info@aditum.org    +1(205)-633 44 24

Index of Socioeconomic Status and Index of Cardiovascular-risk factors as Predictors of Cardiovascular Disease

Authors

Satyendra Nath Chakrabartty
Indian Ports Association, Indian Statistical Institute.

Article Information

Corresponding author: Satyendra Nath Chakrabartty, ORCID: 0000-0002-7687-5044

Present Address: Flat 4B, Cleopatra, DC 258, Street No. 350, Action Area 1, New Town, Kolkata.

Received: November 06, 2026           |        Accepted: November 15, 2026        |         Published: November 20, 2026

Citation: Satyendra N Chakrabartty (2026) “Index of Socioeconomic Status and Index of Cardiovascular-risk factors as Predictors of Cardiovascular Disease.” journal of clinical cardiology interventions, 6(1). DOI: 10.61148/2836-077X/JCCI/057.

Copyright:  © 2026 Satyendra Nath Chakrabartty. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Abstract

Background: Associations among components of socioeconomic status (SES) and cardiovascular risk factors (CVRFs) gave contrasting results. Understanding of socioeconomic determinants of cardiovascular disease (CVD) and their association is crucial to mitigate CVD burden and achieving national targets for CVD. 

Method: Avoiding problems of scaling and selection of weights, the paper describes index of CVD-risk factors (ICVD-RFt) and another Index of socioeconomic status (SESt) for a country or region at t-th year combining subjective and objective measures of SES and CVRFs in different forms by multiplicative aggregation. The unit-free indices can cover data in different units and forms like ordinal scores, count data, cardinal data in proportion, percentages, ratios, rates, etc. avoiding normalization and weights of the chosen indicators.

Results: The proposed indices reduces substitution effect among the chosen indicators and satisfy desired properties like translation invariance, aggregation consistency, time-reversal test, formation of chin indices, etc. Significant correlation between ICVD-RFtand SESt enables undertaking multivariate regression of ICVD-RFt on components of SESt or canonical correlation analysis (CCA) considering p-numbers of CVRFs (likely effects) and q-numbers of SES-components (possible factors).  Empirically fitted regression equation of CVD death rate on the components of proposed two indices facilitates making sound decisions regarding identification of critical areas to plan appropriate interventions and direction of improvement providing support in clinical decision‑making and public‑health policies towards mitigation of CVD-burden and to achieve equitable and effective disease prevention.

Conclusion: Novelty of the method lies in known probability distribution of proposed indices facilitating parametric analysis for comparisons, classifications and enabling statistical tests of equality of mean, variance, progress made across groups or time periods. Professionals and researchers can take advantages of the proposed method including detection of changes by longitudinal data. The method with theoretical advantages is recommended

Keywords:

multiplicative aggregation, lognormal distribution, regression, canonical correlation, relative importance

Introduction:

Introduction:

Socio-economic status (SES) is inversely related with cardiovascular disease burden [1]. Coronary heart disease (CHD) is more frequent among people with low SES. SES can be regarded as a cardiovascular risk factors (CVRFs) especially for male adults [2]. In addition to traditional CVRFs like hypertension, dyslipidaemia, diabetes, smoking, family history of CHD, social factors reflected by SES on cardiovascular risk is significant [3].

While people with poor financial, educational, occupational status (low SES) are prone to (CVD), individuals with higher SES recognize the hazard ratio of myocardial infarction and help initiation of medical care [4]. Stress resulting from job insecurity, financial instability, social marginalization, etc. has been found to be associated with higher level of cortisol affecting sympathetic nervous system, endothelial dysfunction, arterial stiffness, atherogenesis, etc. [5]. However, contrasting results have been observed regarding association between SES and CVD and CVRFs over genders and age-profiles. While low SES was related to metabolic syndrome and hypertension, higher SES and obesity were positively associated [6].

In India, Ischemic heart disease contributes maximum towards death due to CVD followed by strokes [7]. As per World Economic Forum (WEF) and Harvard School of Public Health (HSPH), India is likely to lose around $2.17 trillion by 2030 due to CVD alone [8].

