IZVIRNI ČLANEK/ORiGiNAL ARTiCLE
Social support in the case of illness: intergenerational solidarity
Socialna opora v primeru boLezni: medgeneracijska solidarnost
Majda Pahor,1 Barbara Domajnko,^ Valentina Hlebec^
1	Univerza v Ljubljani, Zdravstvena fakulteta
2	Univerza v Ljubljani, Zdravstvena fakulteta
3	Univerza v Ljubljani, Fakulteta za družbene vede
korespondenca/ Correspondence:
izr. prof. dr. Majda Pahor Zdravstvena fakulteta Zdravstvena pot 5 1000 ljubljana tel. 01 300 11 11 majda.pahor@zf.uni-Lj.si
ključne besede:
obvladovanje bolezni, družinska podpora, medgeneracijski odnosi
key words:
coping with illness, family support, intergenerational relations
Citirajte kot/Cite as:
Zdrav Vestn 2011; 80: 75-83
Prispelo: 31. mar. 2010, Sprejeto: 30. jun. 2010
abstract
Background: Social support received through different forms of help from members of one's social network is an important element of coping with illness. In the case of illness, family members are the main providers of support, both within the same generation, but also, and increasingly so, between generations. This informal social support is related to socio-economic conditions of individuals: it is more common in lower economic and educational groups. Members of the middle generation, who help both the young and the old, are the main support providers. Also, female gender is the most significant predictor of the care burden. Withdrawing role of the welfare state in the postmodern society means shifting more responsibilities for care from the formal to informal sector. The aim of our study was to look into the characteristics of intergenerational support in illness in Slovenia.
Methods: A cross-sectional study on personal support networks of the residents of Slovenia, sample size 5013, data collection by computerassisted telephone interviews, respondents above 18 years of age. Multiple Classification Analysis (MCA) was used for data analysis to find out how much the dependent variable deviated from the mean as a result of a given respondent characteristic while controlling for the effects of all others.
Results: The analysis showed the proportion of respondents' social network that would provide support in the case of illness and could be
defined as intergenerational network. Intergen-erational ties represent about 35 % of the whole support netork in illness. The most frequent receivers are the youngest group of respondents (18-29), followed by the 60+ age group. Women receive more help than men, especially those who are widows, living alone or living in multi-generational households. Intergenerational support is more frequent among the less educated respondents.
Discussion: Our results comply with the findings in the literature, and are indicating that the actual trends in the changing structure and composition of the family, combined with less support from institutional health- and social care, is increasing the care burden of the informal carers within families.
Conclussions: Health and social care policy and practice need awarness of the contextual factors of health care outcomes, taking into consideration social support networks' functions.
Izvleček
Izhodišča: Socialna opora, ki jo v različnih oblikah nudijo člani socialnega omrežja, je pomemben dejavnik spoprijemanja z boleznijo. V primeru bolezni so družinski člani glavni viri opore znotraj iste generacije, vedno bolj pa tudi med generacijami. Neformalna socialna opora je povezana s posameznikovimi socialno-ekonom-skimi pogoji: pogostejša je pri pripadnikih nižjih socialnih slojev in manj izobraženih. Glavni viri so pripadniki srednje generacije, ki poma-
gajo tako mlajšim kot starejšim. Ženski spol je najmočnejši napovedni dejavnik obremenjenosti z dajanjem opore. V postmoderni družbi se z zmanjševanjem vloge države blaginje vse več odgovornosti za skrb prenaša iz formalnega v neformalni sektor. Namen raziskave je proučiti značilnosti medgeneracijske opore v primeru bolezni v Sloveniji.
Metode: Izvedli smo presečno študijo osebnih socialnih omrežij prebivalcev Slovenije. Vzorec je obsegal 5013 anketirancev. Podatke smo zbrali s telefonskimi računalniško podprtimi intervjuji, anketiranci pa so bili starejši od 18 let. Za analizo podatkov smo uporabili multiplo klasifikacijsko analizo. Z njo smo ugotavljali stopnjo odstopanja odvisne spremenljivke od povprečja kot posledico danih značilnosti anketirancev ob nespremenjenih drugih značilnostih.
