1 INTRODUCTION

India is a rapidly developing country; this has led to a rapid change from reduction in communicable diseases to an increase in non-communicable diseases. Following this, cardiovascular diseases have become the number one cause of death, and responsible for a quarter of all mortality.1 This number is still increasing due to factors related to developing such as industrialization, urbanization, and changes in lifestyle, so-called epidemiological transition.2 A nationally representative study among 1.3 million Indians showed prevalence values of diabetes and hypertension of 7.5% and 25%, respectively.3 In addition, the prevalence of people with diabetes in India had increased from 26 million in 1990 to 65 million in 20164 with a prevalence of 12% in those older than 50 years of age.5 Consequently, cardiovascular diseases are responsible for a massive economic burden with estimated costs of $2.17 trillion between 2012 and 2030 being the major cause of the economic loss.6 However, despite these high figures, awareness of both diabetes7 and hypertension8 is still low among Indians.

Public screening activities may help to increase awareness of cardiovascular diseases that are often symptomless and may improve the insight into the prevalence of these diseases and its control. Following this, an initiative was undertaken to set up health care screening booths in eight airports and one hospital around India. The current paper presents the obtained data of more than 100 000 participants screened for cardiovascular risks.

2 METHODS 2.1 Study design and participants

The study design envisaged a retrospective analysis of a database created from a community-based screening campaign focusing on blood pressure (BP), blood glucose measurement and body mass index (BMI), and the collection of basic information.

Data were obtained in the period between August 2019 and March 2020. A visible temporary booth was established at 8 airports (Ahmedabad, Cochin, Coimbatore, Delhi Metro, Goa, Lucknow, Mumbai, and Trivandrum) and one hospital (KG Hospital Coimbatore).

Approval for this retrospective study was obtained from Ripon Independent Ethics Committee (RIEC), Chennai, India. A digital signature was obtained from each participant while data were electronically collected using a tablet. Data from the participants were collected in a non-identifiable manner to protect the privacy of all participants. The number of participants screened per state of India can be found in Figure Supplement S1.

2.2 Participants

The screened population consisted of an unselected sample of 101,982 participants. Since most of the screening was done at the airport, the visitors could come from anywhere. However, it appeared that only 0.8% (817) of all visitors were non-Indians. As the present analysis was aimed at improving the insight into the prevalence of cardiovascular diseases in India, non-Indians were excluded from the analysis for the present study. A table in which the characteristics of subjects from “elsewhere” were compared to those with residence in India are shown in the Table S1. In total, 100 107 Indian participants were analyzed for the present study (see Figure 1, flow chart).

Flow chart of the study

2.3 Booth visit

When entering the booth to be measured the following participant's characteristics were registered: sex, age, nationality, city of residence, profession, and flying frequency. In addition, participants were asked if hypertension or diabetes was diagnosed previously and in all except three booths (Cochin, Coimbatore, Delhi Metro), it was registered if participants received treatment for either one of these conditions.

2.4 Measurements 2.4.1 Blood pressure measurement

The booth attendants were well trained to take BP and blood glucose readings as per routinely followed protocols at clinic or lab. BP was measured with a validated oscillometric BP monitor that automatically measured 3 times in a row with 15-s interval times and provided the average of three BP readings (Circa 120/80, Watch BP Home A, Eris Ahmedabad).9-11 Patients were measured in a sitting position, with their back supported and the arm supported at heart level. Participants were instructed not to talk during BP measurement and keep their legs uncrossed with both feet flat on the floor. A cuff-size was selected appropriate to the participant's arm circumference (small size cuffs for upper arm circumferences between 14 and 22 cm, medium size 22- 32 cm, large size 32–42 cm, and extra-large size 32–52 cm).

2.4.2 Blood glucose measurement

After the BP measurement had been taken, blood glucose was determined using the Accu-Chek Performa blood glucose monitor (Roche, Germany). Participants were asked when they had last eaten. If the participant reported not to have eaten within the last 8 h their glucose value was considered as Fasting Blood Glucose (FBG), all other values were considered Random Blood Glucose (RBG).

Thereafter, participants were asked to take off their shoes for measuring their height (Thermocare, height measurement scale) and weight (Omron weighing scale) to assess BMI.

Results were discussed with the participant and appropriate advice followed the obtained measurements (for example, participants were recommended to visit their doctor if the values obtained fell outside the normality range). Finally, a printed report was handed to the participants or sent by SMS if this was preferred (see Figure S2).

