2093 smear-positive pulmonary tuberculosis patients were conversion rate of smear at the end of the second months. Results of Cox model showed relationship between gender, age, weight, density bacillus smear and culture results at the start of treatment and negative smears [P<0.05]. Old age, low weight, high density of bacillus in primary smear in the beginning of treatment and increasing of the number of colonies in culture of positive smear patients were considered as predictor factors in changing positive smear to negative one

Conclusion: This study revealed that Conversion rate and treatment success in Golestan province was less than what was expected

[Relationship between socio-economic factors and tuberculosis using negative binomial and Poisson regression models]

Tuberculosis [TB] is the most important cause of death worldwide. The main reason for the increasing global burden of TB are severe poverty and class distinctions between rich and poor population groups in various communities. This study was performed to determine the relationship between socio-economic factors and TB using negative binomial and Poisson regression models. This descriptive - analytic study was conducted on 11320 TB affected patients in Iran during 2010. Data was gathered from the Iranian Ministry of Health and Medical Education. The relationship between the numbers of cases with socio-economic indicators was determined using negative binomial and Poisson models. Fitting models were compared using AIC [Akaike Information Criterion] and BIC [Bayesian Information Criterion]. The Poisson regression model showed a significant relationship between the TB mortality rate and socio-economic factors [P<0.05]. Negative binomial regression model showed a significant relationship between TB and unemployment, illiterate, immigration and urban residency [P<0.05]. Negative binomial regression model showed no relationship between TB and family size, physicians' ratio to the number of population centers and annual average income. There is a significant impact of socio-economic factors with the number of TB cases. Negative binomial regression model is suitable for accountable data in comparision with Poisson regression model

Incidence of acute myocardial infarction in Islamic Republic of Iran: a study using national registry data in 2012

Population-based data on myocardial infarction rates in the Islamic Republic of Iran have not been reported on a national or provincial scale. In a cross-sectional study, data were collected on 20 750 new cases of myocardial infarction [ICD10 codes I21-22] admitted to hospitals and registered by the Iranian Myocardial Infarction Registry in 2012. The crude and age-adjusted incidence for the 31 provinces and the whole country were directly calculated per 100 000 people using the WHO standard population. Overall, males comprised 72.4% of cases and had a significantly lower mean age at incidence than women [59.6 [SD 13.3] years versus 65.4 [SD 12.6] years]. The male: female incidence ratio was 2.63. The age-standardized myocardial infarction incidence rate was 73.3 per 100 000 in the whole country [95% CI: 72.3%-74.3%] and varied significantly from 24.5 to 152.5 per 100 000 across the 31 provinces. The study provides baseline data for monitoring and managing cardiovascular diseases in the country

[Interaction of APOC3 polymorphism and dietary fats on the risk of metabolic syndrome]

The aim of this study was to evaluate the interaction between dietary fatty acids and the genetic variant of APOC3 rs5128 3238C>G in relation to metabolic syndrome [MetS] components in adults. In this matched nested case-control study, 755 MetS subjects and 755 controls were selected from among participants of the Tehran Lipid and Glucose Study. Dietary intake was determined using a valid and reliable food frequency questionnaire. APOC3 was genotyped by the conventional polymerase chain reaction and restriction fragment length polymorphism. Mean ages of men and women were not different in cases and controls. The frequency of C allele was 81%, which did not differ in cases and controls or in men and women. Compared to CC genotype, low HDL-C risk was increased in women with the CG+GG genotypes and with cholesterol intakes >/=208 mg/day [OR: 1.93]. In men with the CG+GG genotypes and saturated fatty acid [SFA] intakes >/=9.8% of energy, OR of high diastolic blood pressure [BP] was 2.15[1-1.46], compared to individuals with SFA intake <9.8% of energy and CC genotype. Compared to the CC genotype, the risk of high diastolic BP was higher in men carrying the G allele and consuming mono-unsaturated fatty acid [MUFA] intakes >/=9.4% of energy. Results demonstrate a nutri-genetic interaction between rs5128 and fat intakes in relation to components of MetS; individuals with G allele carriers and higher intakes of cholesterol, MUFA or SFA had higher risk of low HDL-C and hypertension than the CC genotype