Rezultati: Analiza je pokazala delež socialnega omrežja anketirancev, ki bi nudil oporo v primeru bolezni in ga lahko opredelimo kot medgene-
racijsko omrežje. Medgeneracijske vezi predstavljajo približno 35 % celotnega opornega omrežja v primeru bolezni. Najpogostejši prejemniki so mladi (18-29 let), sledi pa jim skupina starejših od 60 let. Ženske prejmejo več pomoči kot moški, posebej velja to za vdove in ženske, ki živijo same ali v multigeneracijskih gospodinjstvih. Medgeneracijska pomoč je pogostejša pri manj izobraženih anketirancih.
Razpravljanje: Rezultati so v skladu z izsledki literature. Nakazujejo, da aktualno spreminjanje strukture in sestave družine v kombinaciji z zmanjšano stopnjo opore s strani institucionalnega zdravstvenega in socialnega varstva predstavlja povečano skrbstveno obremenitev neformalnih virov opore znotraj družin.
Zaključki: Pri oblikovanju politik in izvajanju zdravstvenega in socialnega varstva se moramo zavedati kontekstualnih dejavnikov pri izidih zdravstvenega varstva in pri tem upoštevati funkcije omrežij socialne opore.
Introduction
Illness means poor health resulting from disease of body or mind. The emphasis of its meaning is on the subjective aspects of health and problems with functioning in everyday life. Social support received through different forms of help from the members of one's social network is an important element of coping with illness. WHO (World Health Organization) defines social support provided by social networks, as a »solid fact« about social determinants of health. 1 Belonging to social networks where communication and mutual support take place, allows people to feel accepted and valued, which has a positive impact on health. Research showed^ that an average person has 5 to 6 members in their immediate support social network including family members, friends, co-workers and neigbours. In the case of illness, family members are the main providers of support.^ In this article, we are interested in the intergenerational solidarity between network members.
Background
The current society is marked by increasing longevity and decreasing birth rate.
How are these changes related to the inter-generational structure of families? The metaphor of "the beanpole family"^ describes the increasing number of generations within a family living at one time with fewer members. People thus have fewer relatives in the horizontal (siblings, cousins) but more in the vertical family line (children, parents, grandparents, grandgrandparents). The absolute number of family members may in fact be decreasing. The same authors also distinguish between two types of families according to fertility timing: age-condensed and age gapped. The first type is characterized by the blurring of generational boundaries, while the second is a consequence of delayed childbearing. Today some adults can have more parents than children, while some people already experience grandpa-renthood for more than half of their lives.
Due to demographic changes and also a high divorce rate, heterogeneity of family/ household structure is greater than ever. Childless adults, multigenerational households, reconstituted or blended families, single parenthood, skipped-generation households-grandparents raising grandchil-dren,5 gay and lesbian families all challenge the prevalence of the traditional nuclear family.
In great part, demographic changes are also a reflection of socioeconomic conditions, which have contributed to a better quality of life and consequently increased longevity. Nowadays we face promotion of privatization, liberalization of trade and monetary policies, deregulation, decrease in welfare programs and public services brought by globalization. These trends have redefined the character of the modern state in relation to the private sector. Some authors® also mention the belief that social protection guaranteed by the welfare state and its distributive policies hinders economic growth. Consequently, the state is supportive of the privatization of the welfare sector, which is a principle also guiding political reforms in the health care sector. Socioeconomic conditions of living are becoming increasingly unequal and unsteady, which is precisely the situation that generates ill health. High unemployment rates and threatening poverty are in close relation to decline of health. Health indicators in Slovenia based on the 2002 data,^ show an improvement in health, however, it is very probable that the economic recession will lead to negative trends similar to those in the early i99os.8 People occupying lower economic positions have less opportunities to afford healthy lifestyle, safe and suitable housing, products assisting independent living, and experience limited access to health care.® Parallel to that, community-based health care, greater participation by service users, families and other informal network ties are stressed. It has become obvious that the state does not perceive itself to be the sole (or even the primary) provider of health care. The responsibilities for providing access to health care has shifted to individuals and families, who sometimes find it hard to secure decent living conditions even for themselves.®
There is a direct link between social class and intergenerational social support pro-vision.9 Middle-class elderly provide more economic support to their adult children, while lower class elderly do not and rely more on their adult children. Middle-class adult children can afford to make use of paid services to a greater degree than lower class adult children. The informal support
sources pool is shrinking due to changed working patterns for women, adult children mobility and childless couples/single house-holds.10 Especially vulnerable is the so-called "sandwich generation", the generation caught between two caring roles - towards own children (grandchildren) and elderly parents (grandparents).
