Modelling the Latent cardio-metabolic profiles of individuals in an epidemiological investigation

doi:10.21203/rs.3.rs-1671006/v1

PDF

Research Article

Modelling the Latent cardio-metabolic profiles of individuals in an epidemiological investigation

https://doi.org/10.21203/rs.3.rs-1671006/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

Introduction: This study aimed to explore the cardio-metabolic risk factors profile to assess the underlying predictors of these profiles in healthcare workers of the AZAR cohort population.

Methods: This cross-sectional study was performed on 500 healthcare providers, official staff, and professors aged 18 to 75 years of Tabriz University of Medical Sciences. Latent profile analysis (LPA) was used to classify the subjects according systolic blood pressure, diastolic blood pressure, diabetes, smoking, hypertension, statin, metabolic syndrome, BMI, lipid profiles, hemoglobin and hematocrit. Logistic regression predicted the underlying predictors of latent profile.

Results: The individuals were classified into two LPA-driven classes, including a low risk profile (n = 210), and a high-risk profile (n = 290). High-risk group were older, and was more likely to be male than those in the low-risk group (P<0.05). The prevalence of metabolic syndrome, diabetes, hypertension, systolic blood pressure, diastolic blood pressure, HDL-cholesterol, LDL-cholesterol, total cholesterol and hemoglobin was significantly higher in the high-risk group than the low-risk group (P<0.05). But statin was not significantly different between the two groups (P>0.05). Compared with the low risk profile; BMI (OR=3.30, 95% CI: 1.10–9.93), triglyceride (OR=1.02, 95% CI: 1.00–1.05), smoking (OR=0.05, 95% CI: 0.01–0.95) and hematocrit (OR=12.07, 95% CI: 3.46–42.11) were higher in the high-risk profile.

Conclusion: Identifying risk factors may prevent the incidence of cardiometabolic disease. Higher BMI levels, higher triglyceride values, higher hematocrit, and lower HDL values are better predictors for identifying a cardio-metabolic high-risk profile.

Cardio-metabolic

Risk factors

Latent profile analysis

Prevent

Non-communicable diseases (NCD) are nontransmissible diseases among people with the highest mortality rate globally and are preventable (1). NCDs, currently have the heaviest economic burden on developing countries (2, 3). According to the WHO, among the NCDs, cardiovascular disease is one of the most critical threats to human health in the world (4, 5). Due to the lack of optimal management and redundant diagnostic tests for cardiometabolic disease, metabolic disorders of this disease often remain undiagnosed and untreated, already placing substantial strain on the health care system (6, 7). Cardiovascular disease and cardio-metabolic comorbidities in 2016, accounted for 31% of all deaths in the world and this percent morbidity and mortality rates are projected to increase by 2030 (8, 9). Cardiovascular disease in 2022 with 28.62%, is the first causes of mortality in Iran (30).

Hypertension, a low HDL cholesterol level, a high LDL cholesterol level, and a high triglyceride level are known underlying risk factors for cardiovascular disease (10-13). Levels of cardiometabolic risk factors vary among individuals based on age, sexual maturity, family history, and high levels of risk factors (14, 15). Epidemiological evidence suggests that risk factors present in childhood and persists into adulthood, and the incidence of the disease increases with age (11, 12, 15, 16). Low levels of physical activity, low carbohydrate diet, socioeconomic status, cardiorespiratory fitness, and smoking are cardiovascular disease modifiers risk factors, and control mortality from this disease (3, 11, 13).

Latent profile analysis (LPA) is a form of finite mixture modelling and a person-centered approach. Person-centered analysis is to identify latent class or groups that is used to discriminate the individual based on the characteristics (17-19). LPA, be utilized to determine whether the risk profiles defined adequately gives details on the cardio-metabolic. In addition, information on risk factors can provide a basis for better identifying high–risk profiles and deal with cardio-metabolic. Therefore, the purpose of the current study was:

1- To identify latent profiles and to investigate associations between cardio-metabolic risk factors, and classify individuals into high-risk and low-risk groups using the LPA.

2- To determine whether these profiles demonstrate associations with cardiometabolic characteristics of the individual.

Study design, aims, and participants information

In this cross-sectional study, we studied a cohort of individuals that are parts of a larger prospective epidemiological research in Iran (the Azar Cohort Study) (20). The current study was carried out on healthcare workers in 2020 as a part of the Azar cohort study, which was conducted by the liver and gastrointestinal diseases research center of Tabriz University of Medical Sciences (TBZMED) (21).

