Association of Depressive Symptomatologywith Receipt of Informal Caregiving AmongOlder American Indians: The Native Elder

Care Study

Marc B. Schure, Ph.D., R. Turner Goins, Ph.D.

Received OctobePuget Sound HeCarolina Universpondence and r1100 Olive Way,

� 2014 Amehttp://dx.d

Am J Geriatr Ps

Objectives: Our study objectives were to identify the primary sources of informal

caregiving and to examine the association of depressive symptomatology with receipt

of informal caregiving among a sample of community-dwelling older American

Indians. Design: We conducted a cross-sectional study of older American Indians.

Participants: Community-dwelling adults aged 55 years and older who are members

of a federally recognized American Indian tribe in the Southeast United States.

Measurements: We collected information on the participant’s primary caregiver,

number of informal care hours received in the past week, depressive symptom-

atology, demographic characteristics, physical health status, and assistance need.

Results: Daughters, spouses, and sons were the most common informal primary

caregivers with distinct differences by sex of those receiving care. Compared with

participants with lower levels, those with a high level of depressive symptomatology

received substantially greater hours of informal care (33.4 versus 11.5 hours per

week). Conclusions: Older American Indians with higher levels of depressive

symptomatology received more informal caregiving than those with lower depressive

symptomatology. The burden of caregiving of older adults is primarily shouldered by

spouses and children with those who care for older adults with depressive symp-

tomatology likely experiencing an even greater burden of care. (Am J Geriatr Psy-chiatry 2014; -:-e-)

Key Words: American Indians, depressive symptomatology, informal caregiving

ccording to the World Health Organization, by

A2020 major depression will surpass many otherserious diseases as the second leading contributor tothe global burden of disease.1 The prevalence ofclinically significant depressive symptoms among

r 23, 2013; revised March 12, 2014; accepted March 31, 20alth Care System, Seattle, WA; and the Department of Sosity, Cullowhee, NC, and Center for Healthy Aging, Moeprint requests to Marc B. Schure, Ph.D., Health ServicesSte. 1400, Seattle, WA 98101. e-mail: Mark.Schure@va.grican Association for Geriatric Psychiatryoi.org/10.1016/j.jagp.2014.03.013

ychiatry -:-, - 2014

older community-dwelling adults in the UnitedStates typically ranges from 8%e16%,2 with occur-rence varying primarily by demographic character-istics. For example, some studies have foundsubstantially higher prevalence among certain racial

14. From the Health Services Research and Development (MBS), VAcial Work, College of Health and Human Sciences (RTG), Westernuntain Area Health Education Center, Asheville, NC. Send corre-Research and Development, VA Puget Sound Health Care System,ov

1

The Native Elder Care Study

and ethnic minority populations.3,4 Late life depres-sion is associated with increased number of chronichealth conditions and physical disabilities.5e8

The presence of comorbid depression appears tohave a substantive impact on receipt of informalcaregiving independent of physical health status.Informal caregiving is unpaid care provided by familymembers and friends to those in need of assistancewith daily activities. Several studies have found thatpersons with chronic health conditions and physicaldisabilities that co-occur with depressive symptomsreceive more informal and formal care than thosewithout comorbid depressive symptoms.9e12 In fact,adults aged 70 years and older with depressivesymptoms have been found to receive twice thenumber of informal caregiving hours than thosewithout depressive symptoms.10 Moreover, addi-tional research has indicated that after controlling fordisease severity, comorbid depression was associatedwith greater outpatient service use and approximately50% increased cost in medical care among adults inthe United States.9,13

The availability of informal caregivers is an impor-tant resource delaying and/or preventing institu-tionalization,14 and therefore identifying caregivingneed is a key component in achieving most olderadults’ desire to “age in place.”Aging in place denotesolder adults’ capacity for continued residence at homeeven until advanced age.15 Informal caregiving needis often determined by the number of activities of dailyliving (ADLs) and/or instrumental activities of dailyliving (IADLs) in which a person has difficulty per-forming. Over 6million adults aged 65 years and olderin the United States need some form of caregiving,16

with approximately two-thirds of them eventuallyneeding such care for an average of two years.17

Adults in the United States aged 65 years and olderwho died at home between 2000 and 2002 received anaverage of 65.8 informal caregiving hours per weekduring their last year of life.18

STUDY OBJECTIVE

Older American Indians experience some of thehighest rates of physical disability19e21 and dispro-portionately suffer from poorer physical and mentalhealth compared with other racial and ethnic pop-ulations.22,23 Evidence suggests that the prevalence of

2

clinically significant depressive symptomatologyamong older American Indians is higher than that ofsame-aged adults of other races and ethnicities.4,24

No identified studies have examined the associationof depressive symptomatology with the receipt ofinformal care among older American Indians. Thereis a critical need to assess receipt of informal care inthis vulnerable population, as evidenced by a highunmet need for long-term care.25,26 Thus, our studyobjectives were to identify the primary sources ofinformal caregiving and to examine the association ofdepressive symptomatology with receipt of informalcaregiving among a sample of community-dwellingolder American Indians.

