A cross-sectional study was conducted to study the critical differences in demographic and clinical characteristics of the SARS-CoV-2 B.1.617.2 (Delta) variant infections compared with those caused by the WT B.1 strain (Wuhan strain with the D614G mutation) in the Indian population. Genomic sequencing of the virus strain isolated from patient samples is a standard method for confirmatory diagnosis of COVID-19 and identification of the causative SARS-CoV-2 variant. We analyzed the patient metadata information attached to the genomic sequence reports for the SARS-CoV-2 Delta and WT (B.1) strains detected in the patient samples from the Indian population. The SARS-CoV-2 genomic sequence reports from India (patient sample collection date no later than July 31, 2021) were accessed from the EpiCoV database of the Global Initiative on Sharing All Influenza Data (GISAID) [24] using the automatic search function for information by geographical location, SARS-CoV-2 lineage, and collection dates. A similar search strategy was used to retrieve comparative metadata for the COVID-19 cases with the WT B.1 strain reported on GISAID since the first COVID-19 wave. The metadata files of the Delta variant (test group) and WT (control group) strain sequences were downloaded from the individual GISAID accession numbers, checked and confirmed for accuracy of the lineage information, and screened for demographic details (age and sex) and vaccination status (for 2 complete doses). Further, the clinical outcomes (information about the illness severity and mortality during the disease course) of the WT and Delta variant strains cases were assessed using a repeat search entering “patient status” as additional input. The collected data were filtered by discarding the sequence reports containing no or incomplete information and duplications (repeat sequencing from the same individuals). To prevent selection bias in the data collection, no specific filters for age, sex, date of collection, and geographical location were used during the search. A team of 3 investigators rechecked and verified the sampling and data entry errors of the complete data set by matching the original details in the metadata files.

The sample size for the study was guided by serosurvey reports from the Indian population during successive waves of the pandemic [25,26]. Of note, nationwide seroprevalences of 24.1% and 67.6% were reported in the studies conducted at the end of the first and second waves, respectively [25,26]. Considering a minimum of 24.1% seroprevalence in the current population of India (approximately 1.35 billion) at 5% relative precision (1.25% absolute precision) and 95% CI, a minimum sample size of 4866 was determined for the study.

Approval from the institute ethics committee was precluded, as the study used open-access data from the EpiCoVTM database from the GISAID [24]. GISAID permits open access to data to all individuals who agree to identify themselves and follow the GISAID sharing mechanism governed by its Database Access Agreement [27].

We assessed the genomic sequence reports for 8269 and 3767 cases with the Delta variant and WT B.1 strains, respectively, that were uploaded on GISAID from the first case to July 31, 2021. The SARS-CoV-2 genomic sequences were noted from nearly all the Indian states and union territories (Figure S1 in Multimedia Appendix 1). After filtering the data, 6238 Delta variant and 3262 WT B.1 strain cases were available for demographic analysis.

Among the screened cases, there were 24 Delta infections following 2 complete vaccine doses; however, no postvaccination infections were noted with B.1. In the database, 659 and 320 sequence reports with information on “patient status” were present for Delta and B.1 strains, respectively, and after filtering the data, 647 and 276 cases, respectively, were available for analysis regarding illness severity and mortality in terms of age and sex.

Frequent incidences of Delta variant as well as B.1 strain infections (N=6238) were noted across the age groups, including in the young group (0-19 years; Figure 1). The prevalence of infections was higher with the Delta variant than with the B.1 strain in the young group (0-19 years), with the highest difference in prevalence in those aged 0 years to 9 years followed by those aged 10 years to 19 years (P<.001). Conversely, the differences in prevalences were relatively low in those aged 20 years to 60 years, with a few exceptions (P<.001; Figure 1). Although a greater proportion of men were infected with both variants (both sexes: N_Delta=6238; N_WT =3262) than women (Delta: 3681/6238, 59% vs 2557/6238, 41%; WT: 2088/3262, 64% vs 1174/3262, 36%), the proportion of women contracting infection was higher with Delta than with B.1 (Delta: 2557/6238, 41% vs B.1: 1174/3262, 36%; P<.001). The mean ages when infected with the Delta variant were 37.9 (SD 17.2) years for men and 36.6 (SD 17.6) years for women and when infected with the B.1 strain were 39.6 (SD 16.9) years for men and 40.1 (SD 17.4) years for women (P<.001).

