Temporal Trends in Mortality Rates among Kaiser Permanente Southern California Health Plan Enrollees, 2001-2016


Wansu Chen, PhD1; Janis Yao, MS1; Zhi Liang, PhD1;
Fagen Xie, PhD1; Don McCarthy, MS1; Lee Mingsum, MD, PhD2;
Kristi Reynolds, PhD1; Corinne Koebnick, PhD1;
Steven Jacobsen, MD, PhD1

Perm J 2019;23:18-213 [Full Citation]

E-pub: 04/04/2019


Background: Temporal analyses of death rates in the US have found a decreasing trend in all-cause and major cause-specific mortality rates.
Objectives: To determine mortality trends in Kaiser Permanente Southern California (KPSC), a large insured population, and whether they differ from those of California and the US.
Methods: Trends in age-adjusted all-cause and cause-specific mortality rates from 2001 to 2016 were determined using data collected in KPSC and those derived through linkage with California State death files and were compared with trends in the US and California. Trends of race/ethnicity-specific all-cause and cause-specific mortality rates were also examined. Average annual percent changes (AAPC) and 95% confidence intervals (CI) were calculated.
Results: From 2001 to 2016, the age-adjusted all-cause mortality rate per 100,000 person-years decreased significantly in KPSC (AAPC = -1.84, 95% CI = -2.95 to -0.71), California (AAPC = -1.60, 95% CI = -2.51 to -0.69) and the US (AAPC = -1.10, 95% CI = -1.78 to -0.42). Rates of 2 major causes of death, cancer and heart disease, also decreased significantly in the 3 populations. Differences in trends of age-adjusted all-cause mortality rates and the top 10 cause-specific mortality rates between KPSC and California or the US were not statistically significant at the 95% level. No significant difference was found in the trends of race/ethnicity-specific, sex-specific, or race/ethnicity- and sex-specific all-cause mortality rates between KPSC and California or the US.
Conclusion: Trends in age-adjusted mortality rates in this insured population were comparable to those of the US and California.


All-cause mortality rates have decreased steadily in the US in the past 3 decades, but the trends of this decrease differed by age, sex, and race/ethnicity.1,2 More recent trends in mortality caused by cardiovascular disease (CVD), heart disease, and stroke in the US between 2000 and 2014 indicate that the decline in death rates has slowed since 2011.3 However, there are large geographic variations in all-cause mortality and disease-specific mortality.4 Because mortality is an important indicator of population health, the overall and cause-specific mortality trends can inform health policy,5 allow the identification of modifiable factors,6-8 and guide the design of population-based and clinical care interventions.

Integrated health care delivery systems are associated with overall better adherence to evidence-based care guidelines, better survival rates, and reduced racial disparities.9-11 Advantages of integrated health systems are their ability to coordinate care and conduct large-scale and sustained care-improvement initiatives to emphasize prevention, improve disease outcomes, and reduce mortality.12,13 One example is the sepsis mortality reduction initiative in 21 hospitals of Kaiser Permanente Northern California (KPNC).14 The sepsis mortality rate at these KPNC hospitals declined from 24.6% to 11.5% in less than 3 years.14 The reported steeper reduction of heart disease, stroke, and all-cause mortality rates among KPNC enrollees compared with those of the US population15 could be related to the implementation of a large-scale hypertension prevention program.16 

Lower age-adjusted mortality rates have been observed in Hispanic17-19 and Asian/Pacific Islander populations19 compared with that of non-Hispanic whites. Mortality rates were even lower for Hispanics than non-Hispanics after adjusting for annual family income.17 However, not all studies support these conclusions because of the differences in the populations being studied.17 For example, differences in mortality between Hispanics who immigrated to another country vs those who did not were noted.20-24 In Southern California, the growth of Hispanic and Asian populations in the past 2 decades has been substantial. Thus, the advantage in mortality of the 2 populations could favorably affect the overall mortality rates of Kaiser Permanente Southern California (KPSC) enrollees if the mortality advantage does prevail among Hispanic and Asian/Pacific Islander populations.

Although absolute values of mortality rates are important, temporal trends in mortality are extremely informative because they could reflect changes of individual, organizational, or societal factors, including individual behaviors, medical practice (eg, change of practice guidelines or introduction of new treatments), medical technology, and the environment where people live. For example, cancer screenings can influence cancer mortality rates, and more advanced drug treatment can have an impact on the mortality rates owing to heart disease. In this study, both mortality rates and trends between 2001 and 2016 were studied for KPSC Health Plan enrollees. However, given the ethnic diversity of the KPSC and California populations compared with that of the US, we focused on mortality trends rather than the absolute values of mortality rates when the 3 populations (KPSC, CA, and the US) were compared. More specifically, the goals of this study were to 1) examine the age-adjusted mortality rates and trends of KPSC Health Plan enrollees; 2) study the age-adjusted sex- and race/ethnicity-specific mortality rates and the trends of all-cause mortality and the leading causes of mortality among the Health Plan enrollees; and 3) compare the age-adjusted overall and sex- and race/ethnicity-specific mortality trends and cause-specific mortality trends between KPSC and California and between KPSC and the US during the study period.