2.5 Statistical analysis

Participants were classified as hypertensive if the average of three BP readings was equal to or higher than 140 mmHg and /or 90 mmHg for systolic and diastolic BP.12 Subsequently, BP values were categorized according to ESH/ESC -standards: (Optimal < 120 and < 80 mmHg, Normal 120–129 and 80–84 mmHg, High normal 130–139 mmHg and 85–89 mmHg Grade 1 hypertension: 140–159 and 90–99 mmHg, Grade 2 hypertension 160–179 and 100–109 mmHg, Grade 3 hypertension: ≥180 and ≥110 mmHg for systolic and diastolic blood pressure, respectively).12

For classifications in glucose ranges guidelines of the American Diabetes Association were applied. For FBG a level of ≥ 126 mg/dL and for RBG a value ≥ 200 mg/dL was labeled as diabetes.13 Weight was divided in 6 classes (underweight, BMI less than 18.5 kg/m2; normal, 18.5-24.9 kg/m2; overweight, 25.0 to 29.9 kg/m2; class I obesity 30–34.9 kg/m2, class II obesity 35–39.9 kg/m2, class III obesity 40 kg/m2 or higher). Prevalence for overweight (yes, no [≥ 25.0 kg/m2]) and obesity (yes, no [≥ 30.0 kg/m2]) were determined and used for comparison. Participants were divided in age groups of less than 30 years old with stepwise 10-years increases until 60 years and older.

Participant's characteristics were compared for sex and other aspects such as screening location (hospital vs airports), BP and weight class. Main demographic and clinical data were summarized by calculating the mean (±SD) in case of continuous variables and the absolute (n) and relative (%) frequency in case of categorical variables. Differences across groups were evaluated using retrospective analysis of variance (ANOVA). For categorical variables, differences across groups were evaluated using Chi-square tests. We calculated the prevalence of hypertension and diabetes for categories of age, BP (for diabetes prevalence), and BMI. To determine the statistical relationship between systolic BP and age and other measured parameters, Pearson's correlation coefficients (CCs) and confidence intervals were calculated. In order to verify if Pearson's CCs would differ between groups younger and older than 40 years of age, CCs were calculated for systolic BP and BMI and RBG and BMI and compared for significant differences. No methods for imputation were performed. Results were presented in p-values if considered relevant, a p value of < .01 was considered significant. Analyses were performed using the statistical package R Studio Version 1.2.5033 for Windows.

3 RESULTS

Data from 100 107 Indian visitors were collected with an average age of 45.9 ± 12.9 years. Among these, there were 8599 (8.6%) who did not undergo BP measurement and 15 203 (15.2%) participants had no blood glucose measurement.

Table 1 shows the unweighted characteristics of all booth visitors and separated for sex (17% female). In total, 30 345 (33.2%) had hypertension and 12 571 participants (14.8%) were classified as diabetes (388 [19.4%] based on FPG and 12 183 [14.7%] based on RBG). Weight and height measurements showed an average BMI of 26.7 ± 4.0 kg/m2, 61 219 (64.8%) participants were overweight, and 17 074 (18.1%) were obese.