[Validity of vision screening in schoolchildren of Iran]

To determine the sensitivity, specificity, and predictive value of visual screening examination in students living in urban areas of Iran. The first grade students of 7 cities of Iran were randomly selected in a population based cross sectional study in 2013 using multistage cluster sampling. After obtaining consent from the students' parents, examinations including corrected and uncorrected visual acuity, cycloplegic and non-cycloplegic refraction, and cover test were performed for all students by an optometrist. The criterion for visual problem was a cut point of 20/32 or worse for uncorrected visual acuity. In addition, the validity of the uncorrected visual acuity measured in the national screening program, which is recorded in the students' health profiles, was calculated. Out of total 4157 individuals who were selected for the study, uncorrected visual acuity was measured in 3645 participants. The sensitivity and specificity of uncorrected visual acuity measured in the national screening program was 35.3% [95% CI 29.2%-41.7%] and 93.9% [95% CI 93.04%-94.7%], respectively. The lowest sensitivity was observed in Dezful [19.6%] and the highest was observed in Ardebil [83.3%]. The predictive value of the examinations was 28.8% [95% CI 23.6%-34.3%] and the best cut point of uncorrected visual acuity was recorded as 20/25 in the national screening program. The validity of the visual screening program which is performed before the first grade of primary school is very low. It should be noted that since many families trust these examinations, they do not follow up their children's condition any more based on the results of these examinations. Therefore, the validity of these examinations, especially their sensitivity, should increase through training the examiners or using more specialized methods

[ association between dairy intake with metabolic syndrome and its components in adolescents: Tehran lipid and glucose study]

Epidemiological studies show that consumption of dairy products plays an important role in prevention and treatment of the metabolic syndrome [MetS]. The objective of this study is to examine the relation between dairy intake with MetS and its components in Tehranian adolescents. In this cross-sectional study, 785 adolescents, aged 10 to 19 years, participants of the fourth phase of the Tehran Lipid and Glucose Study were selected. Usual dietary intake was assessed using a valid semi-quantitative food frequency questionnaire. Total dairy, low fat and high fat dairy, milk, yoghurt and cheese were and anthropometrics, biochemical measurements and blood pressure were assessed. MetS was defined according to the de Ferranti criteria. Mean age of subjects was 14.8 +/- 2.9 years. The prevalence of MetS was 22.2% [girls: 19.5% and boys: 25.2%]. The most prevalent component for MetS among boys was high waist circumference [53.4%] and among girls was low HDL-C [53.1%]. Energy density and intake of protein, total fat, saturated fat, cholesterol, calcium and phosphor were higher in the highest quartile of dairy consumption, than in the lowest quartile. After adjustment for confounders, odds ratios with 95% confidence interval for MetS in the highest quartiles of total dairy, low fat dairy, high fat dairy, milk, yoghurt and cheese were 0.97 [0.57-1.66], 1.44 [0.83-2.49], 0.97 [0.56-1.67], 0.70 [0.42-1.18], 1.62 [0.99-2.64], 0.72 [0.44-1.18] respectively, compared to the lowest quartile. Results of this study do not support the hypothesis that consumption of dairy products protects against MetS and its components

[Diet quality and its relationship with metabolic syndrome in adolescents: Tehran lipid and glucose study]

The "Dietary Guidelines for Americans Adherence Index [DGAI] " was developed based on the adherence to the dietary recommendations of the Dietary Guidelines for Americans 2005 [DGA-2005] to assess the contribution of dietary patterns to chronic diseases. The objective of this study was to evaluate the association of dietary patterns as measured by the modified DGAI with the prevalence of metabolic syndrome [MetS] and its components in Tehranian adolescents. In this cross-sectional study, 706 adolescents aged 10 to 19 years, participated from the fourth phase of the Tehran Lipid and Glucose Study. Usual dietary intake was assessed using a valid food frequency questionnaire, and the DGAI score was calculated for all participants. MetS was defined according to de Ferranti criteria. The most prevalent risk factor for MetS was high waist circumference [51.8%]. After adjusting for confounding variables, those in the highest quartile category of DGAI had a 36% lower risk of low HDL-C [odds ratio=64%, 95% CI: 0.41-0.98 and P for trend=0.03] and 50% lower risk of hypertension [odds ratio=50%, 95% CI: 0.25-0.99 and P for trend=0.02], compared with the lowest quartile. Increasing adherence to DGAI showed no significant difference in risk of hypertriglyceridemia, hyperglycemia, high waist circumference, MetS, mean body mass index [BMI] or physical activity scores. In adolescents, increasing adherence to Dietary Guidelines for Americans 2005 [DGA-2005] had only an inverse association with low HDL-C and hypertension, but had no relationship with the risk of MetS and its other components