Study Setting and Data Sources

KPSC is an integrated health care organization that currently serves approximately 4.6 million enrollees, about 19% of the population in Southern California. The demographics and socioeconomic status including race/ethnic composition of the enrollees are representative of those living in the region.25 

Date of birth, sex, and race/ethnicity were collected administratively as part of Health Plan enrollment and/or patient care. Race and ethnicity information was based on a combination of administrative and self-reported data.26 

The study protocol was approved by the KPSC institutional review board.

Study Subjects and At-risk Person-Time

Health Plan enrollees up to 110 years of age who had more than 1 day of enrollment between 2001 and 2016 were included. Enrollees whose age was missing were excluded. Those who were older than age 110 years were also removed from the analysis because they represented a very small number of enrollees, and some may have had an incorrect date of birth. Enrollees whose sex was labeled “other” or “unknown” were also excluded.

For each enrollee, the total number of days enrolled was considered the at-risk time for each calendar year. The at-risk time started from the date when the enrollee joined KPSC or January 1, whichever occurred later. The at-risk time ended at the disenrollment from the Health Plan, date of death, or December 31, whichever occurred first. Gaps in enrollment of 45 days or shorter were bridged (ie, continuous enrollment was assumed) even if the gaps spanned 2 consecutive years. For example, an enrollee who joined KPSC on July 1, 2010; terminated enrollment on November 30, 2010; rejoined on January 1, 2011; and died on October 30, 2011, contributed 6 (instead of 5) and 10 months of at-risk time in 2010 and 2011, respectively. If an enrollee had multiple enrollment periods within a calendar year that were greater than 45 days apart, the lengths of these enrollment periods were summed and the total length was designated as the at-risk time contributed by the enrollee for that specific year.

For age-specific analyses, an enrollee could contribute to at-risk periods that belonged to different age groups. For example, an enrollee who turned 65 years of age on July 1 could contribute 6 months of at-risk time to the group aged 55 to 64 years and another 6 months to the group aged 65 to 74 years.

Finally, the individual-level age-, sex-, and year-specific at-risk times for each calendar year were combined to form the total at-risk person-time for that specific year. The same process was repeated for each calendar year between 2001 and 2016.


The KPSC death records (in the Research Data Warehouse) were derived by identifying deaths that occurred at KPSC-owned facilities, outside facilities that submitted claims to KPSC, or deaths reported to the Health Plan. These records were supplemented by linking the enrollees with the decedents in the California Death Statistical Master Files (up to 2014), the California Comprehensive Death File (since 2015), the State Multiple Cause of Death File, the State Fetal Death Files, and the Social Security Administration (SSA) Death Master Files. The linkage process is described in Appendix 1.a All death records of Health Plan enrollees since 1988 are kept, including neonatal and fetal deaths.

The cause of death was determined by the underlying cause of death obtained from state death records. The state death records used the Tenth Revision of International Classification of Diseases, Clinical Modification (ICD-10-CM) for underlying cause of death since 1999. Thirty-three cause-of-death categories were identified using the classification defined by the Centers for Disease Control and Prevention (CDC; Appendix 2a). The residual groups “all other diseases” (Item Number 29 in Appendix 2a) and “all other external causes” (Item Number 33 in Appendix 2a) were not included in the ranking process in which we selected the top 10 causes of deaths for each calendar year. In addition, we also examined all cardiovascular conditions by using ICD-10 codes I00-I99 between 2001 and 2016.

Although the linkage process described here is capable of capturing deaths after disenrollment, our intention was to estimate death rates during Health Plan enrollment. However, deaths that occurred 1 month after Health Plan disenrollment were considered, because the Health Plan coverage may not have been renewed for some of the patients who were under end-of-life care. For these patients, their at-risk periods were extended to the date of death. For example, if an enrollee disenrolled on June 30 and died on July 7, his/her at-risk window was expanded from January 1 to June 30 (on the basis of enrollment records) to January 1 to July 7 (on the basis of actual date of death).

Statistical Analysis

The overall, sex-stratified, and race/ethnicity-stratified age-adjusted mortality rates were calculated using the direct method27 and the projected year 2000 US population as the standard population (Appendix 3a). Enrollees with race/ethnicity other than non-Hispanic white, Hispanic, African American, and Asian/Pacific Islander (ie, Native American, Alaskan, other, or multiple) and those with unknown race/ethnicity were included in the analyses that were not specific to race/ethnicity.