The purpose of this cohort study was to evaluate 3000 participants, including healthcare employees in hospitals, schools, and health networks of TBZMED. A total of 500 persons participated in this study. Our baseline assessment consisted of a face-to-face health interview or a health examination regarding a broad range of established and novel risk factors for NCDs.

Eligibility criteria

Participants of this study include full-time and long-term contract employees aged 18 to 75 years who are not pregnant or breastfeeding, and who are not planning to retire within the next five years. Patients who reported a history of debilitating psychiatric disorders or physical illnesses by a health professional were excluded from this study.

Ethical considerations

All involved participants provided written informed consent, and the Institutional review board (IRB) of TBZMED (IR.TBZMED.REC.1396.1263) approved the study.

Main variables

Age, sex, systolic blood pressure, diastolic blood pressure, diabetes, smoking, hypertension, statin, metabolic syndrome, HDL cholesterol, LDL cholesterol, total cholesterol, triglycerides, hemoglobin, and hematocrit circumference were assessed and analyzed in our PLA model.

Demographic Characteristics of the Participants

We used a questionnaire for evaluating the demographic characteristics, including age (years), sex (1: Male, 2: Female), marital status (1: Single, 2: Married, 3: Widow, 4: Divorce, 4: Other), and educational level (1: Illiterate, 2: Primary, 3: Secondary, 4: High school graduated, 5: Associate degree, 6: Bachelor, 7: Master, 8: PhD). Moreover, lifestyle patterns, i.e., smoking (1: Past or current, 2: Never), diabetes (1: Yes, 2: No), hypertension (1: Yes, 2: No), statin (1: Yes, 2: No), metabolic syndrome (1: Yes, 2: No), drug use (1: Daily, 2: Weekly, 3: Monthly), hookah use (1: Past or current, 2: Never), alcohol consumption (1: Past or current, 2: Never), were also assessed by the questionnaire.

Measurements

Height is recorded to the nearest 0.5 cm using a mounted tape measure, subjects without shoes and their arms hanging freely by their sides. Barefoot subjects with only light clothing have their weight recorded to the nearest 0.1 kg on a Seca scale (Seca®, Germany). Body mass index was determined using the standard formula, i.e., weight (kg)/height2 (m) (21). Waist of subjects is measured according to the USA National Institutes of Health (NIH) guidelines. Blood pressure was measured by a trained nurse twice with a two-minute interval, i.e., twice for each arm in a sitting position after 10 minutes of rest using a mercury sphygmomanometer (Rudolf Richter, DE-72417, Germany). Lipid profiles (total cholesterol, high-density lipoprotein cholesterol (HDL-cholesterol), and triglycerides) were assessed using serum samples, which were analyzed using Miura One automated equipment (I.S.E., Rome, Italy) and a commercial DiaSys kit (DiaSys Diagnostic Systems, Hamburg, Germany) (22). Low-density lipoprotein cholesterol (LDL-cholesterol) was calculated according to the Fried Ewald equation (23). Hematocrit was measured using (EKF Diagnostics UltraCrit ™ Hematocr, USA). A photometric device (HemoCue® Hb 201+) is used to determine the hemoglobin content of the blood (24). Insulin was measured with an enzyme immunoassay (microtiter plate format; Dako Diagnostics, Ely, United Kingdom).

Sample size

The census was conducted during a specific period on all eligible health care providers, official staff, and lecturers of Tabriz University of Medical Sciences.

Statistical analysis

Continuous variables are presented as mean and standard deviation (SD) and categorical variables as the frequency count and percentage. The continuous variables were checked and confirmed for normality by distribution measures, namely skewness within ±1.5 and kurtosis within ±2. Chi-square tests with exact p-value and one-way analysis of variance (ANOVA) followed by Tukey post hoc test were used to assess the difference among high risk and low risk factors, for categorical and normal continuous variables, respectively. We used the Akaike information criterion (AIC), Bayesian information criterion (BIC) and Adjusted Bayesian information criterion (ABIC) To compare and select the best-fitted model. For these statistics, the smaller values show better fit of the model. Each class was described in terms of entropy measurements that ranges from 0.00 to 1.00, and also sociodemographic factors. Furthermore, the LPA-driven classes were subsequently entered into a logistic regression model (backward LR) to test for associations between each class and the risk of cardiometabolic with adjustment for smoking, HDL cholesterol, triglyceride, BMI, and other clinical variables. The significance thresholds for variables were set at P <0.05. LPA was performed using MPlus 7.4 (Muthen and Muthen, Los Angeles, CA, USA) and the other analyses using SPSS 17 (SPSS Inc., Chicago, IL, USA) software.

