What do you get when you fall in love? Some health status effects

18
Theodore D. Kemper and Roslyn WaUach Bologh What do you get when you fall in love? Some health status effects Abstract Students of the effects of life change events on health have asked for studies in which single, desirable, life events that are appropriate to the life stage of the population sampled are examined in relation to health status. We do this with a single event romantic involvement - that moves through a number of phases. We found in our college student sample that adverse health reactions are more likely to occur among those who have never been in love or who have been in love more than three times; among those involved in a relationship of two to six nionths duration as opposed to a shorter or longer period; among those who have just broken up a relationship; among those who attri- bute the break-up mainly to the other person; and among those who, having broken up, have not found a new romantic interest, or have plunged deeply into a new relationship. Some of the results are moderated by interaction effects with sex of respondent. What do you get when you kiss a guy? You get enough germs to catch pneumonia .. . From: 'I'll Never Fall in Love Again,' Burt Bacharach song Is it good to be romantically involved? Specifically, is it good for one's health? By cultural and professional fiat, love and the ability to love are among our highest values. From Freud onward, clinicians and psycho- therapists have endorsed the ability to love as a sign of emotional health (Maslow, 1953; Fromm, 1956), or as contributing to emotional health (Horney, 1946). Popular conceptions as conveyed in song, drama, and fiction also frequently elevate love and romantic involvement as the most desirable of fates and fortunes. Researchers investigating the health effects of cumulative life change have similarly scored romantic involvement as a 'plus', a positive life event (Holmes and Rahe, 1967). Yet there appears to be another side of the story, as virtually all who have had romantic involvement are aware. Pop songs, from blues to ballads, lament the vicissitudes of romance and the fate of lovers: in- ability to concentrate, loss of sleep or appetite, despondency when Sociology of Health and Illness Vol. 3 No. 1 1981 ©R.K.P. 1980 0141-9889/81/0301-0072 $1.50/1

Transcript of What do you get when you fall in love? Some health status effects

Theodore D. Kemper and Roslyn WaUach Bologh

What do you get when you fall in love?Some health status effects

Abstract Students of the effects of life change events on health have asked forstudies in which single, desirable, life events that are appropriate to thelife stage of the population sampled are examined in relation to healthstatus. We do this with a single event — romantic involvement - thatmoves through a number of phases. We found in our college studentsample that adverse health reactions are more likely to occur amongthose who have never been in love or who have been in love more thanthree times; among those involved in a relationship of two to sixnionths duration as opposed to a shorter or longer period; amongthose who have just broken up a relationship; among those who attri-bute the break-up mainly to the other person; and among those who,having broken up, have not found a new romantic interest, or haveplunged deeply into a new relationship. Some of the results aremoderated by interaction effects with sex of respondent.

What do you get when you kiss a guy?You get enough germs to catch pneumonia . . .From: 'I'll Never Fall in Love Again,' Burt Bacharach song

Is it good to be romantically involved? Specifically, is it good for one'shealth? By cultural and professional fiat, love and the ability to love areamong our highest values. From Freud onward, clinicians and psycho-therapists have endorsed the ability to love as a sign of emotional health(Maslow, 1953; Fromm, 1956), or as contributing to emotional health(Horney, 1946). Popular conceptions as conveyed in song, drama, andfiction also frequently elevate love and romantic involvement as themost desirable of fates and fortunes. Researchers investigating thehealth effects of cumulative life change have similarly scored romanticinvolvement as a 'plus', a positive life event (Holmes and Rahe, 1967).

Yet there appears to be another side of the story, as virtually all whohave had romantic involvement are aware. Pop songs, from blues toballads, lament the vicissitudes of romance and the fate of lovers: in-ability to concentrate, loss of sleep or appetite, despondency when

Sociology of Health and Illness Vol. 3 No. 1 1981©R.K.P. 1980 0141-9889/81/0301-0072 $1.50/1

What do you get when you fall in love? Some health status effects 73

apart from the other, hurt and pain due to rejection or the break-up ofa relationship etc. Failed love relationships lead to deeply felt distress-ful emotions such as depression, anger, shame, guOt, and anxiety(Kemper, 1978). Significant numbers of persons enter psychotherapyor seek therapeutic guidance when romantic involvements becomeimpaired. There is, thus, no single answer to the question of whether aromantic involvement is good for one's health.

The purpose of this paper is to present some data on the relationshipbetween several indices of romantic involvement and health status. Wesee our work as a contribution to the understanding of how life eventsaffect emotional and physical health (Holmes and Rahe, 1967; Dohren-wend and Dohrenwend, 1974; Eaton, 1978).