For KPSC enrollees, we ranked the age-adjusted, cause-specific mortality rates for each calendar year. For California and the US populations, we obtained the rates only for those causes that were ranked at the top 10 in 2016 for KPSC enrollees. Deaths with unknown causes were removed from the cause-specific analyses. For each cause, we reported age-adjusted rates standardized to the projected year 2000 US population overall as well as stratified by race/ethnicity. The analyses of CVD mortality rates (defined as the sum of heart disease and cerebrovascular disease mortality rates) were limited to adults aged 45 years or older. For comparison, we included relevant age-adjusted mortality rates for the entire US and the State of California populations. The US and California mortality rates between 2001 and 2016 were derived from the CDC’s Wide-Ranging Online Data for Epidemiologic Research CDC WONDER dataset (https://wonder.cdc.gov/ucd-icd10.html). To estimate an average annual percentage change (AAPC) in mortality rates, we first calculated the annual percentage change in age-adjusted mortality rates of 2 consecutive years (ie, slope) on a log scale and then derived the geometric mean of the annual percentage changes and their 95% confidence intervals (CIs).28 For comparison of AAPCs from 2 populations, a Z-statistic was formed by dividing the difference in the 2 AAPCs and the standard error (SE) of the difference (ie, sqrt [SE12 + SE22], where sqrt is the square root and SE1 and SE2 are the standard errors of the 2 individual estimates). The analysis was performed on a log scale. A test was considered statistically significant if the p value was < 0.05.

Because all-cause and cause-specific mortality rates vary considerably by age and some specific causes of mortality are more relevant for adults, we conducted sensitivity analyses for overall and sex-specific all-cause mortality, and for overall cause-specific mortality for the top 10 causes by including only adult Health Plan enrollees who were 25 or more years of age.


The number of KPSC Health Plan enrollees increased from nearly 3.4 million in 2001 to almost 4.6 million in 2016 (Table 1). During the same period, the mean age increased from 34.8 to 38.2 years and the proportions of Hispanic and Asian/Pacific Islander enrollees increased (37% to 43% and 8% to 12%, respectively).

All-Cause Mortality

Age-Adjusted All-Cause Mortality

From 2001 to 2016, the age-adjusted, all-cause mortality rate per 100,000 person-years in KPSC decreased from 684 to 521 (AAPC -1.84, 95% CI -2.95 to -0.71; Table 2). During the same period, the corresponding rates in the US and California decreased from 859 to 729 (AAPC = -1.10, 95%, CI = -1.78 to -0.42) and from 783 to 617 (AAPC = -1.60, 95% CI = -2.51 to -0.69), respectively. The differences in trends between KPSC and California and between KPSC and the US were not statistically significant at the 95% level.

In KPSC, the AAPCs for males and females were -1.94 (95% CI = -2.94 to -0.93) and -1.78 (95% CI = -3.35 to -0.18), respectively, between 2001 and 2016 (Table 2). The trend estimates did not differ statistically from those of California and the US. In all 3 populations, mortality rates for males appeared consistently higher than those of females (Table 2). When the analyses were limited to adults aged 25 years or older, all the comparisons of AAPCs mentioned in this section yielded the same conclusions (data not shown).

Race/Ethnicity-Specific All-Cause Mortality

In KPSC, Asian/Pacific Islanders had the lowest age-adjusted mortality rates during the study period (377/100,000 person-years in 2016), followed by Hispanics (445/100,000), non-Hispanic whites (568/100,000) and African Americans (652/100,000; Figure 1, Supplemental Table E1a). For all racial/ethnic groups (non-Hispanic whites, Hispanics, African Americans, and Asian/Pacific Islanders), the overall and sex-specific all-cause mortality rates in KPSC seemed to be consistently lower compared with those of California and the US (Figure 1, Supplemental Table E1a). Asian/Pacific Islanders seemed to have a more rapid decline in mortality rates between 2001 and 2016 (AAPC = -1.95, 95% CI = -5.27 to 1.49), compared with African Americans (AAPC = -1.36, 95% CI = -3.26 to 0.58), non-Hispanic whites (AAPC = -1.30, 95% CI = -2.67 to 0.08), and Hispanics (AAPC = -1.28, 95% CI = -7.60 to 5.48). However, the trend estimates in the 4 race/ethnicity populations were not significantly different at the 95% level (Supplemental Table E1a). No statistically significant difference was found in the trends of race/ethnicity-specific, or race/ethnicity- and sex-specific age-adjusted all-cause mortality rates between KPSC and California and between KPSC and the US (Supplemental Table E1a).

Leading Causes of Mortality

Top 10 Leading Causes of Death

Supplemental Table E2a displays the top 10 causes of death in 2016 for KPSC and the corresponding mortality rates in California and the US. Cancer and heart disease were the leading causes of death between 2001 and 2016 for all 3 populations (KPSC, California, and the US; Supplemental Table E2a, Figure 2). During the last 10 years of the study period (2007 to 2016), the rank of the top 5 causes of death in KPSC remained the same (Supplemental Table E2a).

