JKACHN Journal of Korean Academy of Community Health Nursing J Korean Acad Community Health Nurs 1225-9594 2288-4203 Korean Academy of Community Health Nursing 10.12799/jkachn.2019.30.4.449 jkachn-30-449 Original Article Trajectories of Self-rated Health among One-person Households: A Latent Class Growth Analysis KimEunjoo1 KimHyang1 YoonJu Young2 Doctoral Student, College of Nursing, Seoul National University, Seoul, Korea Associate Professor, College of Nursing · Research Institute of Nursing Science, Seoul National University, Seoul, Korea Corresponding author: Kim, Hyang College of Nursing, Seoul National University, 103 Daehak-ro, Jongno-gu, Seoul 03080, Korea. Tel: +82-2-740-8456, Fax: +82-2-766-1852, E-mail start2010@snu.ac.kr 122019 12122019 30 4 449 459 10062019 07102019 07112019 Copyright © 2019 Korean Academy of Community Health Nursing 2019 This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/3.0), which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited. Abstract Purpose

The aim of this study is to explore different types of self-rated health trajectories among one-person households in Korea.

 Methods

We used five time-point data derived from Korea Health Panel (2011~2015). A latent growth curve modeling was used to assess the overall feature of self-rated health trajectory in one-person households, and a latent class growth modeling was used to determine the number and shape of trajectories. We then applied multinomial logistic regression on each class to explore the predicting variables.

 Results

We found that the overall slope of self-rated health in one-person households decreases. In addition, latent class analysis demonstrated three classes: 1) High-Decreasing class (i.e., high intercept, significantly decreasing slope), 2) Moderate-Decreasing class (i.e., average intercept, significantly decreasing slope), and 3) Low-Stable class (i.e., low intercept, flat and nonsignificant slope). The multinomial logistic regression analysis showed that the predictors of each class were different. Especially, one-person households with poor health condition early were at greater risk of being Low-Stable class compared with High-Decreasing class group.

 Conclusion

The findings of this study demonstrate that more attentions to one-person households are needed to promote their health status. Policymakers may develop different health and welfare programs depending on different characteristics of one-person household trajectory groups in Korea.

