The geography of suicide in older adults in Hong Kong: An ecological study
Abstract
Objectives
The geography of suicide has been widely explored among the general population. However, little is known of the geographic variations in suicides among the older adults and their spatial correlates. This study aims to explore the spatial variations in the elderly suicide rates and their correlates in Hong Kong.
Methods
Bayesian hierarchical models have been used to estimate smoothed standardized mortality ratios (2006-2015) on suicide in people aged 65 years or older in each geographic unit in Hong Kong. Their associations with the Social Vulnerability Index and the accessibility of eight types of services (ie, recreational services, rehabilitation services, food services, daily necessity services, community services, and transportation services) were further analyzed.
Results
The results suggested that compared with the simple “inner-city high suicide rate and suburban low” pattern in the Western studies and the “central low suicide rate and peripheral high” pattern in the Asian studies, the spatial variations of elderly suicides in Hong Kong exhibit a much more complicated pattern. In Hong Kong, higher elderly suicide clusters were found in both the lower-density areas located in the New Territories and in some inner-city areas. The spatial variations of suicide in the older adults cannot be explained by the Social Vulnerability Index. Instead, service provision such as recreational services, daily necessity resources, and community centers played a more significant role in affecting suicides in the older adults.
Conclusions
Strengthening public services, providing more public spaces and activities, and making good use of the community resources might be key and efficient strategies in elderly suicide prevention in Hong Kong.
Key points
- The spatial variations of elderly suicides in Hong Kong show a much more complicated pattern compared with the simple “inner-city high suicide rate and suburban low” pattern in the Western countries and the “central low suicide rate and peripheral high” pattern in some of the Asian countries.
- In Hong Kong, suicide rates in the city centers were not higher than the average in the city. Clusters of higher suicide rates were mainly found in the New Territories, which is somewhat disconnected from the city and, in some inner-city neighborhoods, with high-density population.
- The spatial variations of suicide in the older adults in Hong Kong cannot be explained by the neighborhood Social Vulnerability Index as in the existing literature.
- Neighborhood service provision such as recreational services, daily necessity resources, and community centers played a significant role in affecting suicides in the older adults in Hong Kong.
1 INTRODUCTION
In most countries throughout the world, the suicide rates are consistently far higher in later life than in any other age groups.1 There are approximately four attempts in each completed suicide in later life.2 In contrast to the adolescent and general population, each suicide death is estimated to comprise around 200 and 25 suicide attempts, respectively.3, 4 These disparate findings illustrate the lethality of suicide in later life. Furthermore, elderly suicides deliver fewer warnings to others on their suicide plan yet execute the suicide methods with greater planning.5 This suggests that elderly suicide prevention may be more difficult than their younger counterparts. The design of effective suicide prevention strategies hinges on the quantification of the risk and protective factors on suicide. Thus, it is of crucial importance to increase in-depth understanding on the measurable and policy-related factors in elderly suicides. Therefore, identifying the risk factors of such a complex, multidetermined, rare, and dire phenomenon (ie, suicide) is a daunting task. At the individual level, the factors in each of these four domains (ie, demographic, mental health, physical health, and social factors) interact within an older individual to determine his/her suicide risk.3
Since the publication of Durkheim's sociological theory, it has been understood that suicide presents geographical variations. The heart of Durkheim's theory is the belief that suicide is a result of structural and contextual factors rather than individual.6 Durkheim constructed an ecological framework for the analysis of suicide. Various ecological studies are available at the national and regional levels.7-11 However, although these studies are at larger geographical levels and have certainly led to the development of new insights into the spatial variations on suicide rates, nonetheless, it can be argued that they do not cover the entire story. First, suicide shows important variations within a society, and these differences tend to be marked if focus is only on the large aggregate average.12 Second, from the theoretical point, social integration is a process that occurs most intensively in an environment when the residents interact directly (eg, among the neighbors), especially for the older adults whose daily mobility has decreased.13