METHODS

Study Design and Data Collection

Data for this study originate from the Native ElderCare Study, a cross-sectional study of community-dwelling older adult members of a federally recog-nized American Indian tribe located in the Southeastregion of the United States.27 Data were collected from2006 to 2008 using in-person interviewer-adminis-tered surveys and included information about de-mographic characteristics, physical disability, mentaland physical health, personal assistance needs, healthcare use, and psychosocial factors. The tribe’s insti-tutional review board, tribe’s health board, tribalcouncil, tribal elder council, and the West VirginiaUniversity institutional review board approved theproject. All study participants provided informedconsent and received a $20 gift card for completing theinterview. The Oregon State University institutionalreview board approved the secondary data analysesfor this study.

Sample

Study inclusion criteria were being an enrolledtribal member, aged 55 years or older, being a resi-dent in the tribal service area, non-institutionalized,and having passed a cognitive screen. We used anage threshold of 55 years rather than 65 years becauseit was requested by the project’s tribal stakeholdersand data has shown rapid declines in health statusand shorter life expectancy among American Indianscompared with other racial and ethnic groups.28,29

Am J Geriatr Psychiatry -:-, - 2014

Schure and Goins

Based on age and residential location, we generated alist of 1,430 eligible tribal members from tribalenrollment records. To ensure equal representationacross the range of ages, an age-stratified randomsample was taken of 680 tribal members from thefollowing age groups: 55e64 years, 65e74 years, and75 years and older. We recruited study participantsby a telephone call or a visit to their home. Of the 680individuals, 47 could not be located and 50 weredeemed ineligible. Of the remaining 583 persons, 78declined participation, resulting in a total of 505participants and an 86.6% response rate. Of the 195persons who had an assistance need, 180 werereceiving informal care at the time of this study.

Measures

Sources and receipt of informal caregiving. Weidentified the most predominantly used caregivers byasking participants who helps them the most (amongthe unpaid persons), with a list including spouse/significant other, daughter, daughter-in-law, son,son-in-law, granddaughter, grandson, neighbor, andother (specifying the relationship to the respondent).Receipt of informal caregiving was measured as thereported number of informal care hours received perweek by study participants with disability-associatedassistance needs. Participants who had difficulties inperforming one or more ADLs or IADLs were askedthe number of hours of unpaid per week they receivefor these activities.

Depressive symptomatology. Depressive symptom-atology was assessed with the Centers for Epidemi-ologic StudieseDepression (CES-D) scale.30 This scalehas been widely used in population-based studiesand its validity and reliability has been confirmedamong older adults and across different racial andethnic groups,31 including older American Indians.32

Participants were asked how often they felt eachsymptom in the past week, with sample itemsincluding (1) I felt fearful, (2) I enjoyed life, and (3)people were unfriendly. The response optionsincluded: 0 ¼ rarely or none of the time; 1 ¼ some ora little of the time; 2 ¼ occasionally or a moderateamount of time; and 3 ¼ most or all of the time.Positive affect items were reverse coded. Therefore,the total sum score ranges from 0 to 60 comprisingboth the count and frequency of experiencing each ofthe CES-D items. For our study sample, the internal

Am J Geriatr Psychiatry -:-, - 2014

consistency was high (a ¼ 0.89). Based on past rec-ommendations, we used a tiered (low: 0e9, moder-ate: 10e19, and high: �20) approach for identifyingsubsyndromal depression in addition to identifyinghigher depressive symptomatology.33

Demographic characteristics. Demographic charac-teristics included age, sex, marital status (married/partner versus unmarried), educational attainment(<12 years versus �12 years), and living arrange-ments (living alone versus living with others).

Physical health conditions. Physical health condi-tions included physical disability and number ofchronic health conditions. We defined physicaldisability as havingdifficultywith one ormore of eightADLs and/or eight IADLs.34,35 The ADLs we exam-ined included bathing/showering, dressing, eating,transferring, walking, toileting, grooming, and gettingoutside and the IADLs we examined included usingthe telephone, light housework, heavy housework,preparing meals, shopping, managing money, man-aging medications, and transportation. We assessedthe number of chronic health conditions among 12common self-reported chronic medical conditions,which included heart disease, stroke, angina, conges-tive heart failure, heart attack, lung disease, Parkinsondisease, cancer, diabetes, high blood pressure, kidneydisease, and liver disease. Respondents were asked if,since age 55, a doctor had told them they had one of thelisted 12 conditions with “yes” and “no” responseoptions.