Models to study SES and CVD could be interactive [9] or additive showing effect of physiological, psychosocial and environmental factors influencing CVD-risks [10]. However combining environmental factors and psychosocial factors are problematic. Understanding CVD-risks help in better clinician-patient discussions for primary prevention of CVD and adoption of preventive strategies towards lipid-lowering and anti-hypertensive therapies [11]. It often involves questionnaires covering items on risk-factors like age, gender, family history, lifestyle habits, etc. (Ordinal level data) and pathological findings like blood Systolic and Diastolic pressure, cholesterol levels, left and right ventricular hypertrophy, levels of blood glucose and triglyceride, etc.(cardinal data in different units). Aggregation of such data following unknown distributions suffers from methodological limitations [12]. Such models differ in form, characteristics, and do not perform well in heterogeneous populations with possibility of overestimation or underestimation which limit applicability of temporal and spatial changes across time and space [13].  New CVD-risk markers evaluated through Systematic Coronary Risk Evaluation (SCORE2), AHA Predicting Risk of CVD Events (PREVENT), lipoprotein(a) (Lp[a]) risk calculator, etc. are not being used frequently in CVD-risk models, despite strong agreement with ASCVD risk scores [14]. For improvement, CVD-risk tools based on all relevant indicators with validation across populations, the PREVENT-CVD equations evaluating 19-risk-factors, treatment status, urine albumin-creatinine ratio (UACR), HbA1c, social deprivation index (SDI) etc. was developed for 10-year and 30-year periods [15]. However, challenges remain on effective CVRFs stratification and finding clinically meaningful risk thresholds for different age-groups across genders especially for pharmacotherapy of hypertension and hyperlipidemia [16].

Dimensions and indicators of SES assessment tools vary. SES indicators chosen by [17], [18], [19] are different. Shortcomings of popular SES scales in India were highlighted [20]. Better understanding of socioeconomic determinants of CVD and their association following sound approach is crucial to achieve national CVD targets [21]. Hence, methodologically sound comprehensive index of SES for a region/country is imperative.

The paper describes index of CVD-risk factors at t-th year (ICVD-RFt) and index of socio-economic status (SESt) satisfying desired properties and facilitating meaningful comparisons, classifications and finding association by correlating ICVD-RFt and SESt. Relationship between the two indices can be derived by multivariate linear regression or by canonical correlation based on CVRF-vector (p ×1) and SES-vector (q ×1), even if ????≠????.   

Literature survey:

Avoiding fragmented results on effect of individual CVRFs, quantification of overall CVRFs by estimation models appears to have advantages. Major limitations of such models were addressed [22]. Illustrative list is as follows:

-Use of non-uniform risk factors and methodological limitations of their evaluation. The obsolete Framingham Heart Study is not recommended [23].

- CVD-risks in different scales and units are not additive.  For example, blood pressure in mmHg, cholesterol in millimoles per litre of blood (mmol/L), Diabetes in (mmol/L), BMI in kg/m2, number of cigarettes smoked per day, etc. are clearly not additive.

- Thresholds for each CVRF differed. For example, thresholds for blood-pressure and cholesterol were different for therapeutic measures [24].

- CVD-risks failed to explain variability across age-groups.

- While PROCAM risk model [25] underestimated CHD-risk among men, Reynolds Risk Scores overestimated the same for women [26].

- Questions raised about applicability of CVRF-models across heterogeneous groups [27]. 

- Prediction of CVRFs by such models cannot be generalised and many are not yet validated externally [23].

-  Selection of CVRF models can influences pharmacotherapy decisions. For example, Lipid-lowering therapy starts depending on estimated risk > model-specific thresholds and other clinical qualifiers [28].

- Statin eligibility dependents on selection of CVRFs prediction models and showed modest inter-model agreement [29]. 

Each CVRF-score was transformed to equidistant scores (Ei) which were standardized to Z –scores and transformed to proposed item score Pi in score range [0, 100] following normal distribution, enabling meaningful addition so that Pi= index to reflect overall cardiovascular health risk ICVD-RFt 30,  for categorical variables, weights assigned were  frequency of the j-th levelTotal sample size ×100 to get E-scores. Measures of biomarkers and laboratory findings in ratio scales satisfying equidistant property, were transferred directly to P-scores via Z-scores. Despite several advantages, the method following arithmetic aggregation suffers from substitution effect allowing higher values to compensate lower values of CVRFs.

