Google trend analysis of climatic zone based Indian severe seasonal sensitive population

Abstract

Background: Our earlier Google Trend (GT) Analytics study reported that the worldwide human population severely subject to four seasonal (sensitive) comorbid lifestyle diseases (SCLD) such as asthma, obesity, hypertension and fibrosis. The human population subject to seasonal variability in these four diseases activity referred as "severe seasonal sensitive population". In India, the estimated burden of these four seasonal diseases is more than 350 million as on the year 2018. It is a growing crisis for India with a projected disease burden of 500 million in the year 2025. This study was aimed to decipher the genuine SCLD seasonal trends in the entire Indian population using GT and validate these trends in Indian climatic zones.

Methods: GT is used to study the temporal trends in web search using weekly Relative Search Volume (RSV) for the period 2004 to 2017. The relative search volume (RSV) of the four-severe seasonal comorbid diseases namely Asthma, Hypertension, Obesity and Fibrosis were collected with and without obesity as the reference. The RSV were collected using the GT selection options as (i) Whole India (ii) Jammu and Kashmir (Cold zone) (iii) Rajasthan (Hot and Dry zone) (iii) West Bengal (Hot and Humid zone) and (iv) Uttar Pradesh state (Composite zone). The time series analysis was carried out to find seasonal patterns, comorbidity, trends and periodicity in the entire India and four of its states (zones).

Results: Our analysis of entire India (2004-2017) revealed high significant seasonal patterns and comorbidity in all the four diseases of SCLD. The positive tau values indicated strong positive seasonal trends in the SCLD throughout the period (Table). The auto correlation analysis revealed that these diseases were subjected to 3, 4 and 6 months period seasonal variations. Similar seasonal patterns and trends were also observed in all the four Indian temperature zones. Overall study indicated that SCLD seasonal search patterns and trends are highly conserved in India even in drastic Indian climatic zones.

Conclusions: The clinical outcome arise out of these observations could be of immense significance in handling the major chronic life style diseases asthma, hypertension, obesity and fibrosis. The possible strong comorbid relationship among asthma, hypertension, obesity and fibrosis may be useful to segregate Indian seasonal sensitive population. In disease activity-based chronotherapy, the search interest of segment of the population with access to Internet may be used as an indicator for public health sectors in the early detection of SCLD from a specific country or a region. As this disease population could be highly subject to the adverse effect of seasons in addition to life style and other environmental factors. Our study necessitates that these Indian populations need special attention from the Indian health care sectors.

Keywords: Comorbid; Google trends; Seasonal sensitive population.

Fig. 1

Entire India 4 weeks (monthly) moving average of weekly RSV (without benchmark) for SCLD from 2004 to 2017 were shown in four different colours. The overall average of the SCLD was shown in black colour (bold). Please note the positive (increasing) trends in the RSV from 2004 to 2017 with sharp hikes after the year 2010

Fig. 2

Seasonal and trend decomposition using TBATS for the SCLD for the weekly average (4 weeks) RSV from 2004 to 2017 without benchmark disease. Four weeks averaged weekly data were displayed for the SCLD in the top panels as observed (trend), level, slope and seasonal components 1, 2, and 3. Please note that strong seasonal patterns in all the SCLD except hypertension viz. 3 months (season 1), 4 months (season 2) and 6 months (season 3)

Fig. 3

Autocorrelation of SCLD diseases for the 4 weeks averaged weekly RSV dataset from 2004 to 2017 without benchmark disease. Observed data were showing strong cyclic patterns in autocorrelation above significant line (dotted) except hypertension (weak) in the entire period

Fig. 4

Entire India seasonal monthly RSV with benchmark were shown in thin grey lines and their corresponding seasonal moving average (4 months) from 2004 to 2017 were shown in four different colours labelled as MA in the brackets. Please note that sudden drop in the RSV with benchmark after 2010 was indicated to divide the period into I and II (dotted vertical line)