[Application of quantile regression model in assessment of urine iodine related factors in Tehran population]

Quantile regression can be applied to model skewed variables, especially, when the objective is to model the tails of a response variable with highly skewed distribution. The aim of this study is to apply quantile regression to analyze urine iodine data and related factors in a Tehranian population. Data was collected in a cross-sectional study, in which 639 subjects, aged 19 years and over, were enrolled through randomized cluster sampling in Tehran between 2008-9. Due to the high skewness of 24 hr urinary iodine concentrations [UIC24] and to evaluate its extreme points, two linear quantile regression models were fitted. In model I, UIC24 was regressed on iodine content of salt and daily salt intake. These variables were replaced by iodine intake in model II, both models were adjusted by age. Model coefficients were estimated using the linear programming method and simplex algorithm. Significancy of the variables were evaluated by the bootstrap method. The Akaike information criterion [AIC] was used to assess the fitting of the models. All analyses were performed using R software version 2.12.2. Model I showed an increase in coefficients of iodine content of salt, daily salt intake, but a decrease in age coefficient in the length of the urinary iodine concentration percentiles. Model II showed similar results, but better fit [smaller AIC] in percentiles lower than median. Compared to ordinary regression, quantile regression models showed better fit, and a more complete picture and are recommended for modeling all parts of UIC24

[Studying the interaction effects of gene polymorphisms on low level of HDL over time using transition logic regression: Tehran lipid and glucose study]

As High-density lipoprotein [HDL] is directly associated with cardiovascular disease, the factors affecting the levels of this fat can be effective in reducing heart diseases. In addition to biochemical and environmental factors, genetic interactions also affect HDL level. Since polymorphism effects can be time-dependent, study of genetic interactions on HDL over time is important. In this study, we proposed Transition Logic Regression to analyze interactions in binary longitudinal data and used it to investigate polymorphism interactions related to low HDL over time. Data of 329 subjects who participated in three phases of TLGS was analyzed using the proposed model. Results showed that subjects with high triglyceride levels and increased waist circumference have an odds ratio of 2.29 [CI 95%: 1.51, 3.48] of having low HDL. Also, being in phase 2 and being a carrier of the minor allele of ApoA1M1 or being homozygous for the common allele of ApoCIII, were associated with an increased odds of having low HDL [OR= 2.30, CI 95%: 1.77, 2.99]. The odds ratio for having low HDL in male subjects with high blood pressure or being homozygous for the minor allele of SRB1 is 0.38 [CI 95%: 0.25,0.59]. Considering the identification of gene interactions in genetic studies and their importance over time, Transition Logic Regression was introduced and used to find gene interactions influencing low HDL over time and the most important models for gene interactions were identified

[Genetic association of some haplotypes to level of hdl-c, triglyceride, and waist in family members with metabolic syndrome using haplotype based association test]

Haploytpes are important elements in study of genetic associations. The haplotype based association test [HBAT] is a method to study genetic association of haplotypes with one or more traits. The test statistic in this method, which is calculated for all haplotypes, follows a standard normal distribution. In this study, in order to find the chromosomal area locus of genes affecting metabolic syndromes, the HBAT method was used to investigate the genetic association of haplotypes of some candidate microsatellites with HDL-C, triglycerides, and waist. A sum of 125 families with at least one member having metabolic syndrome according to ATPIII, and at least two members with low HDL-C levels were selected from among participants of the Tehran Lipid and Glucose Study [TLGS]. The genetic association of HDL-C, triglycerides, and waist with haplotypes of some microsatellites of chromosome 8, 11, 12, and 16 was studied, using HBAT. Data was obtained for 125 families, consisting of 563 individuals, aged 20 years or above [269 males and 294 females]. Genetic association of the haplotype 2-2-2-2 of chromosome 8 showed significant association with HDL-C and triglycerides. Haplotypes 2-2-1 and 2-2-2 of chromosome 12 showed significant association with triglycerides. In addition, haplotype 1-1-2 of this chromosome was found to be associated with waist [P<0.05]. Researching haplotypes provides more information on genetic associations, and identification of haplotypes influencing HDL-C level, triglycerides, and waist may be helpful in designing future research aimed at determining the genes predisposing persons to metabolic syndrome.