TABLE 1. Subject characteristics separated for sex Female (No. = 17 321) Male (No. = 82 786) Total (No. = 100 107) p Age [y] < .001 45.3 (13.8) 46.0 (12.6) 45.9 (12.9) Systolic BP [mmHg] < .001 123.4 (18.7) 131.0 (16.6) 129.7 (17.2) Diastolic BP [mmHg] < .001 76.7 (10.7) 83.6 (10.6) 82.4 (11.0) Heart Rate [BPM] < .001 84.9 (12.8) 85.5 (13.6) 85.4 (13.5) Random Blood Glucose [mg/dL] < .001 (no. = 82 907) 139.1 (58.6) 147.9 (65.1) 146.5 (64.2) Fasting Plasma Glucose [mg/dL] .053 (no. = 1997) 108.0 (23.8) 110.4 (24.6) 110.0 (24.5) Height [cm] < .001 158.9 (6.4) 170.7 (7.2) 168.6 (8.3) Weight [kg] < .001 66.5 (12.0) 78.0 (12.5) 75.9 (13.2) Body Mass Index [kg/m2] < .001 26.3 (4.5) 26.7 (3.9) 26.7 (4.0) Overweight < .001 Non-Overweight 6688 (40.6%) 26 548 (34.0%) 33 236 (35.2%) Overweight 9780 (59.4%) 51 439 (66.0%) 61 219 (64.8%) Obesity < .001 Non-Obese 13 261 (80.5%) 64 120 (82.2%) 77 381 (81.9%) Obese 3207 (19.5%) 13 867 (17.8%) 17 074 (18.1%) Random Blood Glucose < .001 Diabetes 1463 (10.8%) 10 720 (15.5%) 12 183 (14.7%) Non-Diabetes 12138 (89.2%) 58 586 (84.5%) 70724 (85.3%) Fasting Plasma Glucose .088 Diabetes 54 (16.1%) 334 (20.1%) 388 (19.4%) Non-Diabetes 282 (83.9%) 1327 (79.9%) 1609 (80.6%) Blood pressure diagnosis < .001 Hypertension 3470 (21.9%) 26 875 (35.5%) 30 345 (33.2%) Normotension 12 399 (78.1%) 48 765 (64.5%) 61 164 (66.8%) Diabetes total < .001 DM 1517 (10.9%) 11 054 (15.6%) 12 571 (14.8%) non-DM 12 420 (89.1%) 59 913 (84.4%) 72 333 (85.2%) Tachycardia [≥100 BPM] < .001 Normal 13 829 (87.5%) 64 014 (85.0%) 77 843 (85.4%) Tachycardia 1967 (12.5%) 11 338 (15.0%) 13 305 (14.6%) Hypertension awareness < .001 No 16 405 (94.7%) 77 748 (93.9%) 94 153 (94.1%) Yes 916 (5.3%) 5038 (6.1%) 5954 (5.9%) Diabetes awareness < .001 No 16 360 (94.5%) 75 727 (91.5%) 92 087 (92.0%) Yes 961 (5.5%) 7059 (8.5%) 8020 (8.0%) Receiving treatment < .001 Hypertension 568 (3.3%) 2840 (3.4%) 3408 (3.4%) Diabetes 600 (3.5%) 4865 (5.9%) 5465 (5.5%) Both 361 (2.1%) 2252 (2.7%) 2613 (2.6%) None 15 792 (91.2%) 72 829 (88.0%) 88 621 (88.5%) Profession < .001 Business person 662 (4.5%) 19 444 (26.5%) 20 106 (22.8%) Corporate Job 5354 (36.1%) 45 648 (62.3%) 51 002 (57.9%) Education 214 (1.4%) 819 (1.1%) 1033 (1.2%) Government Service 611 (4.1%) 5278 (7.2%) 5889 (6.7%) Home Maker 7536 (50.8%) 1233 (1.7%) 8769 (10.0%) Student 458 (3.1%) 850 (1.2%) 1308 (1.5%) Flying Frequency < .001 Occasional 13 221 (92.6%) 55 731 (77.2%) 68 952 (79.7%) Once a month 813 (5.7%) 12 000 (16.6%) 12 813 (14.8%) Once a week 146 (1.0%) 2692 (3.7%) 2838 (3.3%) Multiple times a week 104 (0.7%) 1805 (2.5%) 1909 (2.2%)

Males were slightly older than females (46.0 ± 12.6 vs 45.3 ±13.8 years), had higher BMI (26.7 ±3.9 vs 26.3 ±4.5 kg/m2), higher RBG (147.9± 65.1 vs 139.1 ± 58.6 mg/dL), and higher systolic and diastolic BP values (131.0/ 83.6  vs 123.4/76.6 mmHg, all p < .001). In contradiction to the higher BMI for males, females showed higher obesity prevalence than males (19.5% vs 17.8% p < .001).

Of all participants 1,428 individuals were measured at a booth located at a hospital entrance hall, their characteristics slightly differed from the airport booth visitors (Table S2).

Table 2 shows that systolic BP had highest correlation values with age (r = 0.27), glucose (r = 0.17), and BMI (r = 0.16). FBG showed a stronger correlation with age (r = 0.29) and BMI (r = 0.16) than RBG (r = 0.22 and r = 0.09, respectively)

TABLE 2. Means, standard deviations, and correlations with confidence intervals for the parameters measured Variable M SD 1 2 3 4 5 6 1. Age (years) 45.87 12.90 2. BMI (kg/m2) 26.66 4.01 0.11 [0.11, 0.12] 3. Systole (mmHg) 129.65 17.22 0.27 0.16 [0.26, 0.28] [0.15, 0.16] 4. Diastole (mmHg) 82.36 10.97 0.04 0.17 0.69 [0.03, 0.04] [0.16, 0.17] [0.69, 0.70] 5. Heart rate (bpm) 85.39 13.47 -0.08 0.12 0.04 0.22 [-0.09, -0.08] [0.11, 0.12] [0.03, 0.05] [0.22, 0.23] 6. RBG (mg/dL) 146.37 64.08 0.22 0.09 0.17 0.07 0.22 [0.22, 0.23] [0.08, 0.09] [0.17, 0.18] [0.06, 0.08] [0.21, 0.23] 7. FBG (mg/dL) 109.96 24.47 0.29 0.16 0.17 0.14 0.16 NA [0.25, 0.33] [0.11, 0.20] [0.12, 0.21] [0.09, 0.18] [0.11, 0.21] Note. M indicates mean; SD, standard deviation; FBG, fasting blood glucose; RBG, random blood glucose; NA, not applicable. Values in square brackets indicate the 95% confidence interval for each correlation. All correlation values are significant at p < .01. 3.1 Treatment