In the studies conducted at the neighborhood level, the majority focused on the neighborhood social attributes such as neighborhood deprivation, social fragmentation, and social vulnerability.9, 14-20 The concept of social vulnerability is to capture the degree of a person's overall social situation. It provides a useful way in understanding the social problems at multiple nested domains. Social vulnerability can be operationalized as an index of social challenges that a neighborhood has. The ecological framework approach can be useful to understand the social problems at multiple nested domains (eg, individual, family and friends, peer groups, institutions, neighborhoods and communities, and society at large). The cumulative approach allows consideration of numerous factors from multiple domains. The more the social problems, the higher the social vulnerability. From this perspective, building a Social Vulnerability Index (SVI) can capture a holistic view of the social circumstances the neighborhood is facing. Regarding the neighborhood built environments, previous studies have highlighted the important role of neighborhood services (eg, especially on the recreational services, health care services, daily necessity services, and public activity services) in affecting the older adults' mental well-being13; however, most of those discussions are theoretical, lack of empirical investigations. Furthermore, comprehensive analyses that include both the social and built attributes (eg, on service and facility provision) are relatively rare.21-23 Also, with only a few exceptions,15, 16, 24, 25 most of the previous neighborhood-level ecological studies on elderly suicides have been conducted within the Western context.7, 12, 26-29 Compared with the West, Hong Kong and other Asian societies are well known for their higher elderly suicide rate.30, 31 In recent years, due to an aging population and high suicide rate among the elderly, it is observed that more and more attention has been drawn to elderly suicides in Hong Kong.32 Nonetheless, only one study has highlighted the important role of the neighborhood features on suicide in Hong Kong16; the main focus was still on the general population rather than on the older adults. Compared with the younger persons, the older adults may be more vulnerable to the adverse neighborhood factors because of retirement and reduced mobility, decline in cognitive functions, visual and hearing impairment, musculoskeletal diseases and other chronic diseases, and a decrease in social support due to their children moving out or deaths of close persons.13 The features of neighborhoods in Hong Kong vary significantly. Hong Kong comprises three main regions: Hong Kong Island, Kowloon, and the New Territories. The inner-city neighborhoods located on Hong Kong Island (ie, 35 700 persons per square kilometer) and Kowloon (ie, 43 033 persons per square kilometer) have significantly higher population density than the general population density of the New Territories (ie, 3801 persons per square kilometer). Given the aforementioned research gaps, the aims of the present study are to (a) explore the geographical variations in elderly suicides in Hong Kong and (b) explain the geographical variations in elderly suicides by neighborhood social and built attributes.
2 METHODS
2.1 Suicide data
In Hong Kong, all unnatural deaths including suicides are investigated by the coroners.16 The suicide data used in this study were provided by the Coroner's Court. All deaths certified as suicides (ie, the tenth version of the International Classification of Diseases [ICD-10] codes X60-X84: intentional self-harm) have been included in the analyses. The suicide data (2006-2015) of the older adults aged 65 years or older in Hong Kong were extracted from the files of the Coroner's Court. On the basis of the residential addresses of the deceased recorded in the Coroner's Court files, each suicide from 2006 to 2015 was assigned to one of the district councils/constituency areas (DCs/CAs).
2.2 Geographical unit
The DC/CA system was developed for district administration and election purposes. The whole Hong Kong is divided into 18 DCs, which are further divided into hundreds of CAs. There are 400 DCs/CAs (median population aged 65 or older = 1972) and 412 DCs/CAs (median population aged 65 or older = 2165) in 2006 and 2011, respectively. Although the names of most of the DCs/CAs were the same in the two time periods, actually, they are not identical areas. The suicide cases “from 2006 to 2010” (below referred to as “2006”) were assigned to the 2006 DC/CA system, and those “from 2011 to 2015” (below referred to as “2011”) were assigned to the 2011 DC/CA system. There were 90 and 40 DCs/CAs without suicides being reported from 2006 to 2010 and from 2011 to 2015, respectively. The population data (N = 852 345 and 940 966 for population aged 65 or older in 2006 and 2011, respectively) by age (ie, 65-69, 70-74, and 75+), gender (male and female), and DC/CA unit were obtained from the Census and Statistics Department (C&SD), and the DC/CA boundary file data were provided by the Lands Department of the Hong Kong government.