The total number of subjects in the initial evaluation was 500, of which 493 (98.6%) remained in the study. The final sample included 493 participants and 63.8% were male. The mean age of patients was 43 (SD 7.2) and the mean BMI of subjects was 28 (SD 4.0). Also, in this study, 89% of patients whose smoking status was never and 9.8% smoked were past or current. Other information about the variables was reported in Table1.

Table 1

Participant demographic characteristics

variables	n/mean	%/SD
Age (years)	43.2	7.2
Sex (male)	319	63.8
Diabetes
Yes	21	4.2
No	479	95.8
Smoking
Past or current	49	9.8
Never	445	89.0
Hypertension
Yes	38	7.6
No	462	92.4
Statin
Yes	13	2.6
No	487	97.4
Metabolic syndrome
Yes	386	77.2
No	114	22.8
BMI (kg/m²)
<=25.00	107	21.4
25.01-30.00	257	51.4
30.01-300	109	21.8
35.01+	24	4.8
SBP (mmHg)	115.5	14.5
DBP (mmHg)	75.5	9.5
HDL cholesterol (mg/dL)	45.2	10.7
LDL cholesterol (mg/dL)	107.8	31.6
Triglyceride (mg/dL)	126.4	60.0
Total cholesterol (mg/dL)	178.3	38.7
Hemoglobin (g/dL)	15.5	1.6
Hematocrit (L/L)	44.9	4.4

LCA identified two classes of study subjects. The model fit indices are summarized in Table 2.

Table 2

Model-fit indices for the latent class analysis mode

Number of classes	Loglikelihood	#free_Parm	LR Chi2 (df)	AIC	BIC	ABIC	entropy
1 Class	-19170.0	28	249.56(308)	38396.0	38514.0	38425.1	---
2 Class	-18638.4	48	190.40(296)	37372.8	37575.1	37422.7	0.910
3 Class	-18409.5	68	118.42(286)	36955.0	37241.6	37025.8	0.916
4 Class	-18308.2	88	130.32(275)	36792.5	37163.4	36884.0	0.888
5 Class	-18214.1	108	125.51(264)	36644.1	37099.3	36756.5	0.870
6 Class	-18125.8	128	120.76(253)	36507.7	37047.1	36640.9	0.869
7 Class	-18036.3	148	109.12(241)	36368.7	36992.4	36522.7	0.899

LCA models with 1–7 classes were compared for model fit and interpretability. We chose a two-class model that showed the best fit measured by a lower Bayesian information criterion with higher entropy.

Figure 1 shows the conditional probabilities for each cardio-metabolic risk factors in each class. Class 1 subjects (low risk, n = 210) generally had lower systolic blood pressure, diastolic blood pressure, and hematocrit, reported higher HDL but lower cholesterol. Class 2 (high risk, n = 290) were more likely to have higher systolic blood pressure, diastolic blood pressure, and hemoglobin. Hypertension and statin levels similar in both classes.

Table 3 summarizes the social, demographic, and clinical factors and cardiometabolic related risk factors in each class. Subjects in the high-risk group were older, had a higher BMI, and was more likely to be male than those in the other group (all P<0.05). Furthermore, the prevalence of metabolic syndrome was significantly higher in the high-risk group (71.9%) than the low-risk group (28.1%). The high-risk class was likely to have more diabetes, hypertension and smoking compared to the other class. There was a significant difference between all variables except statin (P>0.05).