Multidimensional SES indicates social standing of individuals/families/communities, on the basis of their income, education, power, occupation, etc. and access to socio-economic-cultural resources like money, shelter, food, healthcare, education, social opportunities, etc. [31]. Higher SES generally correlates with better living standard, better access to services, and increased opportunities for development of cognitive and socio-emotional processes [32]. To measure SES, subjective variables are obtained by survey using questionnaires/scales developed for the purposes, objective measures are collected from secondary data. Inclusions of more variables to measure comprehensive SES was suggested for effective socio-economic strategies of local and national governments [33]. Most important issue is aggregation of these variables in different levels of measurement, different score-ranges and unknown distributions of the component variables ensuring the resulting composite index (CI) to satisfy aggregation consistency (i.e. index obtained in stages = Index computed in a single stage) facilitating meaningful measurement across space and time. 

Arithmetic aggregation of ordinal scores of a chosen set of indicators is not meaningful since levels of items in questionnaires/scales are not equidistant. For example, trait distance between two consecutive levels of Zung Self-Rating Depression Scale (ZSDS) [34] are non-uniform. If Di,j denotes distance between i-th and j-th rung of 10-rungs MacArthur Scale of Subjective Social Status, then D10,9≠ D9,8 ≠ 2*D9,7 and may not show behaviour of the latent construct in accurate fashion. Hierarchy of occupations in six classes [35] does not represent equal distance between successive classes. Even for objective measures of SES, converting people or households in a fixed score-range could be non-equidistant. Assigning scores of 0 – 8 for each chosen objective measures is arbitrary [36]. Meaningful addition like Z=X+Y and indication of bias are not possible for unknown and different sampling distributions [37].

Other SES measuring scales consider monetary dimensions like income, wealth, assets, etc. and non-monetary dimensions like education, maternal education, occupation, household prestige, water and sanitation, etc.  However, too much reliance on income-based criteria which changes with Consumer Price Index (CPI) limits accurate assessment of socioeconomic disparities. Each such scale with limitations may not be appropriate for comparisons across space and time [38]. Other limitations are:

  • May be biased due to social desirability deviating from actual situations [39]
  • Equal importance to all items is unjustified against different values of like item-total correlations, inter-item correlations, factor loadings, etc.
  • Increase in number of levels changes score distribution and influence reliability, validity more than the underlying variable [40].
  • Classification of scores of Socioeconomic Status Composite Scale (SES-C), In Charge Financial Distress‑Financial Well‑Being (IFDFW) scale etc. disclosed no measure of classification efficiency [41-42].
  • Income is common to both Modified Kuppuswamy scale and BG Prasad scale. It requires frequent updations based on monthly Consumer Price Index for Industrial Workers (CPI-IW), (http://labourbureau.gov.in/LBO_indnum.htm).

Weighted sum approach of SES follow different ways to select weights which influence properties of CI. Different selection of weights may not ensure minimum variance of the resultant CI or equal correlation of CI with the component indicators.  To measure monthly per capita household income (HI) as weighted sum, [43] selected weights from PCA, conditional random forests (RF) and fitted linear regression equations to predict HI where prediction error was minimum for RF implying RF analysis of SES could be better than PCA. Moreover, weighted sum approach is also associated with compensability among the targets/indicators and is undesirable [44]. Asset indices using PCA weights and unweighted indices performed with similar efficiency and SES of households can be well assessed without weights [45]. PCA weights ignore indictors which are weakly correlated with CI, and may be distorted by presence of outliers, normalization methods, etc. Different transformations for normalizing SES components resulted in different coefficients of logistic regression analyses of SES on health [46]. Thus, assessing SES without normalization and choosing weights is desirable.

Proposed method:

Ensure positive relationship of each indicator with CI i.e. higher value indicates better performance. Without resorting to normalization and selection of weights, aggregation is proposed by SESt= 1qYitXi0   where Y1t, Y2t, …….., Yqt denote values of the chosen indicators in t-th period and Y10, Y20, …….., Yq0 denote corresponding values of those indicators in base period (Yit,Yi0>0) of a region/country.  Thus,

 SESt=  Y1t, Y2t,…….., YqtY10 Y20 …….. Yq0                                                                             (1)

Similarly, denoting observed values of p-indicators of CVD-risk factors (both traditional and new CVD-risk markers) at t-th year of the a region/country by X1i, X2i, ……,Xpi  and the corresponding base period values as X1i0, X2i0, ……,Xpi0, the index ICVD-RFt is defined as

ICVD-RFt= 1pXiXi0                                                                                                (2)

Note that

  1. Each proposed index (SESt and ICVD-RFt) is geometric mean (GM) of unit-free ratios avoiding the q-th or p-th root.
  2. (1) and (2) become additive and differentiable when logarithm is taken on both sides of the equations like

logICVD-RFt=j=1plogXji-j=1plogXji0                                                               (3)

  1. Each of SESt and ICVD-RFt follows lognormal distribution since distribution of GM approaches lognormal distribution when effects of different factors on the multidimensional index are multiplicative [47] and X ~lognormal distribution implies log(X) ~ normal distribution and vice versa [48].
  2. Each index can be multiplied by 100 to maintain parity of composite indices.