 One-person household Subjective health Latent class analysis References ByunMSingle person household and urban policy in SeoulKorean Journal of Culture and Social Issues2015213551573 Statistics Korea2015 pupulation and housing census result [Internet]DaejeonStatistics Korea2016[cited 2019 Jan 24] Available from:http://kostat.go.kr/portal/korea/kor_nw/3/index.board?bmode=read&aSeq=356062&pageNo=&rowNum=10&amSeq=&sTarget=&sTxt= Organisation for Economic Co-operation and DevelopmentOECD Family Database 2013 [Internet]ParisOrganisation for Economic Co-operation and Development2013[cited 2019 Jan 24] Available from: www.oecd.org/social/family/database YeungWJCheungAKLiving alone: One-person households in AsiaDemographic Research2015324010991112https://doi.org/10.4054/DemRes.2015.32.40 KangENLeeMHSingle-person households in South Korea and their policy implicationsHealth and Welfare Policy Forum20162344756 KangETKangJKMaKRSubjective well-being of one-person households: Focus on non-married and married one-person householdsJournal of Social Science2016271323https://doi.org/10.16881/jss.2016.01.27.1.3 ParkBYKwonHJHaMNBurmEAA comparative study on mental health between elderly living alone and elderly couples: Focus on gender and demographic characteristicsJournal of Korean Public Health Nursing2016302195205https://doi.org/10.5932/JKPHN.2016.30.2.195 BailisDSSegallAChipperfieldJGTwo views of self-rated general health statusSocial Science & Medicine2003562203217https://doi.org/10.1016/S0277-9536(02)00020-5 BenyaminiYIdlerELLeventhalHLeventhalEAPositive affect and function as influences on self-assessments of health: Expanding our view beyond illness and disabilityThe Journals of Gerontology: Series B2000552107116https://doi.org/10.1093/geronb/55.2.P107 KaplanGBaron-EpelOWhat lies behind the subjective evaluation of health status?Social Science & Medicine200356816691676https://doi.org/10.1016/S0277-9536(02)00179-X SimonJGDe BoerJBJoungIMABosmaHMackenbachJPHow is your health in general? A qualitative study on self-assessed healthEuropean Journal of Public Health2005152200208https://doi.org/10.1093/eurpub/cki102 HaJKLeeSThe effect of health-related habitual consumption and lifetime on subjective health of one person households: Focusing on comparison between non-one personFamily and Environment Research2017552141152https://doi.org/10.6115/fer.2017.011 KimEGParkSKComparison of health behaviors, disease prevalence between one-person women and multiple households women in KoreaJournal of Korean Public Health Nursing2016303483494https://doi.org/10.5932/JKPHN.2016.30.3.483 KwonTTrajectories of self-rated health and its preditors: Focus on life spanThe Korean Journal of Humanities and the Social Sciences2011354203231 McAvayGRodinJDeterminants of change in perceived health in a longitudinal study of older adultsJournal of Gerontology1992476373384https://doi.org/10.1093/geronj/47.6.P373 HanBDepressive symptoms and self-rated health in community-dwelling older adults: A longitudinal studyJournal of the American Geriatrics Society200250915491556https://doi.org/10.1046/j.1532-5415.2002.50411.x KimJSChoBAssociation between self-perceived health status and health related behavior in routine health examineesKorean Journal of Family Medicine2010319688696https://doi.org/10.4082/kjfm.2010.31.9.688 Abu-OmarKRuttenARobineJMSelf-rated health and physical activity in the European UnionSocial and Preventive Medicine2004494235242https://doi.org/10.1007/s00038-004-3107-x HaYParkHAssociation between obesity and self-rated health in Korean males and femalesJournal of Korean Biological Nursing Science2012143203211https://doi.org/10.7586/jkbns.2012.14.3.203 Korean Ministy of Health & WelfareKorean Institute for Health and Social Affairs. Social security factbook 2017SejongKorean Ministy of Health & Welfare2017Dec. Report No.: 11-1352000-001385-10. KimYSYangYJParkHKKimJWDevelopment of physical activity guidelines and self-prescription guides for Korean. Research ReportSeoulKorean Health Promotion Foundation2012Nov. Report No.: Health Promotion Research Project 12-37. MäntyselkäPTTurunenJHAhonenRSKumpusaloEAChronic pain and poor self-rated healthThe Journal of the American Medical Association20032901824352442https://doi.org/10.1001/jama.290.18.2435 World Health OrganizationThe Asia-Pacific perspective: Redefining obesity and its treatment [Internet]SydneyHealth Communications Australia2000[cited 2019 Jan 23]. Available from:https://apps.who.int/iris/bitstream/handle/10665/206936/0957708211_eng.pdf?sequence=1 DuncanTEDuncanSCStryckerLAAn introduction to latent variable growth curve modeling: Concepts, issues, and applications2nd edNew YorkRoutledge2006261 JungTWickramaKASAn introduction to latent class growth analysis and growth mixture modelingSocial and Personality Psychology Compass200821302317https://doi.org/10.1111/j.1751-9004.2007.00054.x KlineRBPrinciples and practice of structural equation modeling4th edNew YorkGuilford publications2016445 SirolaJTuppurainenMRikkonenTHonkanenRKoivu-maa-HonkanenHKrogerHCorrelates and predictors of self-rated health and ambulatory status among elderly women: Cross-sectional and 10 years population-based cohort studyMaturitas2010653244252https://doi.org/10.1016/j.maturitas.2009.11.014 RobertsGAge effects and health appraisal: A meta-analysisThe Journals of GerontologySeries B199954B1S24S30https://doi.org/10.1093/geronb/54B.1.S24 Statistics Korea2010 pupulation and housing census result [Internet]DaejeonKorea Statistics2011[cited 2019 Jan 24]. Available from:http://kostat.go.kr/portal/korea/kor_nw/1/10/1/index.board?bmode=read&aSeq=249070 Figure and Tables

Latent classes of self-rated health trajectories.

General Characteristics of One-person Households on 2011 KHP data (N=698)

Domains Variables Categories One-person households
 Self-rated health
 t or F or r‡ p
n† (%)‡ or M±SD‡ M±SD‡
Demographic· socio-economic characteristics Gender Male 157 (28.5) 3.41±0.99 -4.07 ＜.001
Female 541 (71.5) 3.04±0.94
Age (year) 62.62±15.78 -.21 <.001
≤29 18 (4.8) 4.12±0.77 6.95 < .001
30~39 32 (8.2) 3.42±0.83
40~49 31 (6.4) 3.20±0.83
50~59 95 (16.2) 3.08±1.02
60~69 170 (21.2) 3.21±0.96
70~79 285 (35.2) 2.99±0.96
80~89 67 (8.1) 2.97±0.89
Education ≥ Elementary school 325 (51.3) 2.96±0.95 16.29 < .001
Middle-High school 169 (32.9) 3.29±0.97
≤ College 56 (15.7) 3.66±0.93
Missing 148
Marital status Single§ 84 (19.4) 3.52±0.91 4.59 < .001
Experience of being married‖ 614 (80.6) 3.06±0.96
Economic activity Active 287 (45.9) 3.37±0.90 -5.12 < .001
Inactive 411 (54.1) 2.96±0.98
Household income quantiles (weighted) Lowest 369 (47.6) 2.97±0.99 7.98 < .001
Second 167 (22.2) 3.10±0.92
Third 73 (12.2) 3.40±0.84
Fourth 51 (10.3) 3.47±0.94
Highest Fifth 38 (7.7) 3.60±0.90
Individual medical expenditure (KRW/year) 722,077.49± -.20 < .001
1,269,974.27
Missing 1 3.24±0.94 5.00 < .001
Medical security National health insurance 581 (84.2) 2.69±0.97
Medicaid 117 (15.8)
Health behavior Current cigarette smoking Yes 111 (19.4) 3.27±1.03 -1.50 .135
No 587 (80.6) 3.12±0.95
Excessive alcohol consumption Yes 88 (17.6) 3.55±0.91 -4.62 <.001
No 610 (82.4) 3.06±0.96
Healthy physical activity possible Yes 287 (40.3) 3.31±0.97 -3.28 .001
No 411 (59.7) 3.04±0.95
Health state Disabled Yes 107 (14.7) 2.72±1.07 4.00 < .001
No 591 (85.3) 3.22±0.93
Chronic disease Yes 617 (83.2) 3.05±0.96 5.80 < .001
No 81 (16.8) 3.66±0.82
Obesity Yes 184 (25.5) 3.11±0.93 0.62 .533
No 513 (74.5) 3.16±0.98
Missing 1
Depression Yes 89 (11.8) 2.65±0.99 4.60 < .001
No 609 (88.2) 3.22±0.94
Pain/discomfort Yes 379 (49.9) 2.77±0.92 10.17 < .001
No 319 (50.1) 3.53±0.86
Self-rated health 3.15±0.97
Very good 38 (6.4)
Good 215 (31.9)
Fair 256 (36.0)
Poor 163 (21.7)
Very poor 26 (4.0)
Missing 1