Table 3

Baseline characteristics according to LPA-derived classes

Variables	Low risk	High risk	p-value #
n	210	290
Age (years)	42.3(7.1)	43.8(7.2)	<0.001
Sex n (%)			<0.001
Male	30(14.3)	289(99.7)
Female	180(85.7)	1(0.3)
Underlying disease
SBP (mmHg)	107.8(11.3)	120.9(13.9)	<0.001
DBP (mmHg)	70.5(7.5)	79.1(9.1)	<0.001
Diabetes n (%)			<0.001
Yes	0(0.0)	21(100.0)
No	210(43.8)	269(56.2)
Smoking n (%)			<0.001
Past or current	6(12.2)	43(87.8)
Never	202(45.4)	243(54.6)
Hypertension n (%)			0.042
Yes	10(26.3)	28(73.7)
No	200(43.3)	262(56.7)
Statin n (%)			0.406
Yes	4(30.8)	9(69.2)
No	206(42.3)	281(57.7)
Metabolic syndrome n (%)			<0.001
Yes	32(28.1)	82(71.9)
No	178(46.1)	208(53.9)
BMI (kg/m²)			0.049
<=25.00	57(53.3)	50(46.7)
25.01-30.00	102(39.7)	155(60.3)
30.01-35.00	39(35.8)	70(64.2)
35.01+	10(41.7)	14(58.3)
HDL cholesterol (mg/dL)	50.0(11.4)	41.6(8.9)	<0.001
LDL cholesterol (mg/dL)	104.3(31.6)	110.2(31.3)	<0.001
Triglycerides (mg/dL)	104.8(42.6)	141.9(65.7)	<0.001
Total cholesterol (mg/dL)	175.4(39.2)	180.3(38.2)	<0.001
Hemoglobin (g/dL)	14.1(1.1)	16.5(1.1)	<0.001
Hematocrit (L/L)	40.9(2.9)	47.6(2.8)	<0.001

BMI body mass index, SBP systolic blood pressure, DBP diastolic blood pressure, HDL high-density lipoprotein, LDL low-density lipoprotein.

For qualitative variables, the data are expressed as frequency (percent), # Chi-square test with exact procedure.

For quantitative variables, the data are expressed as mean (SD), #T-test.

The results of logistic regression evaluating the associations between the two LPA-driven classes and cardio-metabolic risk factors are depicted in Table 4. Logistic regression analysis showed that the high-risk profile was significantly associated with higher BMI, triglycerides, and hematocrit, but lower HDL cholesterol, and past or current smoking.

Table 4

Adjusted odds ratio and 95% confidence interval for cardio-metabolic in association with latent profiles

variables	Adjusted odds ratio	95% confidence interval	P-value
Smoking
Never smoked	Referent
Past or current smoke	0.05	0.01,0.95	0.046
BMI (kg/m2)	3.309	1.10, 9.93	0.033
HDL cholesterol (mg/dL)	0.88	0.79, 0.99	0.035
Triglyceride (mg/dL)	1.02	1.00, 1.05	0.039
Hematocrit (L/L)	12.07	3.46, 42.11	<0.001

The backward LR was carried out; the other variables (SBP, DBP, LDL cholesterol, total cholesterol, hemoglobin, metabolic syndrome, hypertension, diabetes, statin, sex, age) were removed from the model due to Linearity.

This study identified two latent profiles of cardio-metabolic risk factors, low-risk and high-risk profiles. Notably, a high-risk profile also has a higher prevalence of metabolic syndrome, smoking, and hypertension, indicating cardio-metabolic risk factors. We considered the cardio-metabolic risk factors as a latent profile rather that their clustering as metabolic syndrome because the metabolic syndrome as predictor of cardiovascular diseases beyond its component abnormalities remains controversial and different definitions of the syndrome may vary considerably depending on the sets of criteria (25, 26). There are only a few studies that have used LPA to detect the latent profiles of cardio-metabolic syndrome.

In the study by Verswijveren et al. Hypertension rem 3 latent profiles for cardio-metabolic risk factors in youth were identified: prolonged sitters, breakers and prolonged movers. The breakers were on average one year younger than the other two groups and have been the highest proportions of girls. Higher BMI, waist circumference, systolic blood pressure, diastolic blood pressure and higher cardio-metabolic risk-score were higher in prolonged sitters than the other two groups. Triglyceride and HDL were equal in prolonged sitters and prolonged movers. LDL was lower in the latent profile prolonged movers than the other two groups. Based on the results of multivariable logistic regression, higher BMI level and waist circumference were significantly connected to the higher risk (27). The results of this study were in the line with the results our study. In breaker profiles similar to low-risk profiles, lower mean age, higher female proportions, and HDL cholesterol value have been the higher. In prolonged sitters profile similar to high risk profile, systolic blood pressure, diastolic blood pressure, BMI, triglyceride and LDL were higher. In addition, prolonged movers had moderate risk profile.

