Changing Patterns in Rates and Means of Suicide in California, 2005 to 2013

Abstract

Objectives. To describe recent trends in suicide throughout California and to compare rates and methods of suicide (“means”) across demographic groups.

Methods. Data from statewide mortality records were used to estimate age-adjusted rates of suicide from 2005 to 2013, overall and by means, age, gender, race/ethnicity, urbanicity, and county.

Results. The suicide rate increased 12.6% between 2005 and 2013, from 11.2 to 12.6 per 100 000 population, but this overall trend masks substantial heterogeneity across subgroups. In particular, rapid increases were observed for individuals of multiple races/ethnicities. Means of suicide changed, trending away from firearms toward suffocation and drug poisoning.

Conclusions. High-risk groups and means of suicide are changing rapidly in California, so appropriate public health programming should prioritize population-based strategies.

Suicide is a leading cause of premature mortality in the United States, accounting for more than 40 000 deaths in 2014.¹ Between 1999 and 2014, suicide rates increased 23%, from 10.5 to 12.9 per 100 000 population,¹ with rates nearly 3 times higher in certain groups such as older White men. Comprehensive explanations for this increase are lacking, but research suggests that several factors are possible contributors, including the rising prevalence of long-term morbidity, physical disability, and pain; growing rates of psychological disorder and substance abuse; declining job prospects; and increasing social conflict, income inequality, and racial inequality.^2,3 Rising rates along with intensified attention to the role of firearms in society have increased the focus on suicide as an important population health issue.^2,4 However, suicide research relevant to prevention continues to be underprioritized.⁵

Recommended population-level strategies for suicide prevention often focus on restricting access to certain methods of suicide (“means”) such as firearms. Although these interventions can be extremely effective, they require timely information on suicide means. In the United States, about half of suicides are completed with firearms, but other means such as drug poisoning and suffocation now make up the majority.¹ These trends, along with increasing overall rates, highlight that the epidemiology of suicide is changing, and prevention approaches may need to adjust as well.

Recent developments in suicide in California are of interest with respect to public health at the national level for several reasons. First, demographic factors such as race/ethnicity and immigration status are strongly associated with suicide.⁶ The US population is expected to comprise greater proportions of racial/ethnic minority groups and foreign-born individuals in the future, similar to California’s current demographic composition.⁷ Thus, California’s experience may help guide planning for future trends in suicide nationally. Second, in 2014, the percentage of suicides attributable to firearms was just under 50% nationally but only 38% in California.¹ California has some of the most restrictive firearm policies in the country, and thus patterns in that state may be informative about what might be expected nationwide if such policies were more broadly adopted.

To our knowledge, only 2 studies have systematically addressed suicide epidemiology in the California context. The first focused only on firearm suicides between 1997 and 1999 among youths.⁸ The second described trends in suicide fatalities by age, by gender, and in 5 large state regions from 1999 to 2009 but failed to address differences across racial/ethnic groups and means.⁹ Other regional and national data also exist but are limited. For example, the Centers for Disease Control and Prevention’s Web-based Injury Statistics Query and Reporting System lacks information on geographic levels smaller than states and on individuals of multiple races/ethnicities, who may be at higher risk for suicide.¹⁰ To address these gaps and inform priority setting for prevention programs, we examined how rates and means of suicide in California varied across demographic groups between 2005 and 2013.

METHODS

We derived data on suicides occurring between 2005 and 2013 from death records provided by the California Department of Public Health’s Office of Vital Statistics. Age-standardized yearly rates were calculated overall and by means, age, gender, race/ethnicity, year of death or injury, county, and urbanicity of residence. We used primary cause of death codes from the International Classification of Diseases, 10th Revision¹¹ to classify means of suicide. To avoid small cell counts and unstable rates and to protect confidentiality, we restricted the study population to Californians aged 15 to 84 years at the time of their death.

RESULTS

Between 2005 and 2013, there were 30 029 suicides in California. Suicides occurred in all subgroups, but the distribution was extremely heterogeneous. The suicide rate increased 12.6% over the study period, from 11.2 to 12.6 per 100 000 population. The percentage of suicides involving firearms decreased from 29.0% to 26.9%, whereas the percentages involving drug poisoning and suffocation increased from 15.9% to 17.1% and from 25.8% to 29.0%, respectively.