There are three issues in the life events literature that bear directlyon our interest in health and romantic involvement: (1) Is healthaffected only by undesirable events or by any kind of life change? (2)Is a single event sufficient to affect health or are aggregates of eventsrequired? (3) Do certain kinds of events affect health at any point inthe life course or are specific events particularly consequential whenthey occur at particular life stages? We discuss these questions now.

Undesirable versus any kind of life change

Most studies have concentrated on undesirable life events and giverelatively little attention to the health effects of desirable hfe changes.Rahe (1974) acknowledges that the original life events protocol.Holmes and Rahe's (1967) Social Readjustment Scale — the prototypefor most others — undersamples positive changes. Yet research into thequestion of undesirabihty versus change per se has produced equivocalresults. Paykel (1974) and Gersten, et al. (1974: 1977) have found thatonly undesirable changes are related to psychiatric disorders. ButRahe (1974) found a positive relationship between desirable life eventsand illness reports. However, these results are not definitive because sofew positive events were studied.

Dohrenwend (1973) found that combining desirability with un-desirability produced a total life-change measure that was more highlycorrelated with mental distress than undesirability alone, suggestingthat change not undesirability affects health negatively. Gersten (1974),however, found that the balance between undesirable and desirablechanges was the critical dimension for predicting health effects. On theother hand, Vinokur and Selzer (1975) and Mueller et al. (1977)reported that undesirable events are the most highly correlated withpsychological health status.

74 Kemper and Bologh

Our own position is that romantic involvement has both desirableand undesirable consequences, and that these are related in part to thephase of the relationship (Rubin, 1969; Centers, 1975). We wereparticularly interested in this aspect of involvement as a predictor ofhealth status.

Single versus aggregate events and changes

Most of the research on the health effects of life changes has used com-posite scores based on a check list of events that are assumed to bestressful. In general the greater the number of events checked, i.e. thegreater the magnitude of life change, the more likely the population atrisk experienced ill health (Holmes and Masuda, 1974: 68). However,there has not been much work linking specific events or life changes tomeasures of health or well being.

Dohrenwend (1974) argued that summary scores based on total lifechange, or total gain versus total loss, or total desirable versus un-desirable events obscure the contribution of specific events to psycho-pathology. Indeed, he found that specific loss events for which theindividual felt responsible, were related to psychiatric disorder.

Brown (1974) too supports the idea of focusing on single events. Hefound that it was not necessarily a high composite score, but a single,severe event, implying a long-term threat, that precedes mental dis-order. A major loss (e.g. of a job or a spouse) was the most importantsingle, long-term threatening event that could lead to emotional illness.Gersten (1977) also found that stressful conditions of long duration,rather than aggregates of short term events, effect behaviour disordersand health impairment among young people, from preadolescencethrough early adulthood.

Romantic involvement, as we conceive it here, is neither a single lifeevent, nor is it a composite of events that can be summed to somemeaningful total score. It comprises a series of stages of increasing (and/or decreasing) emotional and behavioural involvement that are stronglycorrelated with the duration of the relationship. Dohrenwend (1974)included 'a broken love relationship' as one of his items. We include thisas well as several other phases of romantic involvement, examining eachas single events that might affect health status.

General versus specific life stage effect

Kellam (1974) claims that the selection oflife events by most researchers

What do you get when you fall in love? Some health status effects 75

has tended to be atheoretical. He argues particularly for the needto take into account the life stage of the individuals being assessed.Events chosen to represent stressful conditions during adolescencewould be different from those at old age. (This was confirmed byMarkush and Favero (1974).)

Kellam proposed that at each stage of hfe the individual is immersedin one or more major social fields characterized by tasks expected ofindividuals in those social fields. He suggested that 'maladapting in aspecific social field' might increase the risk of psychiatric symptoms orphysical illness.

We propose that college students, from whom we drew our sample,are involved in a social field with a distinct configuration of life tasksand criteria of social adaptation, one of which is the development of asuccessful romantic relationship. According to Erikson (1963), the twostages encompassed by this age group are identity versus role confusion(adolescence) and intimacy versus isolation (young adulthood). Eriksonproposed that in the former stage 'falling in love' constitutes an impor-tant attempt to arrive at a definition of one's identity. In the latterstage, it represents the accomplishment of intimacy.

With respect to the three issues discussed above, we differ from pre-viously reported research in that: (1) We investigate the health con-sequences of a putatively desirable life event; (2) We focus on single lifeevents that are aspects of a larger condition as our independent variable,rather than an aggregate score of life events; and (3) We have strongtheoretical support for choosing romantic involvement as a significantevent in the life stage of our sample.