Intergenerational relations are complex not only due to variations in age but also to values and norms that govern expectations related to behaviour. On the individual level, age and gender are the two most important variables influencing intergenerational relations. Age refers to changes in roles and responsibilities and gender (female) is the primary criterion for incurring the care-giving role.11 Women today find themselves in a paradoxical situation. They tend to live longer, but their later life is characterized by chronic illness and disability. At the same time, they are the ones of whom it is expected to provide informal support to their family in case of illness. They are therefore at a greater risk to have unmet care needs and tend to be institutionalized at higher rates. Gender is the single most important predictor of hours spent providing care.® In relation to gender and the allocation of caring responsibilities in case of illness, there is supposed to be a common preference order.12 Prefered care givers are daughters, followed by daughters-in-law, and sons if the elderly is male; in intimate care men-to-men and women-to-women care is preferred, also women-to-men but never men to women or non-related women to men.
The need to rely on personal (close family) relations in case of illness is therefore partly also determined by what one gets from societal sources.^ Within welfare regimes, state and institutional support enhanced independence by providing at least some means directly to the people in need. New political tendencies shift the burden progressively from the formal towards informal support networks. If care outside health care institutions is not to be covered by general public health insurance, the financial as well as caring burden gets down directly to people and their close families.
Table i: MethodoLogicaL characteristics of survey and data (Social Support Networks of the Residents of SLovenia 2002). table adapted from kogovšek and Hlebec, 20051^
Survey	Social Support networks of the residents of Slovenia, 2002
Research institute	CMi - Centre for Methodology and informatics, Faculty of Social Sciences, University of Ljubljana SPiRS- Social Protection institute of the Republic of Slovenia
Data	CMi, SPiRS
Data coLLection	CATi center, Ljubljana
SampLe	Random sample of telephone users in Slovenia (fixed telephone coverage in Slovenian households in 2002 was 91 %)*
n	5013
age	18 +
Data coLLection mode	Computer assisted telephone interview
Data coLLection	February 2002 - April 2002
Support in the case of an iLLness	Name generator, Actual (usual) provision of social support by an informal network
" Source: Statistical Yearbook of the Republic of Slovenia 2003, Statistical Office of the Republic of Slovenia, Ljubljana, 2003.
The aim of the article is to answer the question: How do intergenerational social networks function as a source of support in the case of illness? Also, we will look into who gives and who receives support and which socio-demographic groups are most vulnerable. We also discuss the implications of such situation for the health care and for health and social policy.
Methods
In this section, data and methods are described. This cross-sectional study includes data on personal support networks of the residents of Slovenia.^^ The sample size was 5013 and it is representative for Slovenian residents after weighting. Data was collected using computer-assisted telephone interviews. Respondents were aged 18 and more and there was no upper age limit.
Data were weighted using post-stratification weights. Ten categories of gender and age were used together with educational structure to correct the realized sample for population structure. The data from 2002 census were used for post-stratification. The characteristics of realized sample after post-
-stratification are presented in the following table.