The cases were noted for all categories for the Delta variant as well as B.1 (asymptomatic/mildly symptomatic: N_Delta=72, N_WT=24; symptomatic/hospitalized: N_Delta=262, N_WT=130; living but symptom status not known: N_Delta=228, N_WT=102; demised: N_Delta =85, N_WT=20). The demographic (age groups and sex) distributions by patient status are shown in Figure S2 in Multimedia Appendix 1.

Of all the screened cases with the Delta variant or B.1 strain, Delta infections following 2 complete doses of vaccines did occur (24/6238: men=12, women=12); however, no postvaccination infections were noted with B.1 (0/3262). Postvaccination infections with the Delta variant were reported across the age groups (except for those aged 0-9 years or 10-19 years, which is explained by the fact that COVID-19 vaccines were not administered to individuals younger than 18 years in India until the period of this study) and for both sexes (Figure 4). The mean ages at postvaccination infection were 44.48 (SD 16.17) years for all cases, 40.4 (SD 11.9) years for men, and 47.8 (SD 20.4) years for women. No significant differences were noted in the frequency of postvaccination infections in terms of age groups and sex. Of all postvaccination infections, 88% (21/24) were with Covishield [28], and the rest (3/24, 12%) were with Covaxin [29].

Our analyses revealed essential observations indicating a shift in clinical and epidemiological characteristics of the COVID-19 pandemic from the WT strain B.1 to the Delta variant. With the Delta variant, more young individuals (<20 years) and adult women contracted the infection. In addition, an increased proportion of the young (<20 years) and adult population developed symptoms and required hospitalization. The mean age at COVID-19 infection and developing symptomatic illness or hospitalization was considerably lower. A disproportionate number of cases who developed symptomatic illness or were hospitalized was noted by sex: An increased proportion of young and adult women was affected. The mortality rate was about 1.8 times higher. Sex-specific contributions to the mortality data were lower for men but higher for women. However, the odds of death increased with age, irrespective of sex. Further, we noted multiple COVID-19 infections with the Delta variant following 2 complete vaccine doses.

Studies analyzing the demographic characteristics of the Delta variant in the Indian population are currently scarce. Recently, Kumar et al [30] presented their observations on the Indian population using data from the national clinical registry for COVID-19. The authors studied the clinical profiles of hospitalized patients during the second COVID-19 wave, mainly driven by the Delta variant [11,21]. Authors have also studied hospitalized cases in the first COVID-19 wave, primarily caused by WT strains [11,21]. However, the diagnosis of COVID-19 in cases included in this study was based on polymerase chain reaction test results, and no viral genomic sequencing was performed to ensure that the analyzed cases only represented those caused by the Delta variant or WT strains in the respective waves. The authors noted a decrease in the mean age of the total population who contracted infection and a reduction in the proportion of cases who were men during the second wave. The authors noted that mortality among hospitalized patients was 13.26% in the second wave, which was 3.1% higher than in the first wave. The increase in mortality was seen in all age groups in their study except for those aged <20 years, for whom mortality decreased [30]. We observed a similar trend in the susceptibility for contracting infection and risk of mortality in terms of age and sex, except we noted greater total mortality with the Delta variant (13.1%) than with the WT strain (7.2%). However, no deaths with either strain were noted in those aged <20 years in this study.

India is one of the world’s most populated countries, with a very high proportion of young people (approximately 45% are <20 years old) [31]. Our analyses show that, in comparison with B.1, Delta variant infections have occurred at considerably higher proportions in those <20 years old, particularly for those aged 0 years to 9 years (Delta: 4.5% vs B.1: 2.3%). The proportion of symptomatic cases and those requiring hospital admission was also higher in the young group (Delta: 10% vs B.1: 4%; Figure 2A). These findings indicate an age shift for the Delta variant, as COVID-19 infections with the WT strains were not common in young people and development of severe symptoms and the need for hospitalization were primarily limited to adults, more specifically older adults [17,18].