Method

The data on health status and romantic involvement were obtainedfrom a questionnaire administered to a 25 per cent sample of sociologyundergraduate classes in an Eastern University in the United Statesduring a single week in April 1978 as part of a larger study of loverelationships. Respondents were 101 males and 126 females. We askedsix questions (Table 1) about romantic involvement' and 29 questionsconcerning health status during the previous week (Table 2). Nineteenof these questions (Nos. 11 to 29) are slightly modified versions of theGeneral Well-Being Schedule (GWBS)^ of the National Center forHealth Statistics.

In order to examine the relationship between health status andromantic involvement, we undertook two different analyses, the resultsof which are shown in Table 2. First, we were interested in the effects

p gTable 1 Romantic involvement questions and response categories

12345

12345

9.435.733.0

8.013.8

11.722.120.723.921.6

(1) Since age 15, how many times have you felt yourself to bein love?^

NeverOnceTwiceThree timesMore than three times

(2) How long have you been romantically interested in him/her?^One month or lessTwo to six monthsSeven to twelve monthsOne to two yearsMore than two years

(3) Which of the statements below best describes the currentstate of your romantic interests?

Don't really know anyone you're interested in 1 8.5Know someone you're interested in, but it is a very early

stage of the relationship 2 26.5Are deeply interested in someone who is deeply interested

in you and things are going well 3 43.5Are deeply interested in someone who is deeply interested

in you and things are not going well 4 7.6Have just about, or already have, broken up with someone

you were mildly interested in 5 4.5Have just about, or already have, broken up with someone

you were deeply interested in 6 9.4

NOTE: If you answered either 5 or 6 for the preceding question,please answer the following three questions.

(4) Who caused or is causing the break-up?Exclusively you yourselfMainly you yourselfBoth of you about equallyMainly the other personExclusively the other person

(5) To what extent did you, or do you, want the break-up?Not at allSomewhatVery much

(6) Are you interested in anyone new?Not at allSomewhatVery much

a N for questions one to three = 228; for questions four to six, 31.b Numbers following response alternatives designate groups for analysis

of variance results shown in Table 2.c Preceding this and the remaining questions was the statement: "The

following questions deal with your current romantic interest...'.

12345

123

I23

6.112.139.436.4

6.1

36.445.518.2

33.354.512.1

What do you get when you fall in love? Some health status effects 77

Table 2 Analyses of variance results, romantic involvement and health

Health Status Items

RomanticInvolvementQuestion^

RomanticInvolvementGroup Means

1. Suffer from a respiratory condition suchas a common cold, virus, flu, etc.''2. Suffer from digestive system problems,such as an upset stomach3. Suffer from injury or accident.4. Have difficulty falling asleep or stayingasleep.5. Take prescribed medications such asantibiotics, tranquilizers, etc.6. Take non-prescribed medications such ascough medicine, aspirin, pillsto help you sleep etc.7. Suffer from headaches

8. Suffer from skin problems, such asrashes or severe itching.9. Stay in bed part or all of the daybecause of illness or injury10. Stay away from work or schoolbecause of illness.11. Have you been bothered by nervousnessor nerves?**12. Have you been in firm control overyour behaviour, thoughts, emotions orfeelings?13. Have you felt so sad, discouraged,hopeless or had so many problems thatyou wondered it anything was worthwhile?14. Have you been under or felt any strain,stress, or pressure?15. Have you been anxious, worried or upset?16. Have you been waking up fresh and rested?17. Have you been bothered by anyillness, bodily disorders, pains or fearsabout your health?18. Has your daily lefe been full ofthings that were interesting to you?19. Have you felt down-hearted andblue?

4(SxR)

4(R)

4,5,3,1,2,6^

2,4,3,5

2,3,4 5 f

2(SxR)4(SxR)64(R)6

l(SxR)4(R)4(SxR)6(R)3

1

4(R)

1(R)2(R)3(R)

1,3,4,5,24,2,3, 52,3 3,12,4,3 52,1 1,33,2,5,1,42,4,5,32,4,3,52,3 3,11,4,2,3,5,6^

5,2,4,3,1

4,3,2 2,5

1,5 5,2,3,42,1 1,5,4,36,2,4,1,5 4,1,5,3

21(R)

3(R)63(R)4(S)5(S)6(S)

1,3,4,5 3,4,5,21,2,5,3,4

5,3,4,2,6,12 1,34,2,6,1,3,5F,MF,MF,M

78 Kemper and Bologh

Table 2 cont.