At KPSC, cancer was the leading cause of death in 2016 with 133 deaths per 100,000 person-years. The second, third, and fourth leading causes of death were heart disease (113/100,000 person-years), Alzheimer disease (39/100,000 person-years), and cerebrovascular diseases (31/100,000 person-years), respectively. During the study period, age-adjusted mortality rates of cancer (AAPC = -1.87, 95% CI = -2.75 to -0.98), heart disease (AAPC = -3.18, 95% CI = -4.78 to -1.54), and influenza and pneumonia (AAPC = -7.08, 95% CI = -13.63 to -0.04) decreased significantly (Supplemental Table E2a). The decrease of mortality rate for cerebrovascular diseases from 57 to 31 per 100,000 person-years was impressive; however, it was not statistically significant (AAPC = -4.42, 95% CI = -8.67 to 0.02). Decreasing trends were observed for cancer, heart disease, and cerebrovascular disease in the US and California populations. A statistically significant increase in the Alzheimer disease mortality was observed in the US (AAPC = 2.91, 95% CI = 0.14 to 5.75) and California (AAPC = 4.80, 95% CI = 1.48 to 8.23) populations, but not in the KPSC population (AAPC = 1.82, 95% CI = -2.66 to 6.50). In 2016, Alzheimer disease was ranked as the sixth leading cause of death in the US and the third leading cause of death in California. No statistically significant difference was found in the trends of any age-adjusted cause-specific mortality rates between KPSC and California, or between KPSC and the US between 2001 and 2016 for the top 10 causes of death (Supplemental Table E2a). When the analyses were limited to adults 25 or more years of age, the comparisons of AAPC between KPSC and the US/California yielded the same conclusions (data not shown).

Top 10 Leading Causes of Mortality by Race/Ethnicity

In KPSC, African American enrollees had the highest age-adjusted cancer and heart disease mortality rates (152 and 146/100,000 person-years in 2016, respectively), followed by non-Hispanic whites (147 and 127/100,000 person-years in 2016, respectively) (Supplemental Table E3a). African American and non-Hispanic whites enrollees had higher mortality rates caused by Alzheimer disease, compared with Hispanics and Asian/Pacific Islanders in all of the years studied. African American enrollees also had the highest mortality rates for diabetes mellitus during the study period, followed by Hispanics. The mortality rates of chronic lower respiratory disease and accidents were highest among non-Hispanic white enrollees.

Figure 3 shows the age-adjusted mortality rates by race/ethnicity for each of the top 6 causes. The reduction in the rates of cancer mortality between 2001 and 2016 seemed to be larger in Asian/Pacific Islander (AAPC = -3.20; 95% CI = -8.81 to 2.76) and African Americans (AAPC -2.74, 95% CI -5.48 to 0.09), compared with those of Hispanics (AAPC = -0.93, 95% CI = -4.03 to 2.28) and non-Hispanic whites (AAPC = -1.67, 95% CI = -2.88 to -0.45); nevertheless, the differences were not statistically significant (Supplemental Table E3a). Age-adjusted mortality owing to Alzheimer disease seemed to increase the most for Asian/Pacific Islander (AAPC = 6.06, 95% CI = -5.71 to 19.31) compared with those of Hispanics (AAPC = 2.70, 95% CI = -5.59 to 11.73), non-Hispanic whites (AAPC = 2.14, 95% CI = -2.56 to 7.07) and African Americans (AAPC = -0.62, 95% CI = -13.82 to 14.61), respectively. However, there was no statistically significant difference.

Trends of age-adjusted CVD mortality rates were similar to those of heart disease (Supplemental Table E4a).

18 213


The current study was conducted in a large cohort of Health Plan enrollees over 16 years. Our findings suggest that despite the fact that the age-adjusted mortality rates declined significantly in all 3 populations (KPSC, CA, and the US), the trends of age-adjusted all-cause and cause-specific mortality rates in KPSC were comparable to those of California and the US. Similarly, when the analyses were stratified by sex and race/ethnicity, the trends of age-adjusted mortality rates in KPSC remained comparable to those of California and the US.

The decline in mortality from heart disease may be attributed to changes in risk factors and progress in treatment. The decreasing prevalence of important cardiovascular risk factors, including cigarette smoking, elevated total cholesterol, high systolic blood pressure, and physical inactivity, were reported to account for almost half of the decrease in death caused by coronary artery disease.29 Improvement in secondary prevention therapies as well as timely revascularization via coronary artery bypass surgery and percutaneous coronary intervention also contributed to the reduction in cardiovascular mortality.30 Unfortunately, this trend is offset by major increases in the prevalence of obesity and diabetes, causing a deceleration in the rate of decline between 2011 and 2014 for all cardiovascular deaths.3 The slowing in rate decline is consistent with the data we reported for the entire US population. The trend in decreasing cardiovascular mortality rate also seemed to slow for KPSC enrollees and California residents since 2011.