2.3 Dependent variables
Crude (unsmoothed) standardized mortality ratios (SMRs) for suicide in people aged 65 years or older for each of the areas were calculated. The expected number of suicides in each DC/CA was calculated by multiplying the whole Hong Kong age-gender–specific suicide rate (65-69, 70-74, and 75+; male and female) and the corresponding age-gender–specific population in each DC/CA. To account for the instability of the small areas of the SMRs, a spatial smoothing process was implemented by using the Bayesian hierarchical model. The model was based on a Poisson assumption on the observed number of suicide cases with random effects allowing for both global variability (ie, heterogeneity across all DCs/CAs) and local variability (ie, heterogeneity across neighboring DCs/CAs). A simple first-order queen contiguity matrix was created associating each DC/CA with all of its immediate neighbors.
2.4 Neighborhood social environment
Social Vulnerability Index proposed by Gusmano et al33 and modified by Chau et al34 was adopted as one of the independent variables. Since it was tailor-made for the older population and had been modified to suit the local context, this SVI was chosen. Seven domains were included in this SVI, namely, the older population size, institutionalization, living alone, poverty, communication obstacles, disability, and access to primary care (Appendix S1). Each of the seven variables of the DCs/CAs was ranked by deciles. The mean decile ranking of the seven variables represented the SVI of each DC/CA. Higher value indicates greater vulnerability of the population residing in that DC/CA area. The data were provided by the Census and Statistics Department and the Central Registry for Rehabilitation under the Labour and Welfare Bureau. The data on access to primary care were collected from previous work based on the data retrived from the Hong Kong Hospital Authority.35
2.5 Neighborhood built environment
The GEO Community 2009 dataset that includes the addresses of the various types of services and the GEO 2009 dataset that includes the street line file provided by the Lands Department were used for calculating neighborhood built measures. Accessibility of eight specific types of services was also used as independent variables. The eight types of services include recreational services (ie, amusement parks, indoor game halls, recreation centers, sports centers, playgrounds, rest gardens, parks, sports grounds, and sports grounds inside parks), primary health care services (ie, clinics, health centers, and dispensaries), tertiary health care services (ie, hospitals), rehabilitation services (ie, rehabilitation centers), food services (ie, cooked food stalls), daily necessity services (ie, convenient stores, markets, and supermarkets), community services (ie, community centers, community halls, rural committees, youth centers, welfare centers, community service complexes, and family service centers), and transportation services (ie, bus termini, light rail stations, green minibus termini, railway station access, indoor car parks, open car parks, and multistory car parks). The road network distance from the centroid of the DC/CA to the nearest service was calculated for each DC/CA based on the ArcGIS version 10.3 Network Analyst extension.
2.6 Spatial statistical analyses
Unsmoothed SMRs (2006 and 2011) were mapped using the standard deviation categories together with extreme low cutoffs (<0.5) and high cutoffs (>5) and a divergent red-blue color scheme. Bayesian hierarchical models were used to estimate the smoothed SMRs for each area. The Bayesian hierarchical models were based on a Poisson assumption of the observed number of suicide with random effects, allowing for structural variability (heterogeneity among the neighboring areas) and nonstructural variability (heterogeneity across all areas in the study region). Sets of areas that share a border were defined as neighbors. The models were estimated using the Markov chain Monte Carlo methods. Moran's I statistics was calculated to test for spatial autocorrelations in the SMRs.36 Local Moran's I was applied to identify the clusters of the high SMRs. The SVI (2006 and 2011) and service accessibility indicator (2009) were plotted in the maps. As there were no apparent differences between 2009 and 2014, only the 2009 service distribution was plotted. The effect of the SVI and the service provisions on the SMRs (ie, combining the two sets of years) was modeled by the hierarchical Bayesian Poisson model to incorporate exploratory variables. A sensitivity analysis was conducted by using the individual domains in the SVI as independent variables. Another sensitivity analysis based on separated datasets (ie, 2006-2010 and 2011-2015) was also conducted. To study the inequalities in the older adult suicides by area quintiles, the DCs/CAs were first grouped into quintiles based on the neighborhood variables; then, the rate ratios for suicide were estimated for each quintile (the first quintile as reference) using the Bayesian hierarchical models. The maps and spatial analyses were produced using ArcGIS 10.3. and GeoDa 1.12. The basic data processes were performed using the software of SPSS 20, Stata 13.0., and R 3.4.2. The Bayesian hierarchical models and the hierarchical Bayesian Poisson model were performed with WinBUGS 1.4.3.
