This study is sequential exploratory mixed method study with three strands [Table/Fig-1]. The first strand involves developing a questionnaire to measure the life course socioeconomic transition of individuals between the age group of 18-45 years. For this, we propose to start with in-depth interview among the residents of Kerala and the experts in social epidemiology, to define the construct of ‘life course socio-economic transition’ and develop a tool to measure it.

The approach used for sampling in this phase will be purposeful sampling procedure. The researcher will invite the individuals between the age group of 18-45 years among the community dwellers of Kerala. Researcher will include maximum individuals to get the maximum representation from the various districts of Kerala. After giving participant information sheet and getting the consent from the participant, researcher will fix the time and location of the interviews. The number of participants will be decided by data saturation. Define the construct LSET based on the available literature and in-depth interview among the experts in social epidemiology and community dwellers. Community dwellers will be the residents of Kerala from the birth itself and belongs to the age group of 18-45 years. This step is due to the absence of a well-accepted single definition of the construct in Indian context. Initial item pool will be created by the process of literature review and in-depth interview among the community dwellers. The in-depth interviews among community dwellers (residents of Kerala) will help to identify the possible markers of LSET and converting them to items which can then become part of the questionnaire.

Qualitative data will be analyzed using theme development procedure and data management conducted using the software of WEFT. We will record all interviews and audio files of the interview will be transcribed. Then the researcher will examine the content and extract the face and latent themes. Each transcribed words will be considered as a unit of analysis. The codes also help in the formation of themes. As the interviews progress and the relationships among the themes emerge it will help to identify the markers of LSET. Findings of this strand will help to define the construct ‘life course socioeconomic position’ and identify the markers of LSET. Initial item pool for the questionnaire will be used for the second phase or quantitative strand 2.

Initial item pool will be given to 10 experts (experts from public health, social epidemiology, sociology and economics) for selecting the most appropriate items on the basis of priority. Based on the scores obtained for each item, we will prioritize, and the highest ranked items will be selected for further step. We will use a 5-point Likert-type format for further part. Translation in to the local language Malayalam will be done, after which it will back translated into English by a second person, and the 2 English versions will be checked for consistency. Conflicts will be resolved by repeating this cycle. Each item in the translated and back-translated questionnaires will undergo strict verification and necessary corrections will be done until both versions became agreeably consistent.

We will examine the LSET-Q to assess its reliability. We will use two estimators of reliability (internal consistency reliability and test-retest reliability) in this study. Internal consistency will be measured by Cronbach’s alpha correlation coefficient. Cronbach’s alpha is equivalent to the average of possible split-half estimates.

Test-retest reliability is estimated by administering the same tool to the same sample on two different occasions on the assumption there will be no substantial change in the construct under study between the two sampling time points. A high correlation between the scores at the two time points indicates the instrument is stable over time. We will do the test-retest reliability of the LSET-Q by administrating the questionnaire to randomly selected 25 adults (between the age group of 18-45 years), from a diabetes outpatient clinic of Medical Trust Diabetes care hospital, Kulanada, Pathanamthitta. They will complete the LSET-Q on two different occasions; at baseline and eight weeks later.

II b 1) Content validity

Content validity indicates the content reflects a complete range of the attributes under study and is usually undertaken by seven or more experts. We will ask eight purposely chosen experts in the areas of social epidemiology, economics, epidemiology and Sociology to review the draft of LSET-Q to estimate its content validity. Each reviewer will independently rate the necessity and relevance of each item on the LSET-Q by using a 3 point rating scale (1=not necessary, 2= useful, but not essential, 3=essential) and a 4-point Likert scale (1=not relevant, 2=somewhat relevant, 3=relevant, 4=very relevant). We will use the Content Validity Index (CVI) to estimate the validity of the items.

It evaluates the appearance of the questionnaire in terms of feasibility, readability, consistency of style and formatting, and the clarity of the language used. We will develop an evaluation form, which will help respondents assess each question in terms of Clarity of the wording, Likelihood the target audience would be able to answer the questions and layout and style. We will randomly select 10 young adults from the outpatient clinics of Medical Trust diabetes care hospital and will complete the face validity form on a Likert scale of 1-4 (strongly disagree= 1, disagree= 2, agree= 3, and strongly agree= 4).

It refers to the degree to which the intended independent variable (construct) relates to the proxy independent variable (indicator). Factor analysis will be used to do the construct validity. The sampling population for factor analysis is (n =100) young adults (18-45 years) from the rural population of Kerala. We will recruit the samples using snowball sampling technique. We will use exploratory factor analysis (Direct oblimn method) for the factor analysis of LSET- Q.

The second quantitative strand is a hospital based matched case control study.