Assistance need. For each of the eight ADLs andeight IADLs, we classified respondents as having anassistance need versus no assistance need. Thosereporting having some level of difficulty performinga task but did not report a need for assistance withthat activity were classified as having no assistanceneed. Those reporting some level of difficulty per-forming an activity and reporting needing assistanceor more assistance for that activity were classified ashaving an assistance need. Then, we created a countvariable of the number of ADL and/or IADL assis-tance needs with a score range from 0 to 16.

Statistical Analyses

First, we generated numbers and weighted per-centages of sample characteristics across tiers (low,moderate, high) of depressive symptomatology. In-dividual responses were weighted to account for

3

TABLE 1. Sample Characteristics by Depressive Symptomatology (N [ 170)

Sample Characteristics

Depressive Symptomatology

p ValueLow (0e9) (N [ 108) Moderate (10e19) (N [ 42) High (‡20) (N [ 20)

Percent (Number)Age, years 0.237

55e64 51.1 (23) 33.3 (15) 15.6 (7)65e74 69.8 (37) 17.0 (9) 13.2 (7)�75 66.7 (48) 25.0 (18) 8.3 (6)

Sex 0.625Male 58.8 (27) 33.6 (10) 7.6 (3)Female 60.8 (81) 24.3 (32) 14.9 (17)

Marital Status 0.978Married/partner 60.0 (40) 26.5 (15) 13.5 (7)Divorced/separated/widowed 60.1 (68) 26.9 (27) 12.4 (13)

Living arrangement 0.306Lives alone 68.6 (33) 25.3 (10) 6.2 (3)Lives with others 58.5 (74) 26.2 (31) 15.4 (17)

Educational attainment 0.028<12 years 50.1 (46) 36.2 (28) 13.2 (9)�12 years 68.2 (62) 18.6 (14) 12.9 (11)

Number of ADL difficulties 0.0070 ADLs 92.7 (26) 3.1 (1) 4.2 (2)1e2 ADLs 59.4 (52) 24.4 (20) 16.1 (12)�3 ADLs 44.8 (30) 41.9 (21) 13.3 (6)

Number of IADL difficulties 0.0010 IADLs 82.2 (13) 7.9 (2) 9.9 (2)1e2 IADLs 78.0 (61) 16.3 (12) 5.7 (5)�3 IADLs 37.8 (34) 41.0 (28) 21.2 (13)

Number of chronic conditions 0.3730 62.0 (15) 24.4 (3) 13.7 (1)1e2 67.6 (55) 24.0 (17) 8.3 (10)�3 46.7 (38) 33.7 (22) 19.6 (9)

Heart disease 0.401Yes 59.3 (25) 21.5 (8) 11.4 (7)No 61.4 (83) 27.2 (33) 19.2 (13)

Stroke 0.537Yes 43.0 (13) 36.4 (7) 11.9 (4)No 64.0 (95) 24.0 (34) 20.6 (16)

Angina 0.110Yes 65.7 (11) 23.2 (8) 11.1 (5)No 40.6 (96) 33.7 (31) 25.7 (15)

Congestive heart failure 0.107Yes 64.2 (12) 23.0 (10) 12.8 (4)No 45.9 (95) 37.6 (30) 16.5 (16)

Heart attack 0.160Yes 64.8 (11) 22.9 (9) 12.3 (3)No 41.1 (95) 42.0 (31) 17.0 (16)

Lung disease 0.720Yes 61.8 (16) 26.4 (5) 11.8 (4)No 55.7 (92) 22.8 (36) 21.5 (16)

Parkinson disease 0.750Yes 100.0 (1) 0.0 (0) 0.0 (0)No 60.8 (107) 26.0 (41) 13.2 (20)

Cancer 0.240Yes 63.2 (19) 21.8 (11) 14.9 (2)No 50.3 (89) 44.7 (30) 5.0 (18)

Diabetes 0.730Yes 60.8 (51) 25.7 (17) 13.5 (8)No 61.1 (57) 26.1 (24) 12.9 (12)

High blood pressure 0.330Yes 68.1 (75) 22.2 (30) 9.7 (17)No 59.1 (33) 26.2 (10) 14.7 (3)

(Continued)

4 Am J Geriatr Psychiatry -:-, - 2014

The Native Elder Care Study

TABLE 1. (Continued)

Sample Characteristics

Depressive Symptomatology

p ValueLow (0e9) (N [ 108) Moderate (10e19) (N [ 42) High (‡20) (N [ 20)

Kidney disease 0.360Yes 61.2 (10) 26.5 (4) 12.3 (4)No 60.5 (97) 18.2 (36) 21.2 (16)

Liver disease 0.380Yes 60.6 (2) 0.0 (0) 25.2 (1)No 74.8 (106) 26.4 (41) 12.9 (19)

Assistance need, hours per week 0.0010 84.8 (14) 10.1 (3) 5.0 (1)1e2 81.6 (51) 11.1 (6) 7.3 (5)3e4 45.3 (19) 38.6 (13) 16.1 (5)�5 38.4 (24) 41.4 (20) 20.1 (9)

Notes: Percentages are weighted per age-stratified sampling design. p value significance levels based on c2 tests. Degrees of freedom ¼ 6 forassistance need; 4 for age, number of ADL and IADL difficulties, and chronic conditions; and 2 for all other sample characteristics. Ten casesfrom the sample were missing information on depressive symptomatology.