3 RESULTS
3.1 Descriptive analyses
From 2006 to 2015, there were in total 2567 suicides in the people aged 65 or older. Of those, 42 (1.6%) suicides had missing or incomplete addresses or with an address that was outside Hong Kong, and they were excluded from the analyses. Of the 2525 cases included in the analyses, 1577 (63%) were male. Of the male suicides, 20%, 22%, and 58% were aged 65 to 69, 70 to 74, and 75+, respectively. The corresponding figures for the females were 16%, 19%, and 65%, respectively. The annual suicide rates for the males and the females aged 65 or older were 40 and 20 per 100 000, respectively. Age-specific annual suicide rates for the males aged 65 to 69, 70 to 74, and 75+ were 26, 30, and 51 per 100 000, respectively. For the females, they were 13, 16, and 24 per 100 000, respectively.
3.2 Spatial patterns of the elderly suicides
The spatial distributions of the elderly suicides were shown in Figure 1A. Excluding the DCs/CAs with no recorded suicides, the middle 90% of the unsmoothed SMRs ranged 0.27 to 3.53 in 2006 (a 13-fold difference) and 0.31 to 2.23 in 2011 (a sevenfold difference), suggesting that after excluding the extreme values, the geographical variations in suicides remained. As many areas recorded no suicide or showed some extreme estimates (eg, the highest SMR was 8.58 in 2006 and 5.11 in 2011), the unsmoothed maps are relatively difficult to accentuate the overall spatial pattern. Smoothed maps however show some apparent spatial patterns where they are hard to be clearly identified in the unsmoothed maps (Figure 1B). The high elderly suicide rates were mainly located in the New Territories. There are also some areas identified in the inner-city areas but fewer in the New Territories. The variation in the elderly suicide rates is larger in 2006 than in 2011, with the standard deviation of the SMRs being 0.9 and 0.4, respectively. The spatial distributions of the areas with the highest rates (smoothed SMR > 2) were more dispersed in 2006 than in 2011. The Moran's I was −0.02 (P = .639) in 2006 and 0.04 (P = .148) in 2011, indicating that the spatial distribution of the elderly suicide rates has dispersed in 2006 and becomes more clustered in 2011. The Local Moran's I identified the clusters of the elderly suicides (Figure 1C).
3.3 Spatial correlates of the elderly suicides
Figure 2 shows the spatial variations in the SVI (2006 and 2011) and the accessibility of the eight types of services (2009). The spatial patterns of the SVI between 2006 and 2011 were in general similar with higher vulnerability areas mainly clustered in the older inner-city areas and the areas in the New Territories. The service maps show that although some of the inner-city areas have a higher SVI, compared with the average situation in Hong Kong, the service provision was actually good with shorter distances to the nearest services. However, for those areas located in the New Territories, service provision was significantly lower than the average Hong Kong situation. Figure 3 shows the adjusted elderly suicide rate ratios by the SVI and service quintiles. Compared with the first quintile, increasing elderly suicide rates were observed in areas with increasing levels of quantiles (ie, longer distance) of recreational services, daily necessity services, and community services. Table 1 shows the associations of the elderly suicide rates with the area characteristics of the SVI and the accessibility of the eight types of services. Area SVI was observed not to be a significant factor of the elderly suicide rate, but the elderly suicide rate was significantly associated with the distance to the nearest services including recreational services, daily necessity services, and community services. Those areas with longer distances to such services have higher suicide rates among the older adults. The results from both the unadjusted and adjusted models were found to be consistent. The results from the sensitivity analysis based on separated datasets were also consistent with the main analysis based on combined dataset (Appendix S2).
| Unadjusted | Adjusted | ||||
|---|---|---|---|---|---|
| Area Characteristics | Rate Ratios | 95% CI | Rate Ratios | 95% CI | VIF |
| Social Vulnerability Index (SVI) | 0.968 | (0.642-1.294) | 0.947 | (0.670-1224) | 1.182 |
| Recreational services | 1.005 | (1.003-1.007) | 1.005 | (1.003-1.007) | 1.942 |
| Primary health care services | 1.000 | (0.231-1.768) | 1.000 | (0.224-1.776) | 2.510 |
| Tertiary health care services | 1.000 | (0.492-1.510) | 1.000 | (0.601-1.399) | 2.068 |
| Rehabilitation services | 1.009 | (0.557-1.444) | 1.009 | (0.652-1.336) | 1.354 |
| Food services | 1.000 | (0.521-1.479) | 1.000 | (0.482-1.518) | 2.092 |
| Daily necessity services | 1.002 | (1.000-1.003) | 1.002 | (1.000-1.003) | 2.963 |
| Community services | 1.001 | (1.000-1.002) | 1.001 | (1.000-1.002) | 3.243 |
| Transportation services | 1.007 | (0.812-1.202) | 1.007 | (0.763-1.251) | 1.229 |
- Abbreviation: VIF, variance inflation factor.