Multiple Classification Analysis (MCA) was chosen for analysis as it allows numerical dependent variable and a combination of nominal, ordinal and numerical independent variables.^^ It is similar to multiple regression analysis, with the advantage of nominal measurement not requiring to be dichotomized. There are several coefficients which would indicate how much the dependent variable deviates from the mean as a result of a given respondent characteristics while controlling for the effects of all other respondents' characteristics. Two measures of the overall effect of each predictor are obtained, and in addition the MCA Eta and MCA Beta. The MCA Eta coefficient measures the strength of the bivariate relationship between a dependent variable and a predictor. MCA Beta coefficients, on the other hand, measure the strength of the relationship, controlled for other independent variables in the model. The rank order of the Betas indicates the relative importance of the independent variables in their explanation of the dependent variable. Finally, the multiple R2, indicating the total proportion
Table 2: Demographic characteristics of a realized sample in % (n=50i3)
Age						
18-29	30-39	40-49	50-59	60-69	70 +	
25.76	17.67	17.39	15.76	12.39	11.03	
Gender						
Male	Female					
48.34	51.66					
Education						
incomplete elementary school	Elementary school	Vocational school	High school	College	University degree	Master degree and more
5.03	28.00	19.27	34.78	5.05	7.24	0,44
Place of living						
Rural	Suburban	Urban				
46.37	20.05	33.59				
Marital status						
Single	Living as married	Married	Divorced/ Separated	Widowed		
28.08	8.30	50.64	3.60	9.39		
Household composition						
Living alone	Single parent and children	Couple without children	Couple with children	Multigenerational household	Other	
11.63	8.76	15.07	46.43	6.51	11.59	
of variance explained by all independent variables together, is estimated.
The dependent variable that we are interested in is the proportion of respondents' informal network that would provide help in the case of illness and can be defined as intergenerational informal network. All possible intergenerational ties in close family were taken into account (parents and their adult children, regardless of their age; grandparents and grandchildren) and the proportion of such ties was calculated with regard to the total informal network (all people that respondents report as providers of social support, see^ for further discussion on social support and social networks).
Results
Several MCA models were estimated owing to the interaction of independent variables. In the first model basic demographic
characteristics are taken into account, age in 10-year categories, gender, place of living and education. Independent variables in the first model explain about 5 »/0 of variability in the proportion of intergenerational ties for provision of social support in the case of illness. The most important predictor variable is age, followed by gender and education. On average, intergenerational ties represent about 35 % of the whole support network in the case of illness. The most important provider of social support in the case of illness is, of course, the partner, who represents about 33 / of the whole support network in the case of illness (not included in this model as the focus is on intergenerational support).
Among age groups, intergenerational support is by far the most frequent for the youngest group of respondents (18-29), followed by the two oldest groups (60+). The middle age group has by far the smallest
		Intergenerational ties					
		Grand mean =	= 35,84		Predicted Mean	Deviation	
		N	Eta	Beta	Adjusted for Factors	Adjusted for Factors	
age	18- 29	1236			46,738	10,901	
	30-39	828			22,282	-13,555	
	40-49	810			29,791	-6,045	
	50-59	707			36,593	0,757	
	60-69	569			37,617	1,781	
	70 +	489	0,205	0,204	38,092	2,255	***
gender	MaLe	2193			31,461	-4,376	
	Female	2446	0,098	0,100	39,759	3,923	***
PLace of Living	RuraL	2145			35,154	-0,683	
	Suburban	939			35,767	-0,070	
	Urban	1555	0,019	0,018	36,820	0,984	
education	VocationaL schooLor Less	2417			37,132	1,295	
	High schooLor more	2221	0,039	0,032	34,426	-1,410	**
MuLtipLe R2				0,053			
percentage of intergenerational ties, indicating that they are the nett providers of intergenerational social support in the case of illness, however, we should not ignore that intergenerational ties still represent from 2236 percents of social support network in the case of illness. Intergenerational social support is more frequent in women (men get it more often from their partner) and less educated respondents. The place of living has no significant impact on the proportion of intergenerational ties in the social support network in the case of illness.
In order to allow also other independent variables to be included in the model, age groups were collated in the next two MCA analyses. In the second MCA, household composition was included together with age and gender. These three variables explain about 7 % of the variability in the proporti-
on of intergenerational ties in social support network in the case of illness. Household composition is as important predictor as age. Intergenerational ties are by far most frequently present in single-parent families. In the absence of partner, which is the most important provider of social support in the case of illness, obviously, intergenera-tional ties within immediate family step in. Among other households, people living in a multigenerational household also have more providers of social support among interge-nerational ties.