We observed that a higher proportion of women than men contracted the infection and developed symptoms or were hospitalized; in addition, mortality was lower in men but higher in women with the Delta variant than with the B.1 strain. Biologically, the severity of illness and mortality risks are considered lower for women than men for infectious diseases that affect mammals, including humans [32]. A higher vulnerability for contracting infection and symptomatic disease for women and narrowing the sex gap in mortality risk with the Delta variant present a paramount epidemiological concern.

Increased involvement of the younger age groups, primarily those aged 0 years to 19 years and of the female sex, in Delta variant infections compared with WT strains was indicated by studies and surveillance reports in the United Kingdom following a surge in Delta variant infections [8,33]. Despite the established immunological advantage for younger individuals [19,20,34] and women [16], as were reflected in infections with WT strains [16,18], increased risks of contracting infection and greater illness severity in these demographic groups indicate enhanced virulence and lethality with the Delta variant.

Last, the multiple Delta variant infections following 2 complete doses of COVID-19 vaccines in our study are supported by previous reports of vaccine breakthrough infections with this variant, including those from the Indian population [9,35].

The findings of this study and others describing demographic and clinical involvement in Delta variant infections confirm a shift from the established pattern for WT strains. Multiple factors could contribute to this epidemiological shift, most importantly, the variant’s intrinsic properties, such as increased transmissibility, virulence, and immune escape capabilities, as have been indicated by recent studies [7-9,11,30,36]. A sudden spurt in the number of cases during the second wave and consequent burden on the emergency health response system could have contributed to the increased mortality noted with the Delta variant. In contrast, the prioritized vaccination of those aged >45 years might have protected the vaccinated older adult population and contributed to lowering the mean age for COVID-19 involvement. On the other hand, in the initial period of the first COVID-19 wave, there was a shortage of diagnostic and COVID-19 care facilities and skilled health personnel in the country as well as limited knowledge about the prevention and therapeutic management of the pandemic. When combined, these limitations might have had an impact on the quality of the available epidemiological and clinical data related to WT strains.

Our study has multiple limitations that must be considered when interpreting the findings. First, our data for reporting postvaccination infections and patient status are limited; hence, related observations may require further validation from studies with a larger sample size. Second, the vaccination status data were unavailable in most of the accessed genomic sequence reports; hence, our data for postvaccination infections do not reflect the incidences of such cases in the total population. Our study does not rule out that actual incidences of postvaccination infections (after 2 complete doses) with the Delta variant could be higher. In addition, the noted 0 postvaccination infections with B.1 in our study could be an underestimation, as COVID-19 vaccines were yet to be made available to the masses in India during the first wave when the WT strains caused most of the infections.

Moreover, owing to the limited availability of vaccination status details for the screened cases in the studied database, we cannot comment conclusively on the efficacy of the vaccines noted in this study. Third, we have not considered the influence of confounding factors, such as comorbidities, vaccination status, and previous COVID-19 infection, on the clinical outcomes of the studied cases, which may have influenced the quality of the results. Last, although we applied no age, sex, nor region-specific filters when acquiring the data, a complete absence of selection bias cannot be assured when uploading the primary data on GISAID. However, any inherent bias in the primary data will likely be nondifferential and may not significantly impact the comparative analysis between the viral strains.

In conclusion, our study reveals greater proportions of infected young individuals and women, a lower mean age for illness, greater mortality, and frequent postvaccination infections with the Delta variant than with the WT (B.1) strain. Collectively, the findings of this study elaborate on the changing demographic characteristics of the COVID-19 pandemic with the emergence of the Delta variant. These findings are essential because we have only considered in the analyses those COVID-19 cases for whom Delta variant infections were confirmed by genomic sequencing, ensuring the accuracy of the presented data.