• caregiving;
  • prevention;
  • mammography;
  • influenza;
  • cholesterol


  1. Top of page
  2. Abstract
  6. Acknowledgments

We examined the association between caregiving for a spouse and preventive clinical services (self-reported influenza vaccination, cholesterol screening, mammography, Pap smear, and prostate cancer screening over 2 years and monthly self-breast exam) for the caregiver in a cross-sectional analysis of the Health and Retirement Study, a nationally representative sample of U.S. adults aged ≥50 years (N= 11,394). Spouses engaged in 0, 1–14, or ≥14 hours per week of caregiving. Each service was examined in logistic regression models adjusting for caregiver characteristics. After adjustment for covariates, there were no significant associations between spousal caregiving and likelihood of caregiver receipt of preventive services.

Over 6 million adults in the United States provide long-term, unpaid care to disabled elderly persons in their families, particularly elderly spouses.1 These caregivers are at increased risk of mortality.2 Possible explanations include higher rates of negative health behaviors and health outcomes among caregivers, such as not having enough time to exercise, forgetting to take prescription medications, fatigue, depression, and worse blood pressure and lipid control.3–10

It is unclear whether the association between caregiving and mortality is also mediated through poorer quality health care to the caregivers, including decreased delivery of preventive clinical services. Caregiving consumes time and financial resources that may interfere with proper preventive care. In the Caregiver Health Effects Study, there was no association between caregiving and caregiver receipt of immunizations3; another study of caregiving found no association between caregiving and mammography, colon cancer screening, or Pap smear performance.11 In contrast, another study noted that caregivers were actually more likely than noncaregivers to get flu shots and receive pneumonia vaccines.12

These studies may have conflicting results due to differing study populations and the services examined. Studies focusing on health services delivered in the medical office involved select populations.11,12 For specific services such as influenza vaccination, caregiving could possibly provide increased opportunities through doctor visits provided for the caregiving recipient. Alternatively, for services that require an appointment such as Pap smears, caregiving could reduce the time available for such visits.

Given the conflicting results of prior studies of caregiving's effect on the use of preventive clinical services, we examined the association between an extended list of preventive clinical services and spousal caregiving in the Health and Retirement Study (HRS). To our knowledge, the HRS has enrolled the largest number of participants in studies of spousal caregiving and health services to date and also has detailed caregiving data as well as clinical service information.


  1. Top of page
  2. Abstract
  6. Acknowledgments

Data Sources and Collection Procedures

The HRS is a biennial longitudinal survey of a nationally representative cohort of the U.S. population aged 50 years or older13; for the purposes of this analysis, we analyzed year 2000 interviews. Both spouses within married households received a full interview. Our study sample included 11,394 married individuals aged 50 years or older with two respondents in the household who lived in the community.

The main outcome variables were caregiver self-reported receipt of preventive clinical services including influenza vaccinations, Pap smears, mammograms, cholesterol screening, and prostate screening over the past 2 years and monthly self-breast exam. The primary independent variable was the amount of informal caregiving provided. Respondents were classified as having received informal home care if they received in-home assistance with any activity of daily living (ADL), such as eating, transferring, toileting, dressing, bathing, walking across a room, or instrumental activity of daily living (IADL), such as preparing meals, grocery shopping, making telephone calls, taking medications, or managing money in the prior month from a relative (paid or not) or unpaid nonrelative with no organizational affiliation.14 Respondents who did not normally perform an IADL, such as shopping, were not classified as having an IADL impairment nor as having received help for that IADL. The number of weekly hours of informal home care was calculated using the average number of days per week (in the prior month), and the average number of hours per day that respondents reported receiving help from their spouse.15 We created tertiles of the number of hours of caregiving received, and from this we were able to create three categories of caregiving for the spouses. Caregivers were grouped into three categories: not providing care (noncaregivers), providing care for 1 to 14 hours a week, and providing care for more than 14 hours per week.

Statistical Analysis

To examine the association between the dependent variable of caregiver receipt of clinical service and the independent variable of caregiving, we examined each clinical service individually using logistic regression models. All analyses were weighted for differential probability of selection and adjusted for the complex sampling design of the HRS.13 We tested for significant interactions between all covariates, and regression diagnostics were performed to check for model assumptions. All covariates reflect characteristics of the caregiver, not the recipient of care, except for the presence of paid (formal) home care (Table 1).

Table 1. Participant Characteristics by Weekly Informal Care Hours Provided to Spouse, Percentages, or Weighted Means
Caregiver CharacteristicsHours of Caregiving per Week
No Care (N= 10,328)1 to 14 Hours (N= 622)>14 Hours (N= 444)P Value
  • *

    Service provided in the past 2 years except for monthly self-breast exam.

  • Standard errors are set in parentheses.