In the third model, marital status was examined as a predictor variable. Age, gender and marital status explain about 8 »%o of variability in the proportion of intergene-rational ties. Marital status is even more important predictor of the proportion of intergenerational ties than age (there is, of
		Intergenerational ties				
		Grand mean = 35,82		Predicted Mean	Deviation	
		N	Eta Beta	Adjusted for Factors	Adjusted for Factors	
age	18- 29	1231		44,342	8,521	
	30-59	2342		28,532	-7,288	
	60+	1058	0,179 0,178	42,043	6,223	***
gender	MaLe	2189		32,213	-3,607	
	Female	2442	0,097 0,082	39,056	3,235	***
HousehoLd composition	Living aLone	501		32,806	-3,014	
	SingLe parent famiLy	402		55,252	19,432	
	CoupLe without chiLdren	702		24,821	-11,000	
	CoupLe with chiLdren	2180		36,382	0,561	
	MuLtigenerationaL househoLd	305		38,269	2,448	
	Other	541	0,179 0,174	34,808	-1,013	***
Multiple R2			0,069			
course, an interaction between age and marital status; younger respondents tend to be single, older respondents are more likely widowed). Intergenerational support is most important for widowed or divorced respondents. Absence of a partner is significant for both categories as well as for the single respondents who have larger proportion of intergenerational ties.
If we sum up the main results, we see that recipients of intergenerational support are mostly the youngest and the oldest, while main support provider is the middle generation. This is compatible with the interpretation of the "sandwich generation" as the most important provider of social support in both directions (to the younger generations and to the elderly).
In terms of gender, the results show that women receive more intergenerational support then men. This might be attributable to the longevity of women (which is very often associated with chronic illness and disability), less support received by partner (men get most support from their partner,
which is an intragenerational tie); also, women experience higher rates of widowhood (marital status confirms that widowhood generates intergenerational social support), living alone or in multigenerational households (such household composition results in more frequent intergenerational support contacts).
Intergenerational support is more frequent in less educated people who rely more on family support, while the more educated have better economic opportunities and can afford other arrangements. Also, intergenerational support depends on household composition: it is more frequently received by single, then those living in mul-tigenerational households, and by couples with children. In terms of marital status, intergenerational support is most frequent in the widowed or divorced, followed by singles.
		Intergenerational ties				
		Grand mean = 35,80		Predicted Mean	Deviation	
		N	Eta Beta	Adjusted for Factors	Adjusted for Factors	
age	18-29	1236		42,135	6,332	
	30-59	2342		32,503	-3,300	
	60 +	1057	0,180 0,097	35,711	-0,092	***
gender	MaLe	2192		33,004	-2,799	
	Female	2443	0,097 0,064	38,315	2,512	***
Marital	SingLe	1319		42,573	6,770	
status	Married or Living as married	2737		28,868	-6,936	
	Divorced, widowed	579	0,264 0,214	53,175	17,372	***
MuLtipLe R2			0,079			
Discussion
Literature on intergenerational relations mentions several myths that can also be addressed by the results of this study. One of them is the belief that elderly parents have been abandoned by their children.^'® As this may seem to be true due to modern living arrangements - smaller households and high geographic mobility - the frequency of intergenerational ties confirms that at least one child maintains regular contacts and gets actively involved in the provision of care in case of illness. The statistics do not tell us much about the quality of these ties, and it would be illusory and also theoretically groundless to exclude conflict and ambivalence from intergenerational relations. Still, the quantitative data show clear evidence of frequent contacts, the nature of which is yet to be explored, at least in the Slovenian context.
Another myth relates to the elderly as primary and sole recipients of social support in the case of illness. Similar to other studi-es,i® our results indicate that intergeneratio-nal support in the case of illness is far more frequent in the younger generation (18-29) than in 60+ group. This is also consistent with the findings on the moral component
of intergenerational ties^^ according to which filial obligation is a very strong norm though not universal nor unconditional. The other type of data to be used in discussing the myth is data about the reciprocity of support. Many authors (e. g. 11, 12) have found out that elderly generations may receive more material support but can at the same time provide emotional support.