In the study by Caspian, conducted on Iranian students, four latent classes of cardio-metabolic components have been identified. These classes were labeled as healthy (59.6%), low risk (20.4%), moderate risk (13.7%) and high risk (6.4%). In high-risk class, individuals had high triglyceride, high LDL, and high cholesterol. In contrast, students in healthy class, were not engaged in any of the cardio-metabolic components. Students in low-risk class, had high triglyceride. In the moderate risk class, abdominal obesity and obesity were high. In all four latent classes, the probability of having low HDL has been reported as low. Based on the results of multivariable logistic regression, sex, parental obesity, high screen time, low physical activity and region (rural) were significantly connected to the moderate risk class. Age and sex were significantly connected to the low-risk class (28). Based on our results, in line with the Caspian study, people in the high-risk group had high triglycerides, high LDL, high cholesterol, and low HDL. The latent profiles among Iranian children and our study reveal a different pattern of cardio-metabolic among the adults and children in Iran, which different age categories may explain these contradictory statements.

The present study showed significant differences among risk factors by sociodemographic characteristics and profiles. Age, sex, diabetes, smoke, hypertension, BMI, and metabolic syndrome distribution were significantly connected with the profiles, and were also significant in these two classes. LDL cholesterol, triglycerides, total cholesterol, systolic blood pressure, diastolic blood pressure, hemoglobin, and hematocrit levels were significantly higher among high-risk class individuals than individuals in low-risk profiles. In contrast, HDL cholesterol levels were significantly higher in low-risk individuals than in those with high-risk profiles. Based on the results of multivariable logistic regression, higher BMI level, higher triglyceride values, higher hematocrit, and lower HDL values were significantly connected to the high-risk profile.

In the study by Burgos et al. a predictive analysis of the early development of cardio-metabolic risk factors, including age, cardiorespiratory fitness, total cholesterol, triglycerides, LDL cholesterol, and fat percentage was lower in males than in females. Also, males showed higher values of systolic blood pressure, waist circumference, HDL cholesterol, and BMI. In this study, males generally showed better cardio-metabolic risk factors values than females. Based on the results of multivariable logistic regression, cardiorespiratory fitness, total cholesterol, triglycerides, LDL cholesterol, age, and percentage of fat were significantly lower among males than females. In contrast, males had a significantly higher level of systolic blood pressure, waist circumference, HDL cholesterol and BMI than females. In this study, HDL cholesterol, age and BMI were not significant between males and females (22). In this study, in contrast to our study, cardio-metabolic risk factors were higher in female than male.

In the study by Sodjinou et al. in the urban adults of Benin, mean BMI, waist circumference, and HDL cholesterol level were markedly higher among women than men. Also, the mean age was higher in women than men. In contrast, triglyceride concentrations were significantly higher in men compared to women. Plasma glucose, systolic and diastolic blood pressure did not vary significantly according to sex. The most prevalent risk factors in the study sample were overall obesity (18%), abdominal obesity (32%), hypertension (23%), and low HDL cholesterol (13%). Risk factors indicated a higher prevalence of cardio-metabolic syndrome incidence in women (29), which was not compatible with the finding of our studies.

In summary, latent profile analysis furthers our understanding of the cardio-metabolic risk factors. These analyses found that a two-profile latent factor best explains associations among the cardio-metabolic risk factors. Therefore, different predictors could be associated with developing the cardio-metabolic risk factors. These results could be considered in public health policies aimed to prevent the early development of cardio-metabolic risk factors.

Limitations

Our study has several limitations. Firstly, due to the cross-sectional nature of this study, it is not possible to assess the temporal relationships. Second, the study sample is only representative of participants related to Tabriz University of Medical Sciences, and thus findings may not apply to the whole urban population of Tabriz. The study may probably lack statistical power due to the modest sample size. The conclusions of the study should therefore be interpreted with caution. Metabolic heart risk factors can be influenced by sex and age. These limitations notwithstanding, our study provides useful data on the prevalence of cardio-metabolic risk factors among participants related to Tabriz University of Medical Sciences correlates.