Table 1 presents trends in age-adjusted suicide rates by means and race/ethnicity. Trends in suicide means were similar in all racial/ethnic groups with the exception of American Indians, among whom use of firearms declined substantially and use of suffocation and other means increased markedly. Also of note, there was a 137% increase in the suicide rate among individuals of multiple races/ethnicities; this group had the lowest rate in 2005 (5.3 per 100 000) and the second highest rate in 2013 (12.6 per 100 000). Additional findings according to gender, age group, birth cohort, urbanicity, and county are presented in the appendix (available as a supplement to the online version of this article at http://www.ajph.org). Suicide rates were higher in rural counties than in urban counties, but rapid increases and decreases were observed in all types of counties over the study period.

TABLE 1—

Trends in Age-Adjusted Suicide Rates, by Means and Race/Ethnicity: California, 2005–2013

Race/Ethnicity and Study Year	All Means	Firearms	Suffocation	Drug Poisoning	Othera
American Indian or Alaska Native
2005	9.75	2.56	1.32	1.42	4.17
2009	12.78	3.23	6.58	1.46	1.46
2013	12.12	0.00	5.45	1.23	5.23
Asian or Pacific Islander
2005	6.21	0.74	2.91	0.50	2.11
2009	7.86	1.11	3.67	0.77	2.51
2013	7.07	0.92	3.23	0.68	2.60
Black
2005	7.53	1.61	1.42	1.41	3.06
2009	7.30	2.27	1.13	1.38	2.86
2013	8.43	2.32	2.72	0.95	2.37
Hispanic
2005	5.72	1.31	2.38	0.60	1.43
2009	6.05	1.42	2.44	0.72	1.59
2013	5.94	1.29	2.55	0.51	1.51
≥ 2 races/ethnicities
2005	5.31	0.90	2.03	1.54	2.18
2009	8.87	2.24	3.59	1.29	2.75
2013	12.59	3.78	4.13	1.99	4.10
White
2005	15.83	4.95	3.49	2.81	5.50
2009	19.10	5.81	4.83	3.77	6.44
2013	19.57	5.63	4.74	4.01	7.07

Note. Data are estimated rates per 100 000 population.

^aIncludes suicides involving poisoning by nonmedicinal substances or drugs; drowning; sharp objects, cutting, or piercing; falls; and other means such as electrocution, exposure to extreme cold, and lying or jumping in front of a moving object such as a motor vehicle or train.

DISCUSSION

In this study, we identified heterogeneous and alarming trends in suicide across demographic groups that indicate the need for increased efforts to mitigate rising rates. Notably, we documented dramatically increasing rates of suicide among individuals of multiple races/ethnicities. This finding is consistent with existing literature suggesting that individuals of multiple races/ethnicities may be at higher risk with respect to numerous problem areas (e.g., see Choi et al.¹⁰); however, the rapid increase we observed is worrisome and underrecognized. We hypothesize that this trend may be related to growing race/ethnicity–related social tension and concerns regarding racial/ethnic identity (see the discussion in the supplemental appendix), but this pattern warrants a thorough public health investigation and response.

We found that many groups can be considered “high risk” for suicide, and high-risk groups are changing, in some cases very rapidly, which makes targeted prevention programs challenging. Thus, population-level strategies such as broad screening and education on warning signs and available resources in primary care centers and schools should be prioritized (e.g., see Portzky and Van Heeringen¹²). Investigation of the drivers of decreasing suicide rates in certain counties may help identify effective population-level prevention strategies. These approaches also have the benefit of reaching individuals in lower risk subpopulations who nonetheless contribute substantially to the burden of suicide as a result of larger population numbers.

Consistent with national trends, we found that suicide means are shifting away from firearms toward suffocation and drug poisoning. This pattern is particularly prominent among American Indians and merits further investigation. The movement toward means that are less feasible to restrict highlights the need to better understand and address the fundamental causes of suicide. For example, the development and implementation of interventions that address key risk factors such as early life trauma and chronic physical conditions could have a more broad-reaching and long-lasting impact on suicide. In addition, further investigation of whether the changing composition of means is driven by means substitution is vital in identifying the most appropriate future prevention efforts, namely whether additional means restriction efforts are expected to be effective.

Our study involved several limitations as well as strengths. Cause of death classifications on death certificates are known to be imperfect. However, authors of previous studies on this topic have generally concluded that the degree of misclassification is not sufficiently substantial to alter major trends and patterns.¹³ It is possible that changes in racial/ethnic classification on death or census records over the study period contributed to the observed patterns; however, they are unlikely to explain the observed 137% increase in suicide rates among individuals of multiple races/ethnicities. We were able to report trends only through 2013, the most recent year of data available at the time of the study. Additional investments in timely health information systems would be valuable in facilitating timely reactions to emerging public health trends.

Despite these limitations, substantial multiracial and American Indian populations live in California, and this allowed us to provide stable estimates of rates in these important, high-risk subgroups that are often overlooked.