Taking into account wider social context, particularly the latest reforms of the health care system, it becomes clear that the burden of providing or organizing provision of care is progressively shifted to individuals and their informal support networks. This raises the question that needs to be addressed urgently: how can the providers of social support be suitably supported themselves in the case of their illness?
Conclusions
Further to our starting question about the role of intergenerational social support in the case of illness, it is possible to conclude that this is as important source of help as the intragenerational one. Probably, the importance of intergenerational solidarity in the case of illness will increase with decreasing role of the public health sector in
providing care, which will increase the care burden of the main providers of support (middle generation). It is also possible that support provided to those who need it most in the case of illness (elderly, disabled, living alone) will not be sufficient. ^ese findings should be an important issue for health and social care policy, as well as for care planning after patients' discharge from hospitals.
reference:
1.	WHO-World Health Organization. Social determinants of health: the solid facts. Dosegljivo na: http://www.who.dk/healthy cities/Documentation/.
2.	Hlebec V, Kogovšek T. Konceptualizacija socialne opore. Družboslovne razprave 2003; 19: 103-25.
3.	Pahor M, Hlebec V. Social networks and health: transition in Slovenia. Zdrav Varst 2006; 45: 17585.
4.	Bengtson VL, Rosenthal CJ, Burton LM. Families and aging: Diversity and heterogeneity. In: Bin-stock RH, George L, eds. Handbook of aging and the social sciences. 3rd ed. San Diego, CA: Academic Press; 1990. p. 263-87.
5.	Hooyman NR, Kiyak HA. Social Gerontology: a multidisciplinary perspective. 8'^ ed. Massachusetts: Allyn & Bacon; 2008.
6.	Chappel NL, Penning MJ. Family caregivers: increasing demands in the context of 21th century globalization? In: Johnson ML, ed. tte Cambridge Handbook of Age and Ageing. Cambridge: CUP;
2005. p. 455-63.
7.	Zatonski W, Manczuk M, Sulkowska U in skupina projekta HEM. Odpravljanje razlik v zdravju v Evropski uniji. Opis značilnosti države: Slovenija. Projekt HEM: Odpravljanje razlik - Zniževanje prezgodnje umrljivosti. Izhodišče za spremljanje sprememb v zdravju po širitvi 2008.
8.	Hanžek M, ed. Human development report. Slovenia 1998. Ljubljana: Institute of Macroeconomic Analyses and Development & United Nations Development Programme; 1998
9.	Matcha DA. tte Sociology of Aging: a social problems perspective. Massachusetts: Allyn & Bacon;
1996.
10.	Lowenstein A. Global Ageing and Challenges to Families. In: Johnson ML, ed. tte Cambridge Handbook of Age and Ageing. Cambridge: CUP; 2005. p. 403-13.
11.	ttomese F, van Tilburg T, van Groenou MB, Knip-scheer K. Network dynamics in later life. In: Johnson ML, ed. tte Cambridge Handbook of Age and Ageing. Cambridge: CUP; 2005. p. 463-69.
12.	Reed J, Stanley D, Clarke C. Health, well-being and older people. Bristol: Policy Press; 2004.
13.	Ferligoj A, Kogovšek T, Dremelj P, Hlebec V, Nova M. Omrežja socialnih opor prebivalstva Slovenije. Ljubljana: Fakulteta za družbene vede in Inštitut Republike Slovenije za socialno varstvo; 2002.
14.	Kogovšek T, Hlebec V. Effects of limitation of number of alters and time frame in the burt name generator. Metodološki zvezki 2005; 2: 59-71.
15.	Andrews FM, Morgan JN, Sonquist JA, Klem L. Multiple Classifications Analysis. Institute for Social Research: Ann Arbor; 1973.
16.	Finch J, Mason J. Filial obligations and kin support for elderly people. In: Gubrium JF, Holstein JA, eds. Aging and Everyday Life. Malden: Blackwell Publishing; 2005. p. 193-211.