Our study showed that higher BMI level, higher triglyceride values, higher hematocrit, and lower HDL values are the main predictors of cardio-metabolic high risk profile. High risk group has a higher risk of developing a cardio-metabolic risk profile, especially in males. With regard to our results, it seems necessary to design cross-sectional studies to determine the incidence rate of cardio-metabolic and to assess the changes in latent profiles by considering time varying effects and related factors. Therefore, our results suggest that risk factors measurement could be a valuable method of identifying individual with cardio-metabolic risk based on epidemiologic studies. Also, identifying risk factors may prevent the incidence of cardio-metabolic disease and inflammatory markers.

Ethics approval and consent to participate

The institutional review board of Tabriz University of medical sciences approved the protocol of the study (ethics code: IR.TBZMED.REC.1400.1006). The participants' privacy was preserved. All participants filled and signed the informed consent and assent. All methods were carried out in accordance with relevant guidelines and regulations.

Consent for publication

'Not applicable.'

Availability of data and materials

The data that support the findings of this study are available from MAJ, but restrictions are applied to the availability of these data, which were used under license for the current study, and are not publicly available. Data are, however, available from the authors upon reasonable request by MAJ.

Competing interests

The authors declare that there is no conflict of interest.

Funding

This study was supported by a research deputy of Tabriz University of medical sciences.

Author contributions

All authors read and approved the final manuscript. MAJ and PE conceived the study and participated in the design and data collection. MAJ, PE, NR, SM, SG, and NMA participated in the data analysis and manuscript preparation.

Acknowledgment

We would like to appreciate the collaboration of the department of biostatistics and epidemiology, faculty of Health, Tabriz University of medical sciences, for providing the environment for modeling data and manuscript writing.