Age, y 65 (0) 68 (1) 71 (0)<.001
Female 47 55 49 .002
Race   <.001
 White 91 87 85 
 African American  6  8 11 
 Other  3  5  5 
Education, y 13 (0) 11 (0) 11 (0)<.001
Net worth, thousands of dollars512 (21)365 (61)253 (36)<.001
Current employment 42 32 18<.001
Resident children, number    .2
 None 78 80 82 
 1 child 16 15 13 
 >1 child  6  5  5 
Caregiver receipt of spousal care  9  9  8 .5
Caregiver health status   <.001
 Excellent 15 12  8 
 Very good 34 26 24 
 Good 30 29 32 
 Fair 15 22 25 
 Poor  6 12 12 
Presence of formal care for the  caregiving recipient  1  2  1 .04
Preventive clinical service*    
 Influenza vaccination 61 69 68 .002
 Cholesterol screening (N= 5,690) 79 78 76 .3
 Mammogram (N= 5,690) 80 75 73 .04
 Pap smear (N= 5,686) 70 62 58 .001
 Prostate (N= 5,674) 76 76 66 .02
 Self-breast exam (N= 5,688) 64 60 64 .4

We performed sensitivity analyses where we excluded data obtained from proxy respondents (N= 1,290) but found little change in the results (results not shown) and therefore present the results including these data. Due to the possibility that we were overadjusting by including both caregiver comorbidities and caregiver health status in the model, we constructed models with and without caregiver chronic medical conditions based on the Charlson comorbidity index16 and caregiver health status and found little difference (results not shown). This finding, combined with our reliance on self-report for caregiver comorbidities, led us to illustrate only models with caregiver health status. Due to concerns that we were overadjusting by including presence of formal care and caregiver receipt of spousal care, we created models that did not adjust for these variables, but this had little effect on the results (results not shown). Finally, in sensitivity analyses, we also adjusted for caregiver ADLs and IADLs, additional caregiving for dependents, hospitalizations, length of hospitalizations, and outpatient physician visits, but addition of these variables did not result in different associations between caregiving and performance of a preventive clinical service, and were not included in the final models. No interactions were significant, including those between net worth and health services or gender and health services. All analyses were performed using STATA statistical software, release 7.0 (STATA Corporation, College Station, Tex; 2001).


  1. Top of page
  2. Abstract
  6. Acknowledgments

Of the 11,394 participants, 10,328 (91%) did not provide care to their spouses, 622 (5%) provided care 1 to 14 hours per week, and 444 (4%) provided more than 14 hours of care per week. Characteristics of participants by caregiving category are shown in Table 1. Older persons (>75 years of age), women, African Americans, and people who had lower education, net worth, and employment rates, more comorbidities, and lower self-reported caregiver health status were more likely engaged in spousal caregiving. Rates of preventive service use across caregiving categories were relatively high (Table 1).

Table 2 shows the unadjusted and adjusted odds of receiving a health care service by caregiving category. In unadjusted analyses, provision of greater than 14 hours of caregiving per week was associated with significantly higher odds of the caregiver receiving influenza vaccination and slightly lower odds of the caregiver receiving cholesterol screening, mammography, Pap smear, and prostate examination, although only the association between Pap smears and prostate exams was significant. Greater intensity of caregiving was not associated with performance of self-breast exam. After adjustment for other covariates listed in Table 1, there were no significant associations between caregiving and caregiver receipt of health service measures (Table 2).

Table 2. Unadjusted and Adjusted Odds Ratios (95% Confidence Interval) of the Association Between Caregiving Hours and Caregiver Receipt of Preventive Clinical Services
Hours of Caregiving Provided per WeekUnadjusted Odds Ratios (95% CI)Odds Ratios (95% CI), Adjusted for Variables Listed in Table 1
1 to 14 Hours>14 Hours1 to 14 Hours>14 Hours
  1. Referent group is spouses who provided no caregiving.

  2. CI, confidence interval.

Influenza vaccination in the past 2 years1.44 (1.12 to 1.84)1.32 (1.03 to 1.69)1.30 (1.00 to 1.68)0.97 (0.77 to 1.24)
Cholesterol screening in the past 2 years0.95 (0.74 to 1.22)0.84 (0.67 to 1.04)1.00 (0.78 to 1.26)0.87 (0.68 to 1.10)
Mammography in the past 2 years0.75 (0.55 to 1.03)0.68 (0.47 to 1.00)0.99 (0.71 to 1.38)1.04 (0.71 to 1.52)
Pap smear in the past 2 years0.68 (0.52 to 0.90)0.60 (0.42 to 0.84)0.91 (0.70 to 1.19)0.97 (0.69 to 1.37)
Prostate cancer screening in the past 2 years0.98 (0.69 to 1.41)0.59 (0.40 to 0.86)1.16 (0.81 to 1.66)0.71 (0.47 to 1.07)
Self-breast exam every month0.84 (0.62 to 1.15)1.00 (0.78 to 1.28)0.91 (0.67 to 1.24)1.10 (0.84 to 1.44)


  1. Top of page
  2. Abstract
  6. Acknowledgments

Despite literature suggesting that caregivers frequently sacrifice their own health and well-being because of the demands of providing care, we found no evidence that caregiving was associated with higher or lower receipt of preventive clinical services for the caregiver. This suggests that the association between caregiving and increased mortality is mediated through other pathways warranting further investigation.