Health Status Items RomanticInvolvementQuestion^

RomanticInvolvementGroup Means^

20. Have you been feeling emotionallystable and sure of yourself?21. Have you felt tired, worn-out, used-up,or exhausted?22. Have you felt that discussingyour problems with your currentromantic interest helps?

23. How have you been feeling ingeneral?24. How concemed or worded about yourhealth have you been? (Low score = notconcerned)25. How relaxed or tense have youbeen? (Low score = relaxed)

26. How much energy, pep, vitality haveyou felt? (Low score = No energy)27. How depressed or cheerful have youbeen? (Low score = depressed)28. How happy, pleased or satisfied haveyou been with your person life?(Low score = happy)

29. How happy, pleased or satisfied haveyou been with your progress at school?(Low score = happy)

6(R)

6(R)

3 1,2

1(S,R)2(R)3(SxR)4(S)5(S)6(S)6

4,3,2,5 5,1 F,M5,4,3 3,2,13 4,5,2 5,2,6,1F,MF, MF,M2,3 3,1

2(S)3(S)6(SxR)5(SxR)

M, FM, F2,133,1,2

3,1 2

2(R)3(R)4(S)6(S,R)l(SxR)

5,4,1,3 4,1,3,23 2,5,6,1,4M, F3,1 2 M, F3,1,2,5,4

a Numbers refer to romantic involvement questions one to six in table one.b Numbers refer to treatment groups derived from response categories of questions in table

one, ordered from low to high according to analysis of variance results.c For items one to 10, a low score = no days ill.d R following a question number means that romantic involvement was a significant main

effect in the two-way analysis of variance; S means sex was a significant main effect; S,Rmeans that both sex and romantic involvement were significant main effects; SxR means asignificant interaction effect.

e Numbers without bar(s) over them indicate order of group means obtained from two-wayanalysis of variance.

f Numbers under bar(s) are treatment groups whose means do not differ significantly accord-ing to Newman-Keuls test, after one-way analysis of variance was significant.

g Where bar covers all treatment groups Newman-Keuls test was not sensitive to specific differ-ences due to unequal size groups N's, despite significant F-ratio in analysis of variance.

h For items 11 to 22, a low score = very much so.

What do you get when you fall in love? Some health status effects 79

of romantic involvement on health status, regardless of other factors.To ascertaih this we perforrhed a one-way analysis of variance on eachhealth status item usirlg the response categories of the six romancequestions to determine 'treatment' groups. For example, romanticinvolvement questions one and two (Table 1) provided five treatmentgroups; question three, six groups, etc. Where significant F-ratios wereobtained (p <.O5), we employed the Newman-Keuls test to estabhshcutting points for significant differences. These are shown in Table 2 bythe following: groups that are not significantly different are shownunder a single bar. Where groups are significantly different they appearunder different bars. In some cases the Newman-Keuls test failed todetect any difference despite a significant overall F-ratio. This was dueto very unequal cell frequencies. When this occurred, we assumed con-servatively that only the two most extreme means were significantlydifferent.

Because it was also probable that sex-linked differences in bothhealth Status and romantic involvement would interact, we sought todisentangle these variables by means of a two-way analysis of variance,in which sex and romantic involvement were the treatment factors. Wewere particularly interested in interaction effects, since these wouldindicate that health status was variably affected by romantic involve-ment status depending on one's sex. The results of this analysis arealso shown in Table 2, where significant main effects and interactioneffects are given.

Where a group number (under Romantic Involvement Question inTable 2) is followed by an R, it signifies a significant romantic involve-ment main effect for that question; followed by an S, it means thesex main effect is significant; followed by S,R, it means both aresignificant; followed by SxR, it means the interaction effect is signifi-cant. Where an R main effect is significant, it is followed in the lastcolumn by the ordered means of the treatment groups. Where theone-way analysis was also significant for the question, the Newman-Keuls results are provided.' For health question nine, for example, theone-way analysis was significant for romantic involvement questionsix, as it was also in the two-way analysis. In addition, there was asignificant interaction effect between sex and romantic involvementquestion four for health status question nine. For purposes of brevity,we shall not discuss results in which main effects for sex only werefound, since this was not a focus of our research.

We shall analyze the results obtained for each romantic involvementquestion in order.

80 Kemper and Bologh

Results

Q. 1. Frequency of Romantic InvolvementTable 2 shows six significant relationships between number of times inlove and health status (Health Q's 8,11,13,17,22, and 29).