4 DISCUSSION
4.1 Summary of the findings
The geographical distributions of suicide in the older adults in Hong Kong showed different spatial patterns from those findings in the Western context (ie, inner-city high suicide rate and suburban low) and some other Asian cities (ie, central low suicide rate and peripheral high). The spatial correlates were also found to be different from those reported in the younger population or the general population in Hong Kong and other countries conducted in previous studies. The suicide rates in the city centers of Hong Kong were also not higher than on average in the city. Clusters of higher suicide rates were mainly found in the New Territories and some inner-city neighborhoods. The spatial variations of suicide in the older adults were not related to the neighborhood SVI. Instead, the neighborhood built features such as recreational services, daily necessity resources, and community centers were significantly associated with suicides in the older adults in Hong Kong.
4.2 Spatial patterns of the elderly suicides
Studies have suggested that the spatial distributions of suicides in older adults showed similar patterns as the general population16 with the simple “inner-city high suicide rate and suburban low” pattern in the Western studies and the “central low suicide rate and peripheral high” pattern in the Asian studies.15 The spatial variations of suicide in older adults in Hong Kong exhibit a much more complicated spatial pattern. In Hong Kong, the “city center” mostly showed below average elderly suicide rates. The “central-high pattern” in the West mainly results from the spatial distribution of deprivation. In Hong Kong, higher suicide clusters were mainly found in the New Territories where the population density was low.37 This pattern is consistent with the findings in China, Japan, and Taiwan where the higher rates of elderly suicide were found to be higher in the rural areas.38 Contributors to the higher rate of elderly suicide in the lower-density population areas include social isolation, lack of services, physical disconnectedness, stigma toward mental disorder, and concentration of the minority groups. Different from the simple spatial pattern in other societies, Hong Kong shows a complicated pattern in the inner-city areas with higher and lower suicide clusters located adjacent to each other, and there is also much heterogeneity in a small area. The 2015 study of Hsu et al conducted in Hong Kong suggested that this complicated distribution mainly results from the complicated distribution of deprivation (ie, the mixed residential pattern between the deprived and the nondeprived groups).
4.3 Spatial correlates of the elderly suicides
In the 2015 study of Hsu et al, social fragmentation and social deprivation were found to be associated with the geographical patterning of suicides in people aged 10 years or older. The same findings have also been observed in other East Asian countries such as in Taiwan15 and South Korea.25, 39 However, in the present study, neighborhood social attribute was not a significant factor on suicides in the older adults. The potential reason might be that for older adults, the feeling of belonging to a community as well as social interactions might be more important than the economic status itself. Despite being mentioned widely in various theories and conceptual models that the provision of neighborhood services is particularly important for the mental well-being of the older adults, empirical research on this was very rare. The present study found that suicide in older adults was not related to health care services; however, it was related to nonhealth care services (ie, daily necessity services). Not only do public resources play the role of providing daily necessity services, but more importantly, they also serve as communal spaces that improve social inclusion and contribute to the development of social capital in the communities.40, 41 Therefore, the findings in the present study suggest that social exclusion or social disconnection might be a more important aspect related to elderly suicides. On the basis of the spatial and social network theory, public services also provide important public spaces that can contribute to creating social equity by providing unrestricted access to space for people from different backgrounds to meet and communicate, thereby also improve social connection.42 Viewing from this aspect, the existence of public services in terms of their availability and accessibility might be beneficial to the older adults to “come out and meet others” especially for those single elderly who live alone and experience loneliness. Public services thus provide good opportunities for the older adults to meet and interact more with each other and in turn reduce loneliness in old age. Regarding the mental health services, in Hong Kong, there are two types of institutions that provide elderly mental health services, including general health care services (eg, primary health care services including clinics, health centers, dispensaries, and tertiary health care services including hospitals) and community mental health centers (ie, psychogeriatric clinics; however, they are dominated by patients with dementia). Due to lack of data, only the association between general health care services with elderly suicides was investigated. The insignificant results suggest that the geriatric mental health services in general health care services are insufficient as well as inaccessible.