Properties:

  • Each of SESt and ICVD-RFt is monotonically increasing continuous variable, with significant reduction of substitution effect among the chosen indicators.
  • Unit-free indices SESt, ICVD-RFt satisfies translation invariance property and aggregation consistency.
  • 1% increase in XitX1i0  ⟹ 1% increase in ICVD-RFt and 1% increase in YjtYj0 ⟹1% increase in SESt when all others are unchanged. Thus, ∆ICVD-RFt and ∆XitX1i0 curve is linear. The same is true for SESt
  • The indices can be computed for disadvantaged groups, regions, genders, urban & rural, etc. satisfying aggregation consistency since:

SEStCountry=SEStRegions =SEStDimensions=SEStindicatorssand

ICVD-RFtCuntry= ICVD-RFtRegions=  ICVD-RFtIndicators

  • Each proposed index satisfy Time–reversal test since SEStSES(t-1) ×SES(t-1)SESt=1 and  ICVD-RFtICVD-RF(t-1) ×ICVD-RF(t-1)ICVD-RFt=1.
  • Chain indices can be formed since ICVD-RF20=ICVD-RF21× ICVD-RF10 
  • Partial derivatives of logSESt with respect to logYjt i.e. δ logSEStδlogYjt.  gives relative importance of j-th indicator of SESt. Similarly, relative importance of the CVD-risk factors can be derived.
  • Based on relative importance, indicators of CVD-risk factors and SES-indicators can be ranked. The indicators getting high ranks can be priotarised for targeted interventions to have significant positive impact in short run.
  • SEStSES(t-1)>1 implies improvement of SES in successive periods. Responsiveness of SES measurement and extent of improvement is reflected by  SESt- SES(t-1)SES(t-1) * 100. Similarly, percentage improvement of index of CVD-risk factors can be measured across years.
  • Plot of SEStSES(t-1) or ICVD-RFtICVD-RF(t-1)  for different years indicate fluctuations of the indices across time and facilitate comparisons with longitudinal data.
  • The i-th indicator becomes critical if  ICVD-RFitICVD-RFit-1<1 for  ICVD-RFtor if 

 SESAAitSESAAit-1<1 for SESAAt. Such critical indicators provide directions for improvements and require appropriate interventions.

  • The indices can also be computed separately for different genders and different age-groups and also based on income, ethnicity, disability, geographic location, or other characteristics.

Association between ICVD-RFtand SESAA can be computed by correlation or by value of R2 obtained from multivariate regression of the form ICVD-RFt=α+βiSESit. However R2 dependents on model and is sensitive to the choice of response variable.  At micro level, canonical correlation between SES (with q-indicators of possible factors) and CVRFs (containing p-indicators of likely effects) indicates association of lifestyle behaviours with CVRFs and mortality. [49]. Both multivariate regression and CCA can be used to predict ICVD-RFt with knowledge of SESt by establishing relationships among them. In addition, relationships can be found between CVD prevalence or CVD death rate and gender-wise, age-group-wise indices ICVD-RFtand/or SESt along with relationship between the two measures.

Cautions need to be taken while fitting linear multivariate regression equation and undertaking CCA in terms of verification of the assumptions of the respective techniques (like linear relationships, absence of multicollinearity, uniform variability, normality, etc.)

and removing bad leverage points [50]. In linear regression equation of the form Y=α+βX, if ri=Yi-(α+βXi) is an outlier in the set of residuals {r1, r2, …….., rn} and the corresponding (Xi,Yi) is a leverage point, then the point (Xi,Yi) is a bad leverage point which result in a poor fit of the linear model.  Similarly, multivariate normality is needed for CCA. Interpretation of different canonical functions giving different patterns of relationship between CVRFs and SES needs caution.