[Logic regression analysis for finding interaction effects of genes polymorphisms and other risk factors on low HDL: Tehran lipid and glucose study]

Logic regression is a generalized regression method that can identify complex Boolean interactions of binary variables. This method has been successfully used for analyzing single-nucleotide polymorphism data, because in SNP association studies interactions are important. The aim of this study is to investigate the associations between some candidate gene polymorphisms and HDL concentration using Logic Regression. Subjects for this cross sectional study, 436 subjects [172 men and 264 women] aged >/= 20 with some polymorphisms, were randomly selected from among participants of the Tehran Lipid and Glucose Study [TLGS]. Logic regression analysis was used to identify combinations of main genetic effects and interactions associated with HDL. Cross validation and randomization test were done to avoid over fitting of the models. Cross validation test suggested that the Logic model with four Boolean combinations and four predictors was the best logic model, which after fitting, showed that individuals who carry Apoe SNP Reversed Ze 3 or have high TG have an odds ratio of 2.35 [CI 95%:1.3-4.25] for having low HDL compared to other subjects. Also subjects with high TG have odds ratio 2.73 [CI 95%: 1.65,4.53] for having low HDL. Results of this study shows that Logic Regression is a powerful method to determine the interaction effect between high TG and ApoE SNP for having low HDL

[ effects of Iran-Iraq war circumstances on adolescent's anthropometric and biochemical indices at the age of 11-18. Tehran lipid and glucose study]

Research on children during war shows that in comparison to the whole population, these children are the most jeopardized group. The aim of this study was to investigate the effects of war circumstances on the anthropometric and biochemical indices of adolescent boys and girls who were born during the war-years. In the TLGS, a sub-sample of adolescent, aged between 11 to 18 years old, was selected and divided into two groups. The first adolescent group [war group] were born between 1981 and 1989, and the second adolescent group were born after war years. Height, weight, serum lipids, fast blood sugar [FBS], systolic and diastolic blood pressure and body mass index [BMI] were compared between the two studied groups. The mean weight and height increased at the age of 12, 13, 14 and 17 years old in boys who were born after war. The mean weight of girls in after war group, increased at the age of 11, 13, 14 years old. However, the differences in their height were not statistically significant [P>0.05]. Systolic and diastolic blood pressure decreased in both groups. Some anthropometric indices like height and weight were increased in the boys who were born after the war, but in the girls, the mean weight in three ages groups were increased. Systolic and diastolic blood pressure and some lipid profile decreased in the boys and the girls who were born after war

Application of FBAT-MM and fbat-lc multimarker methods in association of selected microsatellites with hdl-c in families with metabolic syndrome members: tehran lipid and glucose study

Studying several linked markers provides more information on locating disease genes locus by using genetic association analysis. The aims of this study were to introduce Multimarker Family Base Association Tests [FBAT-MM] and its Linear Combination [FBAT-LC] in multimarker genetic association analysis and to examine the association of selected microsatellites with HDL-C in an Iranian population. One hundred twenty five [1 25] families having at least one member with metabolic syndrome and at least two members with low HDL-C were selected from participants of the Tehran Lipid and Glucose Study [TLGS]. Multimarker genetic association of HDL-C level with some microsatellites in the chromosomes 8, 11, 12, and 1 6 were examined using FBAT-MM and FBAT-LC methods. The families consisted of 563 individuals [269 males and 294 females]. FBAT-MM showed significant genetic association only between HDL-C and three microsatellites in Chromosome 1 1 [P<0.05]. The microsatellite Dl 1 SI 304 was found as the significant factor for multimarker genetic association. FBAT-MM and FBAT-LC did not show shortcomings such as excessive conservatism and low power which are, usually, observed in other multimarker methods. Finding microsatellites associated with HDL-C level can provide background for further researches on the role of predisposing genes in metabolic syndrome