FIGURE 1. Percent of sample using informal care by primary caregiver.

Notes: Percentages are weighted per age-stratified sampling design.

Schure and Goins

differential sampling rates across the three agegroups. We used c2 tests to evaluate the bivariate re-lationships of depressive symptomatology with eachof the sample characteristics. We then used Poissonregression models to compare the unadjusted andadjusted relationship between depressive symptom-atology and number of informal care hours received.Predicted unadjusted and adjusted mean number ofinformal care hours was then computed using

Am J Geriatr Psychiatry -:-, - 2014

StataCorp statistical software’smargins command andcorresponding effect sizes were calculated using theCohen’s d effect size calculation method. Specifically,we adjusted for physical disability, chronic healthconditions, assistance need, and demographic char-acteristics. We dropped cases for which informationon the number of informal caregiver hours receivedwas completely missing, and imputed missing dataon the remaining variables36 with the multiple

5

FIGURE 2. Poisson models showing adjusted and adjustedmean number of informal care hours received perweek by levels of depressive symptomatology and95% confidence intervals. [A] Unadjusted meannumber of informal care hours. [B] Adjusted meannumber of informal care hours.

The Native Elder Care Study

imputation chained equations method.37 We thenestimated the variance inflation factor to test formulticollinearity among the independent variables,which was not found to be a substantive issue. Of the180 subjects in the study sample, 10 had completemissing data on depressive symptomatology. Weused t tests to compare the completely missing caseson the dependent variable with the others by de-mographic variables. Comparisons from these t testsshowed that those with completely missing data onthe dependent variable were more likely to be ofyounger age (t ¼ �3.21, df ¼ 260, p ¼ 0.002), male sex(t ¼ �2.64, df ¼ 260, p ¼ 0.009), and have highereducational attainment (t ¼ 2.69, df ¼ 259, p ¼ 0.008).All analyses were conducted using StataCorp’s sta-tistical software package version 12.0.38

Notes: Poisson regression models indicate that the associationbetween depressive symptomatology and number of informal carehours received is statistically significant for both the unadjusted(IRR ¼ 1.74, df ¼ 177, p <0.001) and adjusted model (IRR ¼ 1.70,df ¼ 177, p <0.001). For the adjusted model, Cohen’s d effects sizesby tiers are: low to moderate (d ¼ �0.28), moderate to high(d ¼ �0.17), and low to high (d ¼ �0.28).

RESULTS

The study sample characteristics by tiers ofdepressive symptomatology are presented in Table 1.Of the 170 participants, 60.3% had low, 26.6% hadmoderate, and 13.1% had high depressive symptom-atology. Depressive symptomatology varied bycertain demographic and physical health characteris-tics. Specifically, we found that tiers of depressivesymptomatology were significantly associated witheducational attainment (c2 ¼ 7.15, df ¼ 2, p ¼ 0.028),ADL difficulties (c2 ¼ 14.06, df ¼ 4, p ¼ 0.007), IADLdifficulties (c2 ¼ 19.62, df¼ 4, p ¼ 0.001), and numberof assistance needs (c2 ¼ 21.68, df ¼ 6, p ¼ 0.001).

Among respondents in our analytic sample, 82.2%reported receiving informal care versus 35.5%receiving formal (paid) care; only 3.9% indicatedreceiving care from both. Two-thirds (n ¼ 120) of oursample reported the number of people they receivedcare from either informal or formal care, or both.Results indicated that 8.3% did not receive care,30.8% received care from one person, 44.7% from twoor three persons, and 16.7% from 4 or more persons.c2 analyses indicated that greater number of care-givers was associated with number of assistanceneeds (c2 ¼ 43.51, df ¼ 9; p <0.001), ADL difficulties(c2 ¼ 128.24, df ¼ 6, p ¼ 0.006), and IADL difficulties(c2 ¼ 229.05, df ¼ 6, p ¼ 0.007).

Figure 1 shows the percent of those who receivedinformal care by type of their primary caregiver. Themost common sources of informal caregiving were

6

daughters, spouses, and sons. Some differences werefound in respect to the proportions ofmen andwomenreporting primary caregiver source. Of those indi-cating spouse as the primary caregiver, 55.9% were

Am J Geriatr Psychiatry -:-, - 2014

Schure and Goins

women and 44.1% were men. For those indicatingother sources other than spouse as primary caregiver,the proportion of women was substantially greaterthan the proportion of men, with women accountingfor 100% of persons indicating son- or daughter-in-laws and neighbors as their primary caregiver.