Knowledge of distribution of each proposed index facilitates:

  1. Statistical inferences like estimation of population parameters such as mean (μ), variance σ2 and testing hypothesis of equality of mean and variance across time and space. Equality of treatment effectiveness can be tested byH0: μICVD-RFPre-group=  μICVD-RFPost-group using paired t-test. 
  2. Possible to test H0: ICVD-RFitICVD-RFit-1=0  or H0: SESAAitSESAAit-1=0 since distribution of ratio of two variables, each following normally is χ2.
  3. Davies-Bouldin Index (DBI) indicating efficiency of classification based on average similarity between classes and dissimilarities between classes [51] is calculated as

DBIK = 1Ki=1Kj=1  i≠jKMax[DiamCi-DiamCjCi-Cj]   

where K denotes number of mutually exclusive classes. Diameter of i-th class with mean Ci and size ni is computed as DiamCi=xi∈ Cixi-Ci2ni

Better classification efficiency is indicated by lower value of DBI. Use of DBI in simulated and also real-world medical datasets was investigated [52] who suggested method to recognize subtle patterns of DBI that might otherwise be overlooked.

Discussion:

Proposed multiplicative aggregation of subjective and objective measures of SES and CVRFs, irrespective of their inter-correlations to get composite indices SESt and ICVD-RFt reflecting overall improvement in a given period over the base period has theoretical advantages. The indices cover data in any units and forms such as count data, ordinal scores, cardinal data in proportion, percentages, ratios, rates, etc. avoiding normalization and weights of the chosen indicators. For each index, substitutability among indicators is reduced significantly. Knowledge of probability distribution of each proposed index helps to undertake testing statistical hypothesis and to find association between ICVD-RFtand SESt by simple correlation or by R2 from multivariate regression of ICVD-RFt  on components of SESt or by canonical correlation analysis (CCA).  Both multivariate regression and CCA can be used to predict ICVD-RFt with knowledge of SESt to help achieving equitable and effective disease prevention.  However, it may be desirable to evaluate the indices and find their association for a particular country since roles played by the components of SES and ICVD-RF vary for different locations and cultures and form regional clustering [53]. 

Conclusion:

The proposed indices with wide coverage of data in any forms satisfy aggregation consistency and offer significant benefits to find association between ICVD-RFtand SESt in better fashion.  Empirically fitted regression equation of CVD death rate on the components of proposed two indices facilitates making sound decisions regarding identification of critical areas to plan appropriate interventions and direction of improvement providing support in clinical decision‑making and public‑health policies towards mitigation of CVD-burden and to achieve equitable and effective disease prevention. Novelty of the method lies in known probability distribution of proposed indices facilitating parametric analysis for group comparisons, classifications and enabling statistical tests of equality of mean, variance progress made across groups or time periods. The method with theoretical advantages is recommended. Future empirical investigations are suggested for finding salient features of the proposed approach including robustness for generalization of findings.

Declarations:

Acknowledgement: Nil

Conflict of Interests: No conflict of interests to declare

Funding: No funds, grants, or other support was received

Data availability: No datasets were generated or analysed in the study.

Use of artificial intelligence (AI): The author declares that no generative artificial intelligence (AI) tools were used in the preparation, analysis, or writing of this manuscript

Authors’ contribution: The single author is involved in Conceptualization, Methodology,