The decline in cancer mortality could largely be attributed to screening31,32 and more advanced treatment.33 Other factors affecting cancer mortality included smoking,34 unhealthy diet,35 and obesity.36 The decline in cancer mortality in the US was reported to be stable between 2000 and 2014 by Sidney et al.3 The same pattern was observed for KPSC enrollees and California residents during the same period.

Although studies have shown associations between air pollution and respiratory and allergic conditions37 and air quality has been poor in California because of traffic and wildfire-related pollution,38 California residents experienced lower mortality rates because of chronic lower respiratory tract diseases compared with those of the US in recent years (Supplemental Table E2a). Research based on adult Californians who responded to the Behavioral Risk Factor Surveillance System in 2011 reported an increased risk of chronic obstructive pulmonary disease among white and black residents, compared with Hispanic residents.39 In KPSC, the mortality rates of chronic lower respiratory diseases in non-Hispanic whites and African American populations were higher compared with those of Hispanic and Asian/Pacific Islander.

Alzheimer disease surpassed chronic lower respiratory diseases in 2004 and cerebrovascular disease in 2007, and became the third leading cause of death among KPSC enrollees. The increasing trend of morality rates of Alzheimer disease was significant in California and the US populations, but not in KPSC. The increase of Alzheimer disease death rates could be attributed to older age at the time of deaths. When advanced medicine prolongs lives and reduces mortality caused by cancer and CVD, people are more likely to die of Alzheimer disease or its complications. Our results, based on KPSC enrollees, showed that the mortality rate due to Alzheimer disease was highest in African American enrollees, followed by non-Hispanic whites enrollees. This result is consistent with what was reported by Taylor et al.40 Patients with Alzheimer disease typically die because of comorbidities (eg, infections), poor functional status, lack of nutrition, delirium, and severe cognitive impairment.41-44 

Our findings also suggest that in the KPSC population, Hispanic and Asian/Pacific Islander enrollees had lower age-adjusted mortality rates, compared with those of African Americans and non-Hispanic white enrollees. This finding is similar to that of a meta-analysis in which the authors reported that Hispanics had lower overall mortality than did non-Hispanic whites and non-Hispanic blacks, but overall higher risk of mortality than did Asian Americans.45 The mortality advantage of Hispanics and Asian/Pacific Islanders may be partially attributable to the healthy immigrant effect,20-22 in which those who choose to migrate to another country are in general healthier than those who decide to stay. However, other studies found only weak evidence to support  such a hypothesis.21,23 Another potential explanation could be the “salmon bias” effect, or reverse immigration hypothesis, in which selective immigrants, especially the less healthy ones, returned to their countries of origin.24 However, authors of other studies believed that the evidence was not enough46 or could explain only part of the advantages.47

The observed lower all-cause and race/ethnicity-specific mortality rates at KPSC compared with those of California and the US should be interpreted with caution. It is very likely that the advantage in mortality rates is attributable to the better coordination and delivery of care within the integrated health care system. However, it may also be possible that people who joined KPSC were healthier than other local residents because of more stable insurance coverage or healthier lifestyles.

Similar to age-adjusted all-cause mortality, the age-adjusted mortality rates caused by cancer and heart disease also varied significantly among racial/ethnic groups in the KPSC population. Our results are consistent with the US national data between 2010 and 2014 showing that African Americans and non-Hispanic whites had the higher CVD mortality rates (African Americans being the highest and non-Hispanic whites the second highest) compared with those of Hispanic and Asian/Pacific Islander.3 A study that evaluated the racial/ethnic differences in the risk of coronary heart disease in a cohort of 1.3 million KPNC enrollees showed that compared with whites, blacks, Latinos, and Asians all had a lower risk of coronary heart disease across all clinical risk categories, with the exception of blacks with prior coronary heart disease and no diabetes having higher risk than whites.48 It is unclear whether or not the lower risk of coronary heart disease may lead to a lower rate of CVD mortality rates in this population.

Some limitations of the present study should be acknowledged. First, one of the factors determining the quality of linkage is the uniqueness of identification of individuals being linked. A higher level of uniqueness is associated with more accurate linkage. The use of common Latino names (surnames and first names) and Asian/Pacific Islander surnames could lead to more false-positive matches and thus affect our ability to identify deaths of Hispanic and Asian/Pacific Islander enrollees. However, when we examined the success rates among 266,398 deaths documented within the KPSC system between 1988 and 2016 for each racial/ethnic subpopulation, the percentage of deaths not found by the linkage process did not differ much. Specifically, the rates were 2.5% for Hispanic, 1.4% for Asian/Pacific Islander, 1.3% for African Americans, and 0.8% for non-Hispanic whites. This is at least partially owing to a feature provided by the linkage software that takes care of the level of uniqueness of matching variables.