Among participants whose measured blood glucose value suggested diabetes and who were asked for their treatment, 44% (n = 4131) reported to be treated for diabetes (Table 3). For those measured with elevated BP 11% (n = 2599) reported to receive antihypertensive treatment. Of the participants treated for diabetes 46% (n = 3574 out of n = 7705) had normal glucose values, whereas 56% (n = 3282 out of n = 5881) of those treated for hypertension measured normal BP values. Of the participants whose measured BP value classified them as having grade 3 hypertension (n = 1383), only 15.1% received anti-hypertensive treatment (Table S3).

TABLE 3. Participants treated and untreated for hypertension and diabetes Diabetes (No. = 12 571) non-Diabetes (No. = 72 333) Total (No. = 84 904) No.-Miss 3182 16 361 19 543 Treated 4131 (44.0%) 3574 (6.4%) 7705 (11.8%) Untreated 5258 (56.0%) 52 398 (93.6%) 57 656 (88.2%) Hypertension (No. = 30 345) Normotension (No. = 61 163) Total (No. = 91 508) No.-Miss 7334 12 257 19 591 Treated 2599 (11.3%) 3282 (6.7%) 5881 (8.2%) Untreated 20 412 (88.7%) 45 625 (93.3%) 66 037 (91.8%) No.-Miss indicates the number of missing values. 3.2 Prevalence values of hypertension and diabetes

Stratification by age-groups and BMI classes and separated for sex, showed a similar trend in hypertension prevalence for both males and females (Figure 2, top), although males showed higher values overall. Hypertension prevalence increased with higher BMI classes and showed a stepwise increase with each age-group. High BMI had a major effect on hypertension at younger age in both males and females, in the age groups younger than 40 years old the hypertension prevalence among obese subjects was about twice the prevalence as compared to those with normal BMI. However, this BMI effect seemed to become less with older age. This was confirmed with Pearson correlation analysis showing that systolic BP correlated stronger with BMI in the group younger than 40 years of age than in the group of 40 years and older (r = 0.27 vs r = 0.06, p < .0001, respectively)

Prevalence of hypertension and diabetes categorized for age groups and BMI class (top and middle picture) and prevalence of diabetes by age-group and blood pressure class (below), separated by sex. For each category the number of participants affected is presented (n) together with the percentage (%) calculated from the total number of participants per category

Among obese (BMI > 30 kg/m2) women aged 60 years and older more than 40% were hypertensive, whereas for obese men a similar hypertension prevalence value of 40% or more was already seen 3 decades earlier in the age-group of 30-40 years old.

A similar trend as for hypertension prevalence was seen for diabetes prevalence (Figure 2, middle). Diabetes showed a clear stepwise increase with each higher age-group. Similar as for hypertension the diabetes prevalence seemed to increase more with higher BMI classes at younger age than at older age. This was confirmed with Pearson correlation analysis that showed that RBG had a higher correlation with BMI for those younger than 40 years of age, as compared to those 40 years and older (r = 0.15 vs r = 0.03, p < .0001, respectively)

Categorization for BP classes and age groups showed a similar stepwise and predictive increasing trend with diabetes prevalence (Figure 2, bottom). Overall, males showed higher diabetes prevalence values than females, but this difference was not as large as was seen for hypertension. When aged older than 50 years and classed hypertension grade 1 or higher both males and females showed a diabetes prevalence of 20% or more.

4 DISCUSSION

The present retrospective analysis from data of 100 107 participants screened showed that one-third had hypertension, 15% had diabetes, and 18% of all participants were obese (65% being overweight). Among those with elevated BP and elevated blood glucose the awareness and treatment rates were low; 44% of the participants measured with elevated blood glucose values received anti-diabetes treatment and 11% of those with elevated BP received anti-hypertensive treatment. This suggests that the awareness of diabetes was higher than for hypertension.

Although significant due to the large sample size, correlation values between parameters in the present analysis were generally poor. However, there were some obvious and si