Writing- Original draft preparation, Writing- Reviewing and Editing

References

  1. Gupta V., Hemraj SR., Voddiraju M., Badugu S., Sharma N., et al. (2025). Incidence, Prevalence and Impact of Socioeconomic Factors on Cardiovascular Disease: A Clinical Study. European Journal of Cardiovascular Medicine, 15, 951-955. DOI: 10.61336/ejcm/25-05-181
  2. Wang P, Gao X, Willett WC, Giovannucci EL. (2024). Socioeconomic Status, Diet, and Behavioral Factors and Cardiometabolic Diseases and Mortality. JAMA Netw Open. 2; 7(12):e2451837. DOI: 10.1001/jamanetworkopen.2024.51837
  3. Schultz WM, Kelli HM, Lisko JC, Varghese T, Shen J, et al. (2018). Socioeconomic Status and Cardiovascular Outcomes: Challenges and Interventions. Circulation.137 (20):2166-2178.DOI: 10.1161/CIRCULATIONAHA.117.029652.
  4. Zhou S, Liang Z, Li Q, Tian W, Song S, et al. (2024). Individual and area-level socioeconomic status, Life's Simple 7, and comorbid cardiovascular disease and cancer: a prospective analysis of the UK Biobank cohort. Public Health. 234:178-186. DOI: 10.1016/j.puhe.2024.06.028.
  5. Wang Tao, Li Y and Zheng X (2024). Association of socioeconomic status with cardiovascular disease and cardiovascular risk factors: a systematic review and meta analysis. Journal of Public Health, 32:385–399 https://doi.org/10.1007/s10389-023-01825-4
  6. Sadabadi, F., Talkhi, N., Omouri-Kharashtomi, M., Mirzaei, M., et al. (2025). Association Between Socio-Economic Status (SES) and the Traditional Risk Factors of Cardiovascular Diseases (CVD): A Cross-Sectional MASHAD Cohort Study Results. Health Science Reports, 8. https://doi.org/10.1002/hsr2.70721
  7. Dandona L, Dandona R, Kumar GA, Shukla DK, Paul VK, et al. (2017). Nations within a nation: variations in epidemiological transition across the states of India, 1990–2016 in the global burden of disease study. Lancet. 390(10111):2437–2460. DOI: 10.1016/S0140-6736(17)32804-0
  8. Bloom DE., Cafiero-Fonseca ET., Candeias V, Adashi E., Bloom L., et al. (2014). Economics of Non-Communicable Diseases in India: The Costs and Returns on Investment of Interventions to Promote Healthy Living and Prevent, Treat, and Manage NCDs. World Economic Forum, Harvard School of Public Health, 2014. http://www.weforum.org/issues/healthy-living
  9. Hertzman C, Power C, Matthews S, Manor O. (2001). Using an interactive framework of society and life course to explain self-rated health in early adulthood. Soc Sci Med. 53(12):1575–85. DOI: 10.1016/s0277-9536(00)00437-8.
  10. Smith GD. (2007). Life-course approaches to inequalities in adult chronic disease risk. Proc Nutr Soc. 66(2):216-36. DOI: 10.1017/S0029665107005460.
  11. Arnett DK, Blumenthal RS, Albert MA, Buroker AB, Goldberger ZD, et al. (2019). ACC/AHA Guideline on the primary prevention of Cardiovascular Disease: a report of the American College of Cardiology/American Heart Association Task Force on clinical Practice Guidelines. Circulation, 140, e596- e646,
  12. 10.1161/CIR.0000000000000678
  13. Chakrabartty, SN. (2025). Aggregation of Biomarkers and Patient Reported Outcomes. Global J Appl. Sci. Technol. 7(3): 194. DOI: https://doi.org/10.36266/GJAST/194.
  14. Betts MB, Milev S, Hoog M, et al. (2019). Comparison of recommendations and use of cardiovascular risk equations by health technology assessment agencies and clinical guidelines. Value in Health, vol. 22, no. 2, 210–219.
  15. Ozpelit ME, Colak A, Uzumcu HI, Kumral Z, Ozpelit E. (2024). Correlation and Agreement Between the SCORE2 and PREVENT 10-Year Atherosclerotic Cardiovascular Disease Risk Scores: Insights from Coronary Computed Tomography Angiography. Diagnostics (Basel). 14(23):2625. DOI: 10.3390/diagnostics14232625.
  16. Khan SS, Matsushita K, Sang Y, S.H. Ballew SH, Grams ME, et al. (2024). Development and validation of the American Heart Association's PREVENT equations. Circulation, 149, 430-449, 10.1161/CIRCULATIONAHA.123.067626
  17. Selvam PV, Sharma R, Ganz P, Blumenthal RS, Gulati M. (2025). Cardiovascular disease risk estimates using the new PREVENT Equation: The good, bad, and the ugly. Am J Prev Cardiol. 24: DOI: 10.1016/j.ajpc.2025.101288.