Second, deaths occurring outside California may not be completely captured, particularly after 2011, when a law was established that prohibited the SSA from disclosing state death records that the SSA receives through its contracts with the states. Given the size of the KPSC population and the lengthy study period, it was not feasible to identify deaths through the National Death Index. However, it is expected that most of the deaths outside California were reported to KPSC for active enrollees by family members, caregivers, doctors from medical facilities outside California, or law enforcement officers.

Third, the cause of death was missing for 27,187 (6.4%) of all deaths. These included deaths that were reported to KPSC or those that were derived through the linkage with SSA records but were not identified through the linkage process with the death records from the State of California. Therefore, the cause-specific death rates could be slightly underestimated.

Fourth, underlying causes of death may be underestimated for certain causes. For example, James et al49 found evidence that supported a larger number of deaths attributable to Alzheimer disease than what was actually reported.

Fifth, the change of underlying cause of death code from ICD-9 to ICD-10 in 1999 may affect the analyses related to cause of death. Anderson et al50 studied the influence of migration from ICD-9 to ICD-10 and concluded that the ranking of leading causes of death was substantially affected for some causes of death.

Sixth, the race/ethnicity information was missing for about 20% of the enrollees in 2001 to 2004 and was reduced to about 6% or 7% in recent years. Finally, although variation in all-cause mortality or cause-specific mortality exists in each ethnic group,18,51,52 our study did not stratify the analyses by subethnic groups.


The trends in age-adjusted mortality rates in this insured population are comparable to those of California and the US. The overall age-adjusted all-cause mortality rates are decreasing, although the cause-specific rates of certain diseases such as Alzheimer disease remained flat or increased during this period. v

a Available online at: www.thepermanentejournal.org/files/2019/18-213-Supp-Mat.pdf.

Disclosure Statement

The author(s) have no conflicts of interest to disclose.


We would like to thank Dianne Taylor for her assistance with formatting the manuscript.

Kathleen Louden, ELS, of Louden Health Communications performed a primary copy edit.

How to Cite this Article

Chen W, Yao J, Liang Z, et al. Temporal trends in mortality rates among Kaiser Permanente Southern California Health Plan enrollees, 2001-2016. Perm J 2019;23;18-213. DOI: https://doi.org/10.7812/TPP/18-213

Author Affiliations

1 Kaiser Permanente Southern California Research and Evaluation, Pasadena

2 Department of Cardiology, Sunset Medical Center, Los Angeles, CA

Corresponding Author

Wansu Chen, PhD (wansu.chen@kp.org)