4.4 Strengths and limitations
To the knowledge of the authors, this is the first small area analyses of suicides of the older adults in Asia. Sophisticated statistical models were used to account for the uncertainties in estimating the suicide rate ratios in the small areas. Furthermore, compared with the limited variables in previous studies, this paper comprehensively investigated the spatial correlates of suicides in the older adults with both the social and the built factors. Nonetheless, some limitations should be noted. First, the aims of the present study are to investigate into the geographical distributions of suicides in the older adults and their possible spatial correlates. Therefore, the reported associations were ecological in nature and might not be replicated in analyses at the individual level, which is widely known as the “ecological fallacy.”43 With more detailed individual-level data, future multilevel studies are thus warranted to explore into the interaction effects between the characteristics of the individual and the neighborhood levels on suicide. Second, the concept of social vulnerability aims to capture the overall vulnerability problem; more specific aspects of the neighborhood social environment should be explored in future studies. A sensitivity analysis was conducted by using the individual domains in the SVI as independent variables. The results were consistent in that recreational services, daily necessity resources, and community centers played a more prominent role in affecting suicides in the older adults. On the basis of the results from the present study, the sensitivity analysis, and the 2015 study of Hsu et al,16 the same conclusion that suicides in older adults are not significantly related to the three aspects of social environment (ie, social vulnerability, social poverty, and social deprivation) was reached. However, more aspects such as measuring the neighborhood social support both objectively and subjectively of the social environment should be further explored in future studies. Third, those specific elderly services (eg, geriatric health and mental health services and geriatric centers) play significant roles in affecting the older adults' mental well-being. However, due to lack of data, the relationship of elderly suicides with those elderly specific services could not be empirically investigated. In future studies, these should be explored with more specific elderly service datasets.
5 CONCLUSION
From the methodological perspective, in contrast to most previous ecological studies that only focused on the socioeconomic aspect, the present study included the built environment such as various types of services. From the theoretical perspective, it calls for more age-specific studies in suicide to complement the prevention and intervention efforts. As the results of the present study suggest, since the underlying mechanism would be significantly different among different age groups, suicide prevention strategies should be age specific. From the practical perspective, efforts aimed at reducing elderly suicides might be more effective if they target neighborhoods rather than the individuals. Sometimes, however, it is difficult, if not impossible, to make any change at this latter part of life. Thus, as the findings clearly suggest that contextual factors can explain the heterogeneity of the spatial patterns in elderly suicides, therefore, area-level strategies from the public health perspective can be valuable. Elderly programs that strengthen recreational services and daily necessity services in the areas with high elderly suicide rates may thus reduce the spatial disparity in elderly suicides in Hong Kong.
ACKNOWLEDGEMENTS
This study was approved by the Human Research Ethics Committee, The University of Hong Kong. This study was funded by the Knowledge Exchange Funding, The University of Hong Kong (2018/19-85), and the General Research Fund, Research Grants Council, Hong Kong (GRF#106160261). The funders assume no role in the study design, collection, analyses, and interpretation of the data, in writing of the article or in the decision to submit the article. This funding body did not contribute to or influence the study design or the findings. Gratitude to the Coroner's Court, the Census and Statistics Department, the Lands Department, the Planning Department, and the Central Registry for Rehabilitation, The Government of the Hong Kong Special Administrative Region, for their support in allowing access to the data. All maps were produced with the permission of the director of the Lands Department. Appreciation also to Dr Mengni Chen and Miss Xinyi Zeng from the Hong Kong Jockey Club Centre for Suicide Research and Prevention, The University of Hong Kong, for their vibrant discussions.
CONFLICT OF INTEREST
None declared.
Open Research
DATA AVAILABILITY STATEMENT
The data were obtained by application from a third party (the Hong Kong government). The authors had no special access privileges to these data. Interested researchers can obtain the data by request or purchase from the Hong Kong government by contacting the relevant departments—the Hong Kong Judiciary for the suicide data, the Census and Statistics Department for the census data, the Lands Department for the service data, the Planning Department for the boundary data, and the Central Registry for Rehabilitation for the Social Vulnerability Index data.