Poisson regression results showed a significantunadjusted and adjusted positive association be-tween depressive symptomatology and informal careuse hours with corresponding incident rate ratios(IRRs) and 95% confidence intervals (CIs). Specif-ically, the unadjusted IRR was 1.74 (95% CI ¼ 1.27,2.38; df ¼ 177, p <0.001). The adjusted IRR was 1.70(95% CI ¼ 1.23, 2.36; df ¼ 177, p ¼ 0.001). Unadjustedand adjusted mean number of informal care hoursreceived per week by levels of depressive symp-tomatology are presented in Figures 2a and 2b,respectively. The unadjusted mean number ofinformal care hours received in the past week by low,moderate, and high levels of depressive symptom-atology was 11.4 (95% CI ¼ 7.76, 15.10; z ¼ 5.32,p <0.001), 19.9 (95% CI ¼ 14.73, 25.05; z ¼ 6.71,p <0.001), and 34.6 (95% CI ¼ 18.24, 51.00; z ¼ 3.96,p <0.001) hours, respectively. When adjusting fordemographic characteristics, physical disability,chronic health conditions, and assistance need, themean number of informal care hours received perweek by those with low, moderate, and highdepressive symptomatology was 11.5 (95% CI ¼ 7.89,15.08; z ¼ 4.48, p <0.001), 19.6 (95% CI ¼ 14.51, 24.65;z ¼ 5.48, p <0.001), and 33.4 (95% CI ¼ 16.92, 49.85;z ¼ 2.95, p <0.001) hours, respectively. Correspond-ing effect sizes of adjusted informal care use hours bytiers of depressive symptomatology were low.Comparisons across each tier generated the followingeffects sizes for the adjusted model: low to moderate(d ¼ �0.28), moderate to high (d ¼ �0.17), and low tohigh (d ¼ �0.28).

CONCLUSION

Our data show the majority of our sample of olderAmerican Indians received care primarily from rela-tives, with few using a combination of informal andformal help. Also, more than 60% of our sampleindicated that they receive help from two or morepersons. Our findings support others that indicatethe burden of care for older adults is primarily that

Am J Geriatr Psychiatry -:-, - 2014

of immediate family members.39e42 The mostcommonly used sources of informal care found inour sample are mostly consistent with previousfindings of informal care networks for oldercommunity-dwelling adults across different racesand ethnicities.40e42 For example, one study showedthat spouses, daughters, and sons were the primarycaregiver sources among U.S. veterans.40 Amongcurrent Medicare beneficiaries, these findingsremained consistent with the noteworthy exceptionthat sons were least likely to be a primary caregiver,below that of friends and other relatives.41 Datafrom the 1989 and 1999 National Long-Term CareSurvey and Informal Caregiver Survey show howprimary caregiver trends have changed over timewith children taking on a greater role compared withspouses.42

We also found that those with a greater number ofdepressive symptoms received a greater amount ofinformal caregiving. Specifically, we found that un-adjusted and adjusted mean number of receivedinformal care hours more than doubled betweenthose with low versus those with high depressivesymptomatology. Only one known population-basedstudy examined the effect of depressive symptoms oninformal care use among the oldest old (�70 years),and it found that those with higher depressivesymptomatology received twice the number ofcaregiving hours compared with those with lowerdepressive symptomatology.10

Plausible explanations for why greater depressivesymptoms lead to greater informal care can best beunderstood with the conceptual framework of theDisablement Process Model,43 whereby greaterdepressive symptomatology expedites pathology’seffect on physical disability and disease severity. Evi-dence supports the idea that depressive symptom-atology has direct adverse physiologic effects, such asincreased inflammationanddecreased immunity levelsaswell as indirect impacts fromdecreased adherence tohealthy lifestyle behaviors (i.e., diet, exercise, medica-tion-taking).9,44,45 Also, a review of existing evidencefound a bidirectional relationship between greaterdepressive symptomatology and physical disability,46

with each found to be independently predictive ofinformal caregiver burden.12 In sum, there is substan-tive evidence to implicate a number of mechanisms bywhich depressive symptomatologymay lead to greaterinformal caregiving use.

7

The Native Elder Care Study

Although not statistically significant, we foundgreater prevalence of depressive symptoms amongthe younger age bracket of our sample. Some evi-dence supports and helps to explain the inverserelationship between age and depressive symptom-atology by showing that social, economic, and psy-chosocial factors correspond with different cohorts ofolder adults.47 Specifically, one study showed thatlower economic hardship and fewer negative re-lationships helped explain the lower prevalence ofdepressive symptomatology among the oldest old.Conversely, this same study showed that as personalmastery decreased with age, it suppressed themagnitude of this trend.47 Such findings align withthe Disablement Process Model43 where psychosocialfactors may either exacerbate or slow the processtowards physical disability and need for care.