1. de Souza HS, Fiocchi C. Immunopathogenesis of IBD: Current state of the art. Nat Rev Gastroenterol Hepatol 2016 Jan;13(1):13-27. DOI: https://doi.org/10.1038/nrgastro.2015.186.
 2. Kaplan GG, Ng SC. Understanding and preventing the global increase of inflammatory bowel disease. Gastroenterology 2017 Feb;152(2):313-21. DOI: https://doi.org/10.1053/j.gastro.2016.10.020.
 3. Bernstein CN, Shanahan F. Disorders of a modern lifestyle: Reconciling the epidemiology of inflammatory bowel diseases. Gut 2008 Sep;57(9):1185-91. DOI: https://doi.org/10.1136/gut.2007.122143.
 4. Hold GL. Western lifestyle: A ‘master’ manipulator of the intestinal microbiota? Gut 2014 Jan;63(1):5-6. DOI: https://doi.org/10.1136/gutjnl-2013-304969.
 5. Mowat C, Cole A, Windsor A, et al; IBD Section of the British Society of Gastroenterology. Guidelines for the management of inflammatory bowel disease in adults. Gut 2011 May; 60(5):571-607. DOI: https://doi.org/10.1136/gut.2010.224154.
 6. Chiba M, Nakane K, Komatsu M. Westernized diet is the most ubiquitous environmental factor in inflammatory bowel disease. Perm J 2019;23:18-107. DOI: https://doi.org/10.7812/TPP/18-107.
 7. Chiba M, Abe T, Tsuda H, et al. Lifestyle-related disease in Crohn’s disease: Relapse prevention by a semi-vegetarian diet. World J Gastroenterol 2010 May;16(20):2484-95. DOI: https://doi.org/10.3748/wjg.v16.i20.2484.
 8. Chiba M, Nakane K, Takayama Y, et al. Development and application of a plant-based diet scoring system for Japanese patients with inflammatory bowel disease. Perm J 2016 Fall;20(4):62-8. DOI: https://doi.org/10.7812/TPP/16-019.
 9. Chiba M, Tsuji T, Nakane K, et al. Induction with infliximab and plant-based diet as first-line (IPF) therapy in Crohn disease: Single-group trial. Perm J 2017;21:17-9. DOI: https://doi.org/10.7812/TPP/17-009.
 10. Chiba M, Nakane K, Tsuji T, et al. Relapse prevention in ulcerative colitis by plant-based diet through educational hospitalization: A single-group trial. Perm J 2018;22:17-167. DOI: https://doi.org/10.7812/TPP/17-167.
 11. Kornbluth A, Sachar DB; Practice Parameters Committee of the American College of Gastroenterology. Ulcerative colitis practice guidelines in adults: American College of Gastroenterology, Practice Parameters Committee. Am J Gastroenterol 2010 Mar;105(3):501-23. DOI: https://doi.org/10.1038/ajg.2009.727.
 12. Faubion WA Jr, Loftus EV Jr, Harmsen WS, Zinsmeister AR, Sandborn WJ. The natural history of corticosteroid therapy for inflammatory bowel disease: A population-based study. Gastroenterology 2001 Aug;121(2):255-60. DOI: https://doi.org/10.1053/gast.2001.26279.
 13. Peyrin-Biroulet L, Cieza A, Sandborn WJ, et al; International Programme to Develop New Indexes for Crohn’s Disease (IPNIC) group. Development of the first disability index for inflammatory bowel disease based on the international classification of functioning, disability and health. Gut 2012 Feb;61(2):241-7. DOI: https://doi.org/10.1136/gutjnl-2011-300049.
 14. Truelove SC, Witts LJ. Cortisone in ulcerative colitis; final report on a therapeutic trial. Br Med J 1955 Oct;2(4947):1041-8. DOI: https://doi.org/10.1136/bmj.2.4947.1041.
 15. Breslow L, Enstrom JE. Persistence of health habits and their relationship to mortality. Prev Med 1980 Jul;9(4):469-83. DOI: https://doi.org/10.1016/0091-7435(80)90042-0.
 16. Moum B, Ekbom A, Vatn MH, et al. Clinical course during the 1st year after diagnosis in ulcerative colitis and Crohn’s disease. Results of a large, prospective population-based study in southeastern Norway, 1990-93. Scand J Gastroenterol 1997 Oct;32(10):1005-12. DOI: https://doi.org/10.3109/00365529709011217.
 17. Henriksen M, Jahnsen J, Lygren I, et al, IBSEN Study Group. Ulcerative colitis and clinical course: Results of a 5-year population-based follow-up study (the IBSEN study). Inflamm Bowel Dis 2006 Jul;12(7):543-50. DOI: https://doi.org/10.1097/01.MIB.0000225339.91484.fc.
 18. Höie O, Wolters F, Riis L, et al; European Collaborative Study Group of Inflammatory Bowel Disease (EC-IBD). Ulcerative colitis: Patient characteristics may predict 10-yr disease recurrence in a European-wide population-based cohort. Am J Gastroenterol 2007 Aug;102(8):1692-701. DOI: https://doi.org/10.1111/j.1572-0241.2007.01265.x.
 19. Solberg IC, Lygren I, Jahnsen J, et al; IBSEN Study Group. Clinical course during the first 10 years of ulcerative colitis: Results from a population-based inception cohort (IBSEN Study). Scand J Gastroenterol 2009;44(4):431-40. DOI: https://doi.org/10.1080/00365520802600961.
 20. Satsangi J, Silverberg MS, Vermeire S, Colombel JF. The Montreal classification of inflammatory bowel disease: Controversies, consensus, and complications. Gut 2006 Jun;55(6):749-53. DOI: https://doi.org/10.1136/gut.2005.082909.