The health symptom, nervousness or nerves (Q. 11), is significantlyrelated to number of times in love, with those who have been in lovemore than three times more susceptible to this condition. While thisfinding offers little comfort to those who regard love and romanticinvolvement as a general palliative, we find also (Q. 17) that those whohave never been in love are more likely to suffer from bodily disorders,pains, and fears about ill health. These are somewhat contrary findings,but it suggests that the two conditions — never, and more than threetimes in love - are both deleterious. Some evidence for this is found inresponses to the item on discouragement (Q. 13). Students who havenever been in love differ significantly in discouragement from all othergroups except those who have been in love more than three times. Weinterpret this finding as indicating that college students regard bothnever having been in love as well as falling in love perhaps too often aspersonal failures. These may be stressful enough to engender healthproblems.

The third health measure significantly related to number of times inlove was the feeling that discussing one's problems with the currentromantic interest 'helps' (Q. 22). While this does not tap health statusdirectly, it is a measure of social-emotional support that may mediatethe relationship between life events and health, perhaps especiallyemotional problems (cf. Rabkin and Streuning, 1976). Eaton (1976)demonstrated that the relationship of stressful life events and mentaldisorder is stronger among the non-married and those living alone.

The curvilinear pattern of responses to the previous question on dis-couragement is repeated here. Respondents who have been in love morethan three times (group 5) do not differ significantly in claiming ithelps to discuss one's problems from those who have never been inlove (group 1). Those who have never been in love have not, despiteany romantic involvement, achieved the requisite level of intimacy forthe exposure of self and one's vulnerabilities that love is said to imply(Levinger and Snoek, 1972). But for those with frequent love experi-ences, a different explanation is possible. They may have problemsthat cannot be helped by talking about them; or these students havestopped looking for support from their romantic interest after manyrelationships have broken up. Finally, their failure to be helped may besymptomatic of a difficulty in maintaining close relationships, whichcould also account for why these respondents have been in love so often.

What do you get when you fall in love? Some health status effects 81

We also found two significant interaction effects. Skin problems(Q. 8) appear to be most prevalent among those in love three times, butthis is explained in part by the high number of females in this group.In regard to school progress (Q. 29), differences are again partitionedby sex, with females in group 3 most pleased with their progress andmales in the 'more than three' and 'never' categories most dissatisfied.

Q. 2. Duration of Current Romantic InterestIf it is true that the course of true love does not run smoothly, itappears that some periods are definitely bumpier than others. Durationof romantic interest in a particular other is significantly related to fivehealth status items (Qs 6, 13, 22, and 28). The most vulnerable periodappears to be when the relationship is between two and six monthsold (group 2). This group reports significantly greater discourage-ment and depression (Q. 13) than all other groups with the exceptionof group 1, those who have been romantically interested for one monthor less.

On the question of waking up fresh and rested (Q. 16), group 2 againhas the worst health score. With respect to how happy one has beenwith one's personal life (Q. 28), group 2 stands out as being the leasthappy, differing most from group 5, those who have been romanticallyinvolved for more than two years. The pattern is confirmed again inthe results for taking non-prescription drugs: group 2 exceeds all others.

A possible contributing factor to this disturbed health pattern forgroup 2 is evident in responses to the question of whether discussingproblems with one's romantic interest 'helps' (Q. 22). Again group 2 isat the low end of the scale, excepting group 1, which is lowest induration of involvement. The subset composed of groups 5, 4, and 3,whose romantic interests have lasted more than six months, not onlyobtain 'help' from discussing their problems, but are generally foundto have the better health status (see Table 2).

One significant interaction is found also. Taking non-prescriptionmedications (Q. 6) is especially hkely among members of group 2, aswe may expect. A low score on this, however, is contributed by anespecially heavy concentration of females in group one, where therelationship has lasted a month or less.

From these results it appears that the two-to-six-month-old relation-ship is fraught with the most problems for its participants. It is obviouslya relationship with some potential for growth into permanent com-mitment; otherwise it would have ended after the first few dates. Yetthere is apparently the growing recognition that permanence impliesobligations and responsibilities that can be as burdensome as theassured presence and commitment of the other is a delight. Blau (1964)

82 Kemper and Bologh

has outlined some of the pragmatic cost-benefit calculations that mayprecede the stage of final commitment in love relationships. They tumon such issues as not wanting to give more than one receives, and con-cern with the exclusiveness of the other's interest in oneself. Kemper(1978) points out that as love relationships develop, they usually entaildependence of the lovers on each other, and, by virtue of this depen-dence, power and threat emerge to cloud the previously idyllic relation-ship. From our data we surmise that those in relationships of two-to-sixmonths duration are particularly vulnerable to problems of commitment,calculation, dependence, and power, and that this leads to stress anddisturbances of health.

Q. 3. Current State of RomanticlnvolvementTable 2 shows that the