Greater informal care receipt among those withhigher levels of depressive symptomatology posesinteresting scientific questions regarding currentAmerican Indian cultural values and practices.Traditional value orientations in American Indiancommunities tend to foster interdependence, ratherthan autonomy, through extended family systems.48

Such value systems may translate into greater reli-ance on informal caregiving for assistance needsamong older adults. We do not know if these valuesystems still function today as they once did, how-ever, and future studies are warranted to examine theextent to which these cultural value systems functiontoday to the benefit of aging American Indians.Similarly, it would be worthwhile determining ifvalues supporting both the identification of andresponse to older adults with higher levels ofdepressive symptomatology differ across racial andethnic groups. An important question remainswhether other factors, such as cultural values andtraditions, act as mediating mechanisms in theobserved associations between higher depressivesymptomatology and greater informal care use.Qualitative studies examining such questions wouldoffer useful theoretical models for quantitative re-searchers who seek to empirically test such questions.

Our results should be viewed in the context ofsome limitations. As a cross-sectional study, we wereunable to establish causality. Longitudinal studieswill contribute to our understanding of depressivesymptoms’ relative contribution to caregiving useand need. Other omitted variables could have been

8

associated with receipt of informal caregiving anddepressive symptomatology and hence reveal aspurious relationship between our main independentand dependent variable. We believe, however, thatcontrolling for assistance need greatly substantiatedour findings. It is particularly worth noting that allthe data we examined were based on self-report andtherefore subject to participant bias. It has beenposited that specific American Indian values couldlead them not to admit or report an assistanceneed,49,50 which may have led our study participantsto have under-reported both use and need of care.Another related concern is differential recall ofdepressive symptoms. Some evidence exists to sug-gest that persons with higher levels of distress aremore likely to over-report use of services comparedwith those with low levels of distress51,52 and suchbias would inflate our findings. Similarly, althoughthe differences in mean number of informal carehours was substantial among those with differentlevels of depressive symptoms, the large variabilityfound among those with high depressive symptomsmay have reduced the calculated effect sizes. Finally,our results are limited to older American Indianadults of a single tribe; therefore, caution should beexercised in generalizing these results to oldermembers of other tribes.

In sum, our study shows that older American In-dians with physical disabilities and higher depressivesymptomatology use greater informal care comparedwith those without or with lower depressive symp-tomatology. Evidence suggests that American Indiancaregivers, compared with American Indians whoare not caregivers, have poorer mental and physicalhealth.53 Compared with their racial and ethniccounterparts, American Indian caregivers may faceadditional stressors such as poverty and decreasedaccessibility to services.54 Needs assessment studieswill help guide future programming and policiesaimed at the provision of long-term care services andaid of informal caregivers’ burden of care. The NativeAmerican Caregiver Support Program, funded by theAdministration on Aging, is designed to offer infor-mational and instrumental support to American In-dian caregivers.55 One non-traditional interventionprogram has shown promising effects on the healthof American Indian caregivers.56 Future research isneeded to track the effectiveness of these types ofprograms on long-term health outcomes on

Am J Geriatr Psychiatry -:-, - 2014

Schure and Goins

caregivers and those whom are receiving care. Simi-larly, little is known about the effectiveness ofdepression treatment interventions among olderAmerican Indians.57 Future studies are warranted toevaluate the impact of existing depression treatmentsin this population.

Although our study focused on informal caregivinguse among American Indians, seen in conjunctionwith existing evidence, we believe that our findingswarrant greater attention for all older adult pop-ulations with caregiving needs. Specifically, werecommend that clinicians and policy makers payclose attention to the increased care needs andinformal caregiver burden of those caring forolder adults with high levels of depressive

Am J Geriatr Psychiatry -:-, - 2014

symptomatology. A broad spectrum of interventions,including stress-reduction strategies and specializedtraining to deal with the special needs of adults withpoor mental health, are recommended to improveoutcomes for the caregiver and the care receiver.58

Provided current aging and health trends, subse-quent generations will be presented with tremendouseconomic and social challenges of providing adequateand quality long-term care to older adults.

This study was funded in part from the NationalInstitute of Aging, NIH (# AG022336) and from OregonState University’s College of Public Health and HumanSciences. We would like to thank Drs. Adam Branscumand Alan Acock for providing analytic recommendations.