 21. Langholz E, Munkholm P, Davidsen M, Binder V. Course of ulcerative colitis: Analysis of changes in disease activity over years. Gastroenterology 1994 Jul:107(1):3-11. DOI: https://doi.org/10.1016/0016-5085(94)90054-x.
 22. Magro F, Rodrigues A, Vieira AL, et al. Review of the disease course among adult ulcerative colitis population-based longitudinal cohorts. Inflamm Bowel Dis 2012 Mar;18(3):573-83. DOI: https://doi.org/10.1002/ibd.21815.
 23. Romberg-Camps MJ, Dagnelie PC, Kester AD, et al. Influence of phenotype at diagnosis and of other potential prognostic factors on the course of inflammatory bowel disease. Am J Gastroenterol 2009 Feb;104(2):371-83. DOI: https://doi.org/10.1038/ajg.2008.38.
 24. Vester-Andersen MK, Vind I, Prosberg MV, et al. Hospitalization, surgical and medical recurrence rates in inflammatory bowel disease 2003-2011—
A Danish population-based cohort study. J Crohns Colitis 2014 Dec;8(12):1675-83. DOI: https://doi.org/10.1016/j.crohns.2014.07.010.
 25. Kitano A, Okawa K, Nakamura S, Komeda Y, Ochiai K, Matsumoto T. The long-term assessment of the patients with ulcerative colitis (> 10 years follow-up, mean follow-up 21.7 years) [in Japanese]. J New Remedies Clin 2011 Jul;60(7):1347-55.
 26. Burisch J, Pedersen N, Cukovic-Cavka S, et al; EpiCom Group. Initial disease course and treatment in an inflammatory bowel disease inception cohort in Europe: The EECO-EpiCom cohort. Inflamm Bowel Dis 2014 Jan; 20(1):36-46. DOI: https://doi.org/10.1097/01.MIB.0000436277.13917.c4.
 27. Liverani E, Scaioli E, Digby RJ, Bellanova M, Belluzzi A. How to predict clinical relapse in inflammatory bowel disease patients. World J Gastroenterol 2016 Jan; 22(3):1017-33. DOI: https://doi.org/10.3748/wjg.v22.i3.1017.
 28. Reinink AR, Lee TC, Higgins PD. Endoscopic mucosal healing predicts favorable clinical outcomes in inflammatory bowel disease: A meta-analysis. Inflamm Bowel Dis 2016;22:1859-69. DOI: https://doi.org/10.1097/MIB.0000000000000816.
 29. Yoshino T, Yamakawa K, Nishimura S, Watanabe K, Yazumi S. The predictive variable regarding relapse in patients with ulcerative colitis after achieving endoscopic mucosal healing. Intest Res 2016 Jan;14(1):37-42. DOI: https://doi.org/10.5217/ir.2016.14.1.37.
 30. Shi HY, Chan FK, Tsang SW, et al. Factors associated with mucosal healing in patients with ulcerative colitis in clinical remission. Inflamm Bowel Dis 2015 Apr;21(4):840-6. DOI: https://doi.org/10.1097/MIB.0000000000000334.
 31. Jowett SL, Seal CJ, Pearce MS, et al. Influence of dietary factors on the clinical course of ulcerative colitis: A prospective cohort study. Gut 2004 Oct;53(10):1479-84. DOI: https://doi.org/10.1136/gut.2003.024828.
 32. Tasson L, Canova C, Vettorato MG, Savarino E, Zanotti R. Influence of diet on the course of inflammatory bowel disease. Dig Dis Sci 2017 Aug;62(8):2087-94. DOI: https://doi.org/10.1007/s10620-017-4620-0.
 33. Monstad I, Hovde Ø, Solberg IC, Moum BA. Clinical course and prognosis in ulcerative colitis: Results from population-based and observational studies. Ann Gastroenterol 2014;27(2):95-104.
 34. Hyman MA, Ornish D, Roizen M. Lifestyle medicine: Treating causes of disease. Altern Ther Health Med 2009 Nov/Dec;15(6):12-4.
 35. Bodai BI, Nakata TE, Wong WT, et al. Lifestyle medicine: A brief review of its dramatic impact on health and survival. Perm J 2018;22:17-025. DOI: https://doi.org/10.7812/TPP/17-025.
 36. Chiba M, Nakane K, Komatsu M. Lifestyle medicine in inflammatory bowel disease [letter]. Perm J 2018 22:18-062. DOI: https://doi.org/10.7812/TPP/18-062.
 37. Desroches S, Lapointe A, Ratte S, Gravel K, Legatee F, Turcotte S. Interventions to enhance adherence to dietary advice for preventing and managing chronic diseases in adults. Cochrane Database Syst Rev 2013 Feb 28(2):CD008722. DOI: https://doi.org/10.1002/14651858.CD008722.pub2.
 38. Shai I, Schwarzfuchs D, Henkin Y, et al; Dietary Intervention Randomized Controlled Trial (DIRECT) Group. Weight loss with a low-carbohydrate, Mediterranean, or low-fat diet. N Engl J Med 2008 Jul;359(3):229-41. DOI: https://doi.org/10.1056/NEJMoa0708681.
 39. Atallah R, Filion KB, Wakil SM, et al. Long-term effects of 4 popular diets on weight loss and cardiovascular risk factors: A systematic review of randomized controlled trials. Circ Cardiovasc Qual Outcomes 2014 Nov;7(6):815-27. DOI: https://doi.org/10.1161/CIRCOUTCOMES.113.000723.

Keywords: death rate, leading cause of death, mortality trend, race- and ethnicity-specific death rate, sex-specific death rate


Click here to join the eTOC list or text ETOC to 22828. You will receive an email notice with the Table of Contents of The Permanente Journal.


2 million page views of TPJ articles in PubMed from a broad international readership.


Indexed in MEDLINE, PubMed Central, EMBASE, EBSCO Academic Search Complete, and CrossRef.




ISSN 1552-5775 Copyright © 2021 thepermanentejournal.org

All Rights Reserved