Direct estimate of population attributable fraction of risk factors for cardiovascular diseases: tehran glucose and lipid study

Cardiovascular disease is the first cause of death and third place in disease burden in Iran. This study estimâtes population attributable fraction [PAF] of Cardiovascular risk factors in Tehran population. PAF is one of the important parameters of measuring population affect of risk factors and evaluating potentiel impact of preventive strategies in community level. In this study 5868 participants above 30 years old of Tehran lipid and glucose study [TLGS] were employed and 501 CVD events detected during 10 years follow-up. Direct estimate of adjusted PAFs using logistic regression which is one of less biased exist methods of PAF calculation were applied. Highest modifiable Cardiovascular risk factor PAFs, in sequence, was smoking [14.16%], hypertension [11.73%], diabètes [7.32%] hypercholesterolemia [6.85%] and central obesity [5.91%] for men, and hypertension [19.25%], diabètes [18.82%], central obesity [9.88%] and hypercholesterolemia [7.95%], for women. Also PAF of hazardous age and premature family history of CVD, as most important nonmodifiable CVD risk factors, were 36.09%, 1 6.61% and 3.95%, 7.56% for men and women respectively. According to the difference of risk factors PAFs ranking in men and women, it is suitable that CVD preventive interventions to be prioritized by sex separately. In this regard, besides special attention to control tobacco use in men, hypertension and diabètes in both sexes and high cholesterol in men and central obesity in women respectively, should be given in priority of preventive strategies

[ impact of low frequency noise on mental performance during math calculations]

The influences of Low Frequency Noise [LFN] on mental performance have been the objectives of many papers in recent years. Many researchers believe that the LFN even at low pressure levels has impacts on mental performance. In the present work the influences of LFN and Reference Noise [RN] at two different sound pressure levels [45 and 65 dBA] on mental performances of college students during solving math problems were studied. Other factors such as annoyance and sensitivity to LFN and hearing status of the participants were also determined. Low Frequency and Reference Noise were generated and verified using cool Edit software program. The number of participants was calculated using data gathered from a pilot study conducted in advance. The mental performance of the students was evaluated while they were exposed to LFN and RN at 45 and 65 dBA sound pressure levels. For this purpose, standard psychological tests were applied. Standard measuring instrumentation including sound level meters, psychological tests, acoustic room and etc were calibrated prior to their application based on standard methods. The results revealed that, Low frequency noise annoyance and sensitivity have no significant differences between different studied ages and sexes. The result also revealed that, a significant correlations exists between sensitivity to low frequency noise and deep mental process [r=0.33, p=0.01]. It was recognized that low frequency nose at 65 dBA can decrease accuracy [p=0.005] and performance [p=0.001] in comparison with reference noise at the same level. Low frequency noise, have impact on mental performance during performing math calculations

[Artificial neural network design for modeling of mixed bivariate outcomes in medical research data]

Mixed outcomes arise when, in a multivariate model, response variables measured on different scales such as binary and continuous. Artificial neural networks [ANN] can be used for modeling in situations where classic models have restricted application when some of their assumptions are not met. In this paper, we propose a method based on ANNs for modeling mixed binary and continuous outcomes. Univariate and bivariate models were evaluated based on two different sets of simulated data. The scaled conjugate gradient [SCG] algorithm was used for optimization. To end the algorithm and finding optimum number of iteration and learning coefficient, mean squared error [MSE] was computed. Predictive accuracy rate criterion was employed for selection of appropriate model. We also used our model in medical data for joint prediction of metabolic syndrome [binary] and HOMA-IR [continues] in Tehran Lipid and Glucose Study [TLGS]. The codes were written in R 2.9.0 and MATLAB 7.6. The predictive accuracy for univariate and bivariate models based on simulated dataset I, where two outcomes associated with a common covariate, were shown to be approximately similar. However, in simulated dataset ?? in which two outcomes associated with different covariates, predictive accuracy in bivariate models were seen to be larger than that of univariate models. It is indicated that the predictive accuracy gain is higher in bivariate model, when the outcomes share a different set of covariates with higher level of correlation between the outcomes