  18. Agarwal A. (2008). Social classification: the need to update in the present scenario. Indian J Community Med. 33(1):50–51.
  19. Daly MC, Duncan GJ, McDonough P, Williams DR. (2002). Optimal indicators of socioeconomic status for health research. Am J Public Health. 92(7):1151–1157.
  20. Galobardes B, Shaw M, Lawlor DA, Lynch JW, Davey Smith G. (2006). Indicators of socioeconomic position (part 1). J Epidemiol Community Health. 60(1):7-12. DOI: 10.1136/jech.2004.023531.
  21. Majumdar S. (2021). Socioeconomic status scales: Revised Kuppuswamy, BG Prasad, and Udai Pareekh’s scale updated for 2021. J Family Med Primary Care. 10(11):3964–7.DOI: 10.4103/ jfmpc.jfmpc_600_21
  22. Government of India. (2023). Primary census abstract indicators search up to town/village level. https://censusindia.gov.in/census.website/data/data-visualizations/PopulationSearch_PCA_Indicators
  23. Talha I, Elkhoudri N, Hilali A. (2024). Major Limitations of Cardiovascular Risk Scores. Cardiovasc Ther. 4133365. DOI: 10.1155/2024/4133365.
  24. Sofogianni A., Stalikas N., Antza C., Tziomalos K. (2022).  Cardiovascular risk prediction models and scores in the era of personalized medicine. Journal of Personalized Medicine, 12(7), DOI: 10.3390/jpm12071180.
  25. Wu Y, Liu X, Li  X, et al. (2006). Estimation of 10-Year Risk of Fatal and Nonfatal Ischemic Cardiovascular Diseases in Chinese Adults. Circulation, 114(21), 2217–2225
  26. Voss R, Cullen P, Schulte H, and Assmann G. (2002). Prediction of risk of coronary events in middle-aged men in the Prospective Cardiovascular Munster Study (PROCAM) using neural networks. International Journal of Epidemiology, 31 (6), 1253–1262.
  27. Ridker PM, Buring JE, Rifai N, and Cook NR. (2007). Development and validation of improved algorithms for the assessment of global cardiovascular risk in women: the Reynolds risk score, JAMA, 297(6), 611–619. DOI: 10.1001/jama.297.6.611
  28. Studziński K., Tomasik T., Krzysztoń J., Jóźwiak J., Windak A. (2019). Effect of using cardiovascular risk scoring in routine risk assessment in primary prevention of cardiovascular disease: an overview of systematic reviews. BMC Cardiovascular Disorders. 19(1), 11. DOI: 10.1186/s12872-018-0990-2.
  29. Mortensen, M.B.; Tybjarg-Hansen, A.; Nordestgaard, B.G. (2022). Statin Eligibility for Primary Prevention of Cardiovascular Disease According to 2021 European Prevention Guidelines Compared with Other International Guidelines. JAMA Cardiol., 7, 836–843.
  30. Navickas, P., Glaveckaitė, S., Lukavičiūtė-Navickienė, L., Šatrauskienė, A., Baranauskas, A., et al.(2026). The Role of Cardiovascular Risk Prediction Model Selection in Primary Prevention: An Observational Study of Statin Eligibility Agreement Across Nine Scores in a Lithuanian Primary-Prevention Cohort. Medicina, 62(5), 979 https://doi.org/10.3390/medicina62050979
  31. Chakrabartty, Satyendra (2024). Association of Index of Cardiovascular risk factors with Socio-economic status index, Journal of Cardiovascular Investigations and Insights, 1(1). DOI: 10.9567/ISSN.2024/WSJ.95
  32. Migeot J., Calivar M., Granchetti, H., Ibáñez, A., & Fittipaldi, S. (2022). Socioeconomic status impacts cognitive and socio-emotional processes in healthy ageing. Scientific reports, 12(1), 6048. https://doi.org/10.1038/s41598-022-09580-4
  33. Judd N, Sauce B, Wiedenhoeft J, Tromp J, Chaarani B, et al.(2020). Cognitive and brain development is independently influenced by socioeconomic status and polygenic scores for educational attainment. Proceedings of the National Academy of Sciences, 117(22), 12411-12418. https://doi.org/10.1073/pnas.2001228117
  34. Buettner, D., Nelson, T. & Veenhoven, R. (2020). Ways to Greater Happiness: A Delphi Study. J Happiness Stud 21, 2789–2806. DOI: 10.1007/s10902-019-00199-3
  35. Zung, WW. (1965). A self-rating depression scale. Arch Gen Psychiatry 12, 63-70.