The hypothesis of the second quantitative strand of the study is ‘in comparison with individuals who remain free of early onset type 2 diabetes, those who develop this disease will have a low childhood SEP and high adult SEP (SEP transitions towards affluence in their life course socioeconomic position)’.

Medical trust Diabetes Care hospital (a rural diabetes care hospital situated near Pandalam), situated in the Central Kerala, which serves more than 3,00,000 population with a catchment area extending to Northern Kerala. The average number of diabetic patients approaching the outpatient department of the hospital is 200/day.

The source of cases for the case control study will be the type 2 diabetes patients, in the age group 18-45 years, who report to the hospital. We hope to recruit 210 cases between April 2015 and December 2015. We will recruit population based age, sex and frequency matched controls from a randomly selected Panchayath (rural (Hospital catchment area =15 kilometeres physical distance from the hospital by Phibbs & Robbinson’s method 1993) [16]. We will recruit one age and sex matched control for each incident case within 15 days of recruitment of the case, thus following an incident case control design. The sample size results in a power of 80% to detect an odds ratio (OR) of 2.0, assuming exposure prevalence of 25% among controls (National Sample Survey Organisation survey – 2014) [17]. We assume incomplete and missing information in 5% of sample pair. We explained the inclusion and exclusion criteria for the cases in [Table/Fig-2].

ASHA worker will approach the first eligible person in the electoral roll and rule out the history of intake of Oral Hypoglycaemic Agents (OHA) or any previous diagnosis of type 2 diabetes. After getting the consent of the potential control, ASHA worker will do the finger prick test for capillary blood glucose monitoring (CBG monitoring) for confirming their non diabetic status on the next convenient day of the participant. We explained the inclusion and exclusion criteria for the controls in the [Table/Fig-3].

Life course socioeconomic transition is the exposure variable.

1. SEP=low (childhood) and SEP =low (adulthood)

2. SEP = high (childhood) and SEP = high (adulthood)

3. SEP = low (childhood) and SEP= high (adulthood)

4. SEP = High (childhood) and SEP = low (adulthood)

We will club the first 2 categories: they have a stable SEP. Third and fourth categories clubbed together will form the group with the changed SEP. We will estimate the unadjusted and adjusted odds ratios for developing early-onset T2D and their 95% confidence intervals using conditional logistic regression to evaluate associations between individual exposures and the risk of T2D. Other attributes like diet, physical activity, tobacco use, alcohol consumption and body mass index will be considered in the analysis to identify confounding, interaction, or mediation effects. We will use SPSS (SPSS 20.0 for Windows, SPSS Inc., Chicago, Illinois, USA) for statistical analysis.

Ethical approval has been obtained from the ethics committee of the Sree Chitra Tirunal Institute for Medical Sciences & Technology (SCTIMST-IEC registration no:ECR/189/Inst/KL/2013). Informed consent will be obtained from the participants prior to the study for both controls and cases.

Cases: We will provide a participant information sheet and informed consent to those who are selected for the study. We will conduct the interview during the waiting time for physician’s visit so that they will not be biased by the physician’s advice.

Controls: Either the ASHA worker or principal investigator will provide the participant information sheet to the selected individuals in the community prior to the capillary blood glucose test. Data collector will visit the house in the next day and will ask about their willingness. We will collect the blood sample in the next convenient day of the potential controls.

Excluded potential controls: Excluded potential controls (who have no h/o diagnosis of type 2 diabetes, but their FBG > 100 mg/dl) will get individual counseling on life style modification to control their blood glucose level.

Data management: Only the principal investigator will have access to the data. Research proposal has been approved by Medical Trust Diabetes Care Hospital. We will maintain the confidentiality of the data in every step of the study.

1. Questionnaire

We will use developed and validated LSEP- questionnaire (strand 1&2) to measure the life course socioeconomic transition of both cases and controls. We will use the medical records to collect the data on sex, age, Fasting Blood Sugar (FBS) and PostPrandial Blood Sugar (PPBS) of the cases

2(a) Blood glucose monitoring for the potential controls

ASHA worker will ask for the h/o oral hypoglycaemic agents intake or diagnosis of T2D to the potential controls. Capillary Blood Glucose (CBG) monitoring is used as a screening tool to identify individual’s normoglycaemia state. We will use Accu-Check Aviva, Roche Diagnostics, for the blood glucose monitoring. Device has a capacity of 200 readings/device. We will purchase all devices and strips from local regular distribution sources (pharmacies, 25 strip lots per device). We will do fasting blood glucose test by using finger prick test of blood glucose monitoring. Blood samples will be obtained after a minimum fasting time of 8 hours. They will wash their hands prior to performing the measurements. We will take precautions to avoid capillary milking. Devices and test strips were examined by trained technician. Equipment calibrated based on the manufactures recommendations. We will strictly follow the calibration and quality control measures as recommended by the supplier of the equipment.