Beishuizen CR, Stephan BC, van Gool WA, Brayne C, Peters RJ, Andrieu S, et al. Web-based interventions targeting cardiovascular risk factors in middle-aged and older people: a systematic review and meta-analysis. Journal of medical Internet research. 2016;18(3):e5218.
Daneshpour MS, Fallah M-S, Sedaghati-Khayat B, Guity K, Khalili D, Hedayati M, et al. Rationale and design of a genetic study on cardiometabolic risk factors: protocol for the Tehran Cardiometabolic Genetic Study (TCGS). JMIR research protocols. 2017;6(2):e6050.
Bovet P, Romain S, Shamlaye C, Mendis S, Darioli R, Riesen W, et al. Divergent fifteen-year trends in traditional and cardiometabolic risk factors of cardiovascular diseases in the Seychelles. Cardiovascular diabetology. 2009;8(1):1-13.
Mathers CD, Loncar D. Projections of global mortality and burden of disease from 2002 to 2030. PLoS medicine. 2006;3(11):e442.
Bubach S, Horta BL, Gonçalves H, Assunção MCF. Early age at menarche and metabolic cardiovascular risk factors: mediation by body composition in adulthood. Scientific Reports. 2021;11(1):1-9.
Reiter-Brennan C, Dzaye O, Davis D, Blaha M, Eckel RH. Comprehensive care models for cardiometabolic disease. Current Cardiology Reports. 2021;23(3):1-11.
Alwan A. World Health Organization Global status report on noncommunicable diseases. 2010.
Murli S. Adopt a Dog: Your Heart Will Thank You. Reading Time. 2019.
Ghasemirad, M., et al. (2021). "The Trend of Frequency and Cost of Various Outpatient Diagnostic Procedures for Cardiovascular Diseases in Isfahan Province, Iran." Health Information Management 17(6): 274-278.
Cook S, Weitzman M, Auinger P, Nguyen M, Dietz WH. Prevalence of a metabolic syndrome phenotype in adolescents: findings from the third National Health and Nutrition Examination Survey, 1988-1994. Archives of pediatrics & adolescent medicine. 2003;157(8):821-7.
Winkleby MA, Robinson TN, Sundquist J, Kraemer HC. Ethnic variation in cardiovascular disease risk factors among children and young adults: findings from the Third National Health and Nutrition Examination Survey, 1988-1994. Jama. 1999;281(11):1006-13.
Jago R, Harrell JS, McMurray RG, Edelstein S, El Ghormli L, Bassin S. Prevalence of abnormal lipid and blood pressure values among an ethnically diverse population of eighth-grade adolescents and screening implications. Pediatrics. 2006;117(6):2065-73.
Tracy RE, Newman III WP, Wattigney WA, Srinivasan SR, Strong JP, Berenson GS. Histologic features of atherosclerosis and hypertension from autopsies of young individuals in a defined geographic population: the Bogalusa Heart Study. Atherosclerosis. 1995;116(2):163-79.
Tell GS. Cardiovascular disease risk factors related to sexual maturation: the Oslo Youth Study. Journal of chronic diseases. 1985;38(8):633-42.
Mahoney LT, Lauer RM, Lee J, Clarke WR. Factors affecting tracking of coronary heart disease risk factors in children. The Muscatine Study. Annals of the New York Academy of Sciences. 1991;623:120-32.
Muntner P, He J, Cutler JA, Wildman RP, Whelton PK. Trends in blood pressure among children and adolescents. Jama. 2004;291(17):2107-13.
Muthén B, Muthén LK. Integrating person‐centered and variable‐centered analyses: Growth mixture modeling with latent trajectory classes. Alcoholism: Clinical and experimental research. 2000;24(6):882-91.
Bauer DJ, Curran PJ. The integration of continuous and discrete latent variable models: potential problems and promising opportunities. Psychological methods. 2004;9(1):3.
Little TD. The Oxford handbook of quantitative methods in psychology: Vol. 2: statistical analysis: Oxford University Press; 2013.
Poustchi H, Eghtesad S, Kamangar F, Etemadi A, Keshtkar A-A, Hekmatdoost A, et al. Prospective epidemiological research studies in Iran (the PERSIAN Cohort Study): rationale, objectives, and design. American journal of epidemiology. 2018;187(4):647-55.
Farhang S, Faramarzi E, Amini Sani N, Poustchi H, Ostadrahimi A, Alizadeh BZ, et al. Cohort profile: The AZAR cohort, a health-oriented research model in areas of major environmental change in Central Asia. International journal of epidemiology. 2019;48(2):382-h.
Burgos MS, Reuter CP, Possuelo LG, Valim ARdM, Renner JDP, Tornquist L, et al. Obesity parameters as predictors of early development of cardiometabolic risk factors. Ciencia & saude coletiva. 2015;20:2381-8.
Gevers Leuven J, Kempen H, de Wit E, Havekes L. The Effect of Cyclandelate on Cholesterol Metabolism in Patients with Familial Hypercholesterolaemia. Drugs. 1987;33(2):131-5.
Curtis CM, Kost GJ, Louie RF, Sonu RJ, Ammirati EB, Sumner S. Point-of-care hematology and coagulation testing in primary, rural emergency, and disaster care scenarios. Point of care. 2012;11(2):140.
Reaven G. Insulin resistance, the insulin resistance syndrome, and cardiovascular disease. Panminerva medica. 2005;47(4):201-10.
Kahn R. Metabolic syndrome: is it a syndrome? Does it matter? Circulation. 2007;115(13):1806-11.
Verswijveren SJ, Lamb KE, Leech RM, Salmon J, Timperio A, Telford RM, et al. Activity accumulation and Cardiometabolic risk in youth: a latent profile approach. Medicine and science in sports and exercise. 2020;52(7):1502-10.
Abbasi-Ghahramanloo A, Heshmat R, Rafiemanzelat A-M, Ghaderi K, Motlagh ME, Ahadi Z, et al. Subgrouping of Iranian children and adolescents based on cardiometabolic risk factors using latent class analysis: The CASPIAN-V study. Caspian Journal of Internal Medicine. 2020;11(4):370.
Sodjinou R, Agueh V, Fayomi B, Delisle H. Obesity and cardio-metabolic risk factors in urban adults of Benin: relationship with socio-economic status, urbanisation, and lifestyle patterns. BMC public health. 2008;8(1):1-13
30. Available from https://www.worldlifeexpectancy.com/iran-coronary-heart-disease [access date 2022/05/15]

No competing interests reported.

PDF

Reviews received at journal
23 Jun, 2022
Reviewers agreed at journal
23 Jun, 2022
Reviewers invited by journal
23 Jun, 2022
Editor assigned by journal
22 Jun, 2022
Editor invited by journal
24 May, 2022
Submission checks completed at journal
24 May, 2022
First submitted to journal
18 May, 2022

You are reading this latest preprint version

Modelling the Latent cardio-metabolic profiles of individuals in an epidemiological investigation

Status:

Version 1

Abstract

Figures

Introduction

Methods

Results

Discussion

Limitations

Conclusion

Declarations

References

Competing Interests

Status:

Version 1