  36. Azeez I.A., Ladiipo M.M.A. and Ige O.M. (2016). Assessment of Socioeconomic Status and control of Asthma in Adults. Annals of Ibadan Postgraduate Medicine.14 (2), 85-91. PMID: 28337093
  37. Zhao, M., Huang, CC., Mendoza, M. et al. (2023). Subjective socioeconomic status: an alternative to objective socioeconomic status. BMC Med Res Methodol. 23, 73. https://doi.org/10.1186/s12874-023-01890-z
  38. Brawley Newlin, Alice M. and Bhattacherjee, Anol (2024). Social Science Research: Principles, Methods, and Practices, Gettysburg College Edition. Open Educational Resources. 17. https://cupola.gettysburg.edu/oer/17
  39. Kishore J, Kohli C, Kumar N, & Gupta N. (2017). Scales Used in India to Evaluate Socio-economic Status in Medical Research: Limitations of Existing Scales and the Need for a More Comprehensive One. JIMSA, 30(2), 66.
  40. Fastame MC, Hitchcott PK. & Penna MP. (2017). Does social desirability influence psychological well-being: perceived physical health and religiosity of Italian elders? A developmental approach. Aging & Mental Health, 21(4), 348– 353. DOI: 10.1080/13607863.2015.1074162.
  41. Chakrabartty, SN. and Gupta, R. (2016). Test validity and number of response categories: a case of bullying scale. Journal of the Indian Academy of Applied Psychology, 4(2), 343-352
  42. Sacre H, Haddad C, Hajj A, Zeenny RM, Akel M & Salameh P.(2023). Development and validation of the Socioeconomic Status Composite Scale (SES-C). BMC Public Health, 23, 1619.DOI:10.1186/s12889-023-16531-9
  43. Prawitz AD, Garman ET, Sorhaindo B, O’Neill B, Kim J, Drentea P. (2006).  In charge Financial Distress/Financial Well-Being Scale: Development, Administration, and Score Interpretation. Financial Counseling Planning. 17:34–50. DOI: 10.1037/t60365-000.
  44. Psaki SR, Seidman JC, Miller M. Gottlieb M, Bhutta ZA, et al. (2014). Measuring socioeconomic status in multicountry studies: results from the eight-country MAL-ED study. Popul Health Metrics 12(1), 8. DOI: 10.1186/1478-7954-12-8.
  45. Mariani F. and Ciommi M. (2022). Aggregating Composite Indicators through the Geometric Mean: A Penalization Approach. Computation 10(4):64. DOI: 10.3390/computation10040064
  46. Muhammad Tareq, Assim Ibrahim Abdel-Razzaq, Md Arafat Rahman, Tonmoy Choudhury (2021). Comparison of weighted and unweighted methods of wealth indices for assessing SOCIO-ECONOMIC status. Heliyon,7 (2). DOI:10.1016/j.heliyon.2021.e06163
  47. Oka M. (2021). Interpreting a standardized and normalized measure of neighbourhood socioeconomic status for a better understanding of health differences. Arch Public Health 79, 226.  https://doi.org/10.1186/s13690-021-00750-w
  48. Harvey J. Motulsky, Trajen Head, Paul B.S. Clarke (2025). Analyzing lognormal data: A nonmathematical practical guide. Pharmacological Reviews, 77 (3). DOI: 10.1016/j.pharmr.2025.100049.
  49. Alf,  E. F. and Grossberg, J. M. (1979). The geometric mean: Confidence limits and significance tests, Perception &Psychophysics, 26 (5), 419-421. https://doi.org/10.3758/BF03204171
  50. Menotti, A., & Puddu, PE. (2024). Canonical Correlation for the Analysis of Lifestyle Behaviors versus Cardiovascular Risk Factors and the Prediction of Cardiovascular Mortality: A Population Study. Hearts, 5 (1), 29-44. https://doi.org/10.3390/hearts5010003
  51. Chakrabartty S.N. and Chakrabarty Anish (2025). Transformation to achieve perfect correlation. SCOPUA Journal of Applied Statistical Research, Vol.1, Issue 3,
  52. DOI: https://doi.org/10.64060/JASR.v1i3.4
  53. Davies, D and Bouldin, W. (1979). A cluster separation measure, IEEE Transactions on Pattern Analysis and Machine Intelligence, 1 (2), 224 – 227. https://doi.org/10.1109/TPAMI.1979.4766909.
  54. Anfossi, A. and Chicco, D. (2026). An easy guide to the Davies-Bouldin index for unsupervised internal clustering evaluation. Discover Computing, 29:212.
  55.  https://doi.org/10.1007/s10791-026-10094-0
  56. Schultz WM, Kelli HM, Lisko JC, Varghese T, Shen J, et al. (2018). Socioeconomic Status and Cardiovascular Outcomes: Challenges and Interventions. Circulation.137 (20):2166-2178.DOI: 10.1161/CIRCULATIONAHA.117.029652.