References

1. World Health Organization: Mental health: Depression. 2012.(online). Available at: http://www.who.int/mental_health/management/depression/definition/en/. Accessed September24, 2012

2. Blazer DG: Depression in late life: review and commentary.J Gerontol A Biol Sci Med Sci 2003; 58:M249eM265

3. González HM, Haan MN, Hinton L: Acculturation and the preva-lence of depression in older Mexican Americans: baseline resultsof the Sacramento Area Latino Study on Aging. J Am Geriatr Soc2001; 49:948e953

4. Curyto KJ, Chapleski EE, Lichtenberg PA, et al: Prevalence andprediction of depression in American Indian elderly. Clin Ger-ontol 1998; 18:19e37

5. Alexopoulos GS: Depression in the elderly. Lancet 2005; 365:1961e1970

6. Brault MW, Hootman J, Helmick CG, et al: Prevalence and mostcommon causes of disability among adults—United States, 2005.MMWR 2009; 58:421e426

7. Chapman DP, Perry GS, Strine TW: The vital link between chronicdisease and depressive disorders. Prev Chronic Dis 2005; 2:1e10

8. Gill TM, Allore HG, Holford TR, et al: Hospitalization, restrictedactivity, and the development of disability among older persons.JAMA 2004; 292:2115e2124

9. Katon WJ: Clinical and health services relationships betweenmajor depression, depressive symptoms, and general medicalillness. Biol Psychiatry 2003; 54:216e226

10. Langa KM, Valenstein MA, Fendrick AM, et al: Extent and cost ofinformal caregiving for older Americans with symptoms ofdepression. Am J Psychiatry 2004; 161:857e863

11. Badger TA: Depression, physical health impairment and serviceuse among older adults. Public Health Nurs 2007; 15:136e145

12. McCullagh E, Brigstocke G, Donaldson N, et al: Determinants ofcaregiving burden and quality of life in caregivers of stroke pa-tients. Stroke 2005; 36:2181e2186

13. Luber MP, Meyers BS, Williams-Russo PG, et al: Depression andservice utilization in elderly primary care patients. Am J GeriatrPsychiatry 2001; 9:169e176

14. Davey A, Femia EE, Zarit SH, et al: Life on the edge: patterns offormal and informal help to older adults in the United States andSweden. J Gerontol B Psych Sci Soc Sci 2005; 60:S281eS288

15. Farber N, Shinkle D: Aging in Place: A State Survey of LiveabilityPolicies and Practices. Washington DC, AARP Public PolicyInstitute, 2011

16. Kaye HS, Harrington C, LaPlante MP: Long-term care: Who gets it,who provides it, who pays, and how much? Health Affairs 2010;29:11e21

17. Wysocki A, Butler M, Kane RL, et al: Long-term Care for OlderAdults: A Review of Home and Community-Based Services versusInstitutional Care. Rockville, MD, Agency for Healthcare Researchand Quality, 2012

18. Rhee YJ, Degenholtz HB, Lo Sasso AT, et al: Estimating thequantity and economic value of family caregiving for community-dwelling older persons in the last year of life. J Am Geriatr Soc2009; 57:1654e1659

19. Denny CH, Holtzman D, Goins RT, et al: Disparities in chronicdisease risk factors and health status between American Indian/Alaska Native and White elders: findings from a telephone survey,2001 and 2002. Am J Public Health 2005; 95:825e827

20. Goins RT, Moss M, Buchwald D, et al: Disability among olderAmerican Indians and Alaska Natives: an analysis of the 2000 censuspublic use microdata sample. Gerontology 2007; 47:690e696

21. Moss MP, Schell MC, Goins RT: Using GIS in a first nationalmapping of functional disability among older American Indiansand Alaska natives from the 2000 census. Int J Health GeogrSeptember 1, 2006:1e11

22. Barnes PM, Powell-Griner E, Adams PF: Health characteristics ofthe American Indian and Alaska Native adult population, UnitedStates, 1999e2003. Adv Data 2005; 356:1e24

23. Centers for Disease Control and Prevention: CDC Health Dis-parities and Inequalities Report—United States, 2011. Atlanta,GA, Centers for Disease Control and Prevention, 2011

24. John R, Kerby DS, Hennessy CH: Patterns and impact of comor-bidity and multimorbidity among community-resident AmericanIndian elders. Gerontology 2003; 43:649e660

25. Goins RT, Bogart A, Roubideaux Y: Service provider perceptions oflong-term care access in American Indian and Alaska Native com-munities. J Health Care Poor Underserved 2010; 21:1340e1353

26. Schure MB, Conte KC, Goins RT: Unmet assistance need amongolder American Indians with physical disabilities: the Native ElderCare Study [Epub ahead of print January 31, 2014]. Gerontology

9

The Native Elder Care Study

27. Goins RT, Garroutte EM, Fox SL, et al: Theory and practice inparticipatory research: lessons from the Native Elder Care Study.Gerontology 2011; 51:285e294

28. Hayward MD, Heron M: Racial inequality in active life amongadult Americans. Demography 1999; 36:77e91

29. Indian Health Service: Facts on Indian Health Disparities. Rock-ville, MD, U.S. Department of Health and Human Services, 2013

30. Radloff LS:TheCES-D scale: a self-reportdepression scale for researchin the general population. Appl Psych Meas 1977; 1:385e401

31. Mui A, Burnette D, Chen L: Cross-cultural assessment of geriatricdepression: a review of the CES-D and GDS, in MulticulturalMeasurement in Older Populations. Edited by Skinner JH,Teresi JA, Holmes D, et al. New York, NY, Springer PublishingCo., 2002, pp 147e178