[Comparison of glucose tolerance in patients with graves disease treated with methimazole with those who have received radioiodine]

One of the major concern of the administration of radioiodine is its complications. Pancras has sodium iodine symporter and may concentrate radioiodine. This study compared glucose tolerance in Graves disease patients on continuous treatment with antithyroid drugs, with those who received radioiodine. Materials and In this study, 132 patients with Graves' disease who relapsed after drug therapy were randomly selected on by the patient preference for long-term treatment with either methimazole or radioactive iodine. In each group, fasting blood sugar, glucose tolerance test, and lipid profiles, TSH, insulin, HOMA-IR and HOMA-B were measured. The two groups showed no significant difference in age sex, BMI and BP. Median FBS and HOMA-IR in the radioactive iodine group were higher than in the methimazole group [94 mg/dl vs. 90 mg/dl, P=0.019] and 1.5[1.2-2.3] vs. 1.3[0.8-2.1, P=0.045, respectively, but after adjustment there was no difference between the two groups; no significant difference was seen in HOMA-B, median two-hour blood glucose and serum insulin levels, between the two groups. Results indicate no relationship between radioactive iodine treatment and glucose intolerance

[Effect of long-term continuous methimazole treatment of thyrotoxicosis: comparison with radioiodine]

The major complication of radioiodine therapy for thyrotoxicosis is hypothyroidism, the long-term management of which is often problematic. In this study, the long-term effects of continuous methimazole [MMI] therapy are investigated. One hundred and thirty-two patients, aged between 36-66 years, with Graves recurrence, were semi randomly randomized in 2 groups for continuous antithyroid and radioiodine treatment. The number of thyroid dysfunctions in each patient were recorded and serum TSH, FT4, Anti TPO, TRAb, FBS, HOMA IR, and lipid profiles were measured. Bone mineral density and echocardiography were performed. There was no significant differences in age, duration of symptoms and thyroid function between the 2 groups. No serious complications occurred in the MMI group and no difference in any of parameters was seen between groups 1 and 2. Goiter rate and anti TPO concentration were higher in group 1 than in group 2. Serum triglycerides and cholesterol were increased in group 2 as compared to group 1. Bone mineral density was more reduced in group 2, especially in the spine. Echocardiography showed diastolic dysfunction in group 2 as compared to group 1. Some parameters in neuro-pschyciatric evaluation were significantly better in the MMI group, as compared to the radioiodine one. Considering it is a safe treatment and has fewer complications, methimazole is another option for patients with recurrent Graves', who do not wish to use radioiodine

[Using principal component analysis to increase accuracy of prediction of metabolic syndrome in artificial neural network and logistic regression models]

Background and aims: In modeling process, correlation between covariates causes multicolinearity that may reduce efficiency of the model. This study was aimed to use principal component analysis to eliminate the effect of multicolinearity in logistic regression and neural network models, and to determine its effect on the accuracy of predicting metabolic syndrome in a sample of individuals participating in the Tehran Lipid and Glucose Study

Methods: A total of 347 participants from the Cohort section of the Tehran Lipid and Glucose Study [TLGS] were evaluated. The subjects were free of metabolic syndrome, according to the ATPIII criteria, at the beginning. Logistic regression, logistic regression with principal components, neural network and neural network with principal components models were fitted to the data. The ability of the models in predicting metabolic syndrome was compared using ROC analysis and kappa statistics

Results: The area under receiver operating characteristic [ROC] curve for logistic regression, logistic regression with principal components, neural network and neural network with principal component were estimated as 0.749, 0.790, 0.890 and 0.927 respectively. Sensitivity of the models was calculated as 0.483, 0.435, 0.836 and 0.919 and their specificity as 0.857, 0.919, 0.892 and 0.964 respectively. The kappa statistic for these models was 0.322, 0.386, 0.712 and 0.886 respectively

Conclusion: the study shows that the prediction accuracy of models based on principal components is better than that of models based on primary covariates, so in the presence of multicolinearity, models based on principal components are efficient for predicting metabolic syndrome