Unweighted;

Sample cross-sectional weights were applied for t-test, ANOVA test, Person r and percentages;

Never married;

Married, widow, divorced, separated.

Trajectory of Self-rated Health among One-person Households (N=698)

Latent variables Mean SE Model fit
x2 (p) CFI NFI TLI RMSEA
Initial intercept 3.09* 0.03 24.78 (.037) .99 .97 .99 .03
Slope -0.05* 0.01

CFI=comparative fit index; NFI=normed fit index; TLI=Tucker-Lewis index; RMSEA=root mean square error of approximation; *p＜.001.

Determining the Number of Latent Classes according to Self-rated Health Development Trajectory (N=698)

Categories 2-class model 3-class model 4-class model 5-class model
Model fit AIC 8,104.20 7,988.71 7,991.44 7,990.61
BIC 8,149.68 8,047.84 8,064.21 8,077.02
Adjusted BIC 8,117.93 8,006.56 8,013.41 8,016.69
n (%) according to Class 1 375 (53.7) 165 (23.6) 1 (0.1) 0 (0.0)
latent class Class 2 323 (46.3) 310 (44.4) 328 (47.0) 231 (33.1)
Class 3 223 (31.9) 204 (29.2) 163 (23.4)
Class 4 165 (23.6) 12 (1.7)
Class 5 292 (41.8)

AIC=Akaike information criterion; BIC=Bayesian information criterion.

Multinomi al Logistics Regression (N N=698)

Domains Variables Class 1: Low-Stable class (n=165)
 Class 2: Moderate-Decreasing class (n=310)
OR 95% CI Wald x2 p OR 95% CI Wald x2 p
Demographic· Age (year) 0.99 0.95~1.03 0.40 .530 1.01 0.99~1.04 0.79 .375
socio-economic characteristics Household income quantiles 1.03 0.74~1.44 0.04 .848 1.06 0.86~1.30 0.30 .583
Gender 2.79 1.05~7.46 4.19 .041 1.44 0.82~2.55 1.58 .208
  (ref. male)
Marital status 0.96 0.20~4.71 0.00 .958 2.93 1.17~7.35 5.22 .022
  (ref. experience of being married)
Education
  ≤Elementary school 0.11 0.01~1.24 3.17 .075 0.91 0.39~2.17 0.04 .838
  Middle~High school 0.26 0.02~2.73 1.28 .259 0.89 0.41~1.94 0.09 .769
  (ref.≥college)
Economic activity 0.43 0.21~0.88 5.37 .021 0.94 0.57~1.54 0.06 .809
  (ref. inactive)
Medical security 1.81 0.79~4.13 1.97 .160 1.19 0.58~2.42 0.22 .639
  (ref. national health insurance)
Health behavior Current cigarette smoking 6.44 2.30~18.04 12.54 ＜.001 1.44 0.75~2.77 1.22 .270
  (ref. no)
Excessive alcohol consumption 0.21 0.05~0.82 5.04 .025 0.58 0.31~1.12 2.65 .103
  (ref. no)
Healthy physical activity possible 0.32 0.17~0.59 13.32 ＜.001 0.57 0.37~0.86 6.99 .008
  (ref. no)
Health state Disabled 3.43 1.46~8.04 8.02 .005 1.45 0.70~2.99 1.01 .316
  (ref. no)
Chronic disease 14.93 1.68~133.02 5.87 .015 3.08 1.53~6.19 9.95 .002
  (ref. no)
Obesity 1.43 0.75~2.71 1.18 .278 0.97 0.61~1.55 0.01 .912
  (ref. no)
Depression 6.51 2.67~15.87 16.96 ＜.001 1.96 0.92~4.19 3.01 .083
  (ref. no)
Pain/discomfort 10.03 4.94~20.33 40.84 ＜.001 2.25 1.45~3.49 12.92 ＜.001
  (ref. no)

Reference group=Class 3: High-Decreasing class (n=223); OR=odds ratio; CI=confidence interval; ref.=reference.