32. Chapleski EE, Lichtenberg PA, Dwyer JW, et al: Morbidity andcomorbidity among Great Lakes American Indians: predictors offunctional ability. Gerontology 1997; 37:588e597

33. Barry LC, Allore HG, Bruce ML, et al: Longitudinal associationbetween depressive symptoms and disability burden among olderpersons. J Gerontol A Biol Sci Med Sci 2009; 64:1325e1332

34. Fillenbaum GG: Multidimensional Functional Assessment of OlderAdults: The Duke Older Americans Resources and Services Pro-cedures. Hillsdale, NJ, L. Erlbaum Associates, 1988

35. Lawton MP, Brody EM: Assessment of older people: self-maintaining and Instrumental Activities of Daily Living. NursRes 1970; 19:278

36. von Hippel PT: Regression with missing Ys: an improved strategyfor analyzing multiply imputed data. Sociol Methodol 2007; 37:83e117

37. Royston P, White IR: Multiple Imputation by Chained Equations(MICE): Implementation in Stata. J Stat Soft 2011; 45:1e20

38. Stata Statistical Software: Version 12 [computer program]. Col-lege Station, TX, StataCorp LP, 2012

39. Schulz R, Martire LM: Family caregiving of persons with demen-tia: prevalence, health effects, and support strategies. Am J Ger-iatr Psychiatry 2004; 12:240e249

40. Van Houtven CH, Oddone EZ, Weinberger M: Informal andformal care infrastructure and perceived need for caregivertraining for frail US veterans referred to home and community-based services. Chronic Illn 2010; 6:57e66

41. Van Houtven CH, Taylor DH Jr, Steinhauser K, et al: Is a home-care network necessary to access the Medicare hospice benefit?J Palliat Med 2009; 12:687e694

42. Wolff JL, Kasper JD: Caregivers of frail elders: updating a nationalprofile. Gerontology 2006; 46:344e356

43. Verbrugge LM, Jette AM: The disablement process. Soc Sci Med1994; 38:1e14

10

44. Segerstrom SC, Miller GE: Psychological stress and the humanimmune system: a meta-analytic study of 30 years of inquiry.Psychol Bull 2004; 130:601e630

45. Zorrilla EP, Luborsky L, McKay JR, et al: The relationship ofdepression and stressors to immunological assays: a meta-analyticreview. Brain Behav Immun 2001; 15:199e226

46. Lenze EJ, Rogers JC, Martire LM, et al: The association of late-lifedepression and anxiety with physical disability: a review of theliterature and prospectus for future research. Am J Geriatr Psy-chiatry 2001; 9:113e135

47. Schieman S, Van Gundy K, Taylor J: The relationship betweenage and depressive symptoms: a test of competing explana-tory and suppression influences. J Aging Health 2002; 14:260e285

48. Red Horse J: Traditional American Indian family systems. Fam SysHealth 1997; 15:243e250

49. Loftin JD: The “harmony ethic” of the conservative EasternCherokees: a religious interpretation. J Cherokee Stud 1983; 8:40e43

50. Moss MP: TOLERATED ILLNESS� concept and theory forchronically ill and elderly patients as exemplified in AmericanIndians. J Cancer Educ 2005; 20(suppl 1):17e22

51. Rhodes AE, Fung K: Self-reported use of mental health servicesversus administrative records: care to recall? Int J Method PsychRes 2004; 13:165e175

52. Rhodes AE, Lin E, Mustard CA: Self-reported use of mental healthservices versus administrative records: should we care? Int JMethod Psych Res 2002; 11:125e133

53. Spencer SM, Goins RT, Henderson J, et al: Influence of caregivingon health-related quality of life among American Indians. J AmGeriatr Soc 2013; 61:1615e1620

54. Jervis LL, Jackson MY, Manson SM: Need for, availability of, andbarriers to the provision of long-term care services for olderAmerican Indians. J Cross Cult Gerontol 2002; 17:295e311

55. Administration on Aging. National Family Caregiver SupportProgram. (online). Available at: http://www.aoa.gov/AoARoot/AoA_Programs/HCLTC/Native_Americans/index.aspx. AccessedOctober 1, 2013

56. Korn L, Logsdon RG, Polissar NL, et al: A randomized trial of aCAM therapy for stress reduction in American Indian and AlaskanNative family caregivers. Gerontology 2009; 49:368e377

57. Fuentes D, Aranda MP: Depression interventions among racialand ethnic minority older adults: a systematic review across 20years. Am J Geriatr Psychiatry 2012; 20:915e931

58. Dunkin JJ, Anderson-Hanley C: Dementia caregiver burden: areview of the literature and guidelines for assessment and inter-vention. Neurology 1998; 51(suppl 1):S53eS60

Am J Geriatr Psychiatry -:-, - 2014