Our results are consistent with previous reports that found little association between caregiving and preventive health measures for the caregiver, either those care practices performed primarily outside the medical office or in the office.3,11,12,17,18 It is possible that this older group of persons has firmly established health routines that are minimally impacted by relatively new demands such as caregiving. It is also possible that any caregiving demands are balanced by increased contact with the health care system through visits for the impaired family member. Contact with the health system can increase perception of the need for services such as influenza vaccination, cholesterol screening, mammography, Pap smear screening, and prostate examinations.12 Such contact can also increase the opportunities for services that are relatively easy to administer, such as influenza vaccination. Finally, those who choose to become caregivers to begin with may have greater awareness of proper health behaviors and necessary services.

Our study has several limitations. Our analysis was cross-sectional, and it is possible that a “healthy worker” effect exists, where persons who are healthier are more likely to become caregivers.18 Although the HRS survey is longitudinal, the previous and subsequent waves dropped questions pertaining either to informal care for spouses or clinical services, limiting our ability to examine caregiving and clinical service use over time. The HRS relies on self-reported information for ADLs and IADLs, which may differ from performance-based measures of functioning. Similarly, our data on health service measures were obtained by self-report, which may underestimate actual services provided, particularly for cholesterol screening.19 Finally, our conclusions do not apply to other outcomes such as quality of life, stress, and depression or other health behaviors such as exercise, smoking, and healthy eating.

We conclude that spousal caregiving is not associated with lower performance of preventive clinical services in the caregiver. Future investigations should focus on the association between caregiving and processes of health care delivery for caregiver chronic disease management and quality of life.


  1. Top of page
  2. Abstract
  6. Acknowledgments

Dr. Kim was supported by an American Diabetes Association Junior Faculty Award. Dr. Langa was supported by a Career Development Award from the National Institute on Aging (K08 AG19180), a New Investigator Research Grant from the Alzheimer's Association, and a Paul Beeson Physician Faculty Scholars in Aging Research Award. The National Institute on Aging provided funding for the Health and Retirement Study (U01 AG09740), data from which were used in this analysis.


  1. Top of page
  2. Abstract
  6. Acknowledgments
  • 1
    Office of the Assistant Secretary for Planning and Evaluation. Informal Caregiving: Compassion in Action. Washington, DC: Department of Health and Human Services; 1998.
  • 2
    Schulz R, Beach S. Caregiving as a risk factor for mortality: the Caregiver Health Effects Study. JAMA. 1999;282: 22159.
  • 3
    Burton L, Newsom J, Schulz R, Hirsch C, German P. Preventive health behaviors among spousal caregivers. Prev Med. 1997;26: 1629.
  • 4
    Teel C, Press A. Fatigue among elders in caregiving and noncaregiving roles. West J Nurs Res. 1999;21: 498514.
  • 5
    Stone R, Cafferata G, Sangl J. Caregivers of the frail elderly: a national profile. Gerontologist. 1987;27: 61626.
  • 6
    Schulz R, O'Brien A, Bookwala J, Fleissner K. Psychiatric and physical morbidity effects of dementia caregiving: prevalence, correlates, and causes. Gerontologist. 1995;35: 77191.
  • 7
    King A, Oka R, Young D. Ambulatory blood pressure and heart rate responses to the stress of work and caregiving in older women. J Gerontol. 1994;49: M23945.
  • 8
    Archbold P, Stewart B, Greenlick M, Harvath T. Mutuality and preparedness as predictors of caregiver role strain. Res Nurs Health. 1990;13: 37584.
  • 9
    Montgomery R. The family role in the context of long-term care. J Aging Health. 1999;11: 383416.
  • 10
    Mittelman M, Ferris S, Shulman E, et al. A comprehensive support program: effect on depression in spouse-caregivers of AD patients. Gerontologist. 1995;35: 792802.
  • 11
    Sarna L, Chang B. Colon cancer screening among older women caregivers. Cancer Nurs. 2000;23: 10916.
  • 12
    Scharlach A, Midanik L, Runle M, Soghikian K. Health practices of adults with elder care responsibilities. Prev Med. 1997;26: 15561.
  • 13
    Juster F, Suzman R. An overview of the Health and Retirement Study. J Hum Resour. 1995;30: S7S56.
  • 14
    Norgard T, Rodgers W. Patterns of in-home care among elderly black and white Americans. J Gerontol B Psychol Sci Soc Sci. 1997;52: 93101.
  • 15
    Wolf D, Freedman V, Soldo B. The division of family labor: care for elderly parents. J Gerontol B Psychol Sci Soc Sci. 1997;52: 1029.
  • 16
    Charlson M, Pompei P, Ales K, MacKenzie C. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chron Dis. 1987;40: 37383.
  • 17
    Altman B, Cooper P, Cunningham P. The case of disability in the family: impact on health care utilization and expenditures for nondisabled members. Milbank Q. 1999;77: 3974.
  • 18
    Baumgarten M, Battista R, Infante-Rivard C, et al. Use of physician services among family caregivers of elderly persons with dementia. J Clin Epidemiol. 1997;50: 126572.
  • 19
    Harwell T, Moore K, Madison M, et al. Comparing self-reported measures of diabetes care with similar measures from a chart audit in a well-defined population. Am J Med Qual. 2001;16: 38.