CBG monitoring is less expensive and a good performing screening tool in population based interventions. Venous and capillary samples will give the same result in the fasting state. But in the non-fasting state capillary will give higher results than venous samples. The current generation of meters and strips are quick to use and require a very small volume of blood. The CBG determinations as screening tool and Venous plasma glucose (VPG) as reference with a cut-off point of 7 mmol/L (126 mg/dl), resulted in a sensitivity of 81.4% and specificity of 97.8% [20].

CBG monitoring is less expensive and a good performing screening tool in population based interventions. Venous and capillary samples will give the same result in the fasting state. We will perform Venous Plasma Glucose (VPG) test for the purpose of validation of Capillary Blood Glucose (CBG) monitoring in 10% of the total controls. We will follow standard laboratory procedures for all blood tests in the field.

According to literature, this is the first attempt to develop a tool for measuring the socioeconomic transition in public health research in Kerala. There is significant change in the socioeconomic status of people living in Kerala over the past 25-30 years. This transition is more relevant in rural areas of Kerala state. This socioeconomic transition is a result of migration of people from Kerala to other countries and states for jobs and also. Land revolution and the increased literacy rate acted as catalyst for both migration and socioeconomic transition. Most of the rural Keralites are following so-called ‘periurbanized’ way of living and practicing conspicuous consumption. This in turn has resulted in a drastic change of life style among the rural population of Kerala. This should reflect in the incidence and prevalence of the non-communicable diseases in Kerala [13,14,21].

Most of the research studies use ‘Socio-Economic Status’ (SES) to represent the socioeconomic background of an individual. Change in the socioeconomic position from the childhood to adulthood (or later) is defined as socioeconomic transition. The dynamic nature of the socioeconomic conditions of an individual’s life course is well accepted in the definition of socioeconomic position. Krieger et al., defined ‘socioeconomic position as the social and economic factors that influence what positions individuals or groups hold within the structure of society’ [22]. This definition is accepting the multidimensional and dynamic nature of the individual’s socioeconomic condition. SEP is a period estimate of a socioeconomic condition of an individual. Socioeconomic status is representing the static nature of the socioeconomic condition of an individual. The static nature of socioeconomic condition is observed in either more economically deprived or super rich category. It is a point estimate of a socioeconomic condition of an individual. Most of Indian socioeconomic scales/indexes are considering SES as an assessment tool. Modified Kuppuswami scale is one of the tools to assess the socioeconomic status of individuals [23]. It might be relevant in research studies conducted in other Indian states. But it is less relevant in a state like Kerala with high HDI and GDP based on the NRI money influx. This study is accepting rapidly changing nature of socioeconomic condition of Kerala society and trying to develop a culturally relevant tool to measure the socioeconomic position of the individual. The variables in this tool may be adequate only for the particular population and time period. But still, some of the parameters to measure the socioeconomic position of the individual childhood and adulthood period will be helpful for making a platform for other tool development studies in future.

Life course studies are very important in non-communicable disease epidemiology. Most of the research studies have proved the effect of adverse life course events on the incidence of noncommunicable diseases. They are using the theoretical models like the critical period model, accumulation model of life course epidemiology. Social mobility model is more relevant in studies related to recently affluent community [5,6,24]. Social mobility frame work is being in this life course socioeconomic transition study. It is very important to conduct the life course studies in a state like Kerala with diabetes prevalence of 22% [25]. In our knowledge, this is the first attempt to make a tool for measuring the life course socioeconomic transition and check its association with early onset type 2 diabetes by using a social mobility framework.

The next step will involve the application of this tool in a hospital based case-control study to know the association between life course socioeconomic transition and early onset type 2 diabetes. The study uses the social mobility theoretical framework. Previous studies in this broad area have focused on static indicators such as socio-economic status or position to explain lifestyle factors which lead to diabetes. In a rapidly changing society such as Kerala, a dynamic indicator like the life-course socio-economic transition is a more appropriate and sensitive indicator to capture the markers of high risk. It is expected that this in turn, will generate insights into possible preventive strategies.

Small sample size for the validation survey is one of the limitations of this study. The validated questionnaire is going to be applied in a hospital based setting. The length of the questionnaire and the questions which are not related to their disease condition may cause lack of interest in diabetic patients.

The study will provide a substantial evidence for the association between life course socioeconomic transition and early onset type 2 diabetes and identifying a high-risk group among Keralites. It will help in developing effective interventions to reduce the incidence of early onset type 2 diabetes among the high-risk group. Application of this tool in other non-communicable diseases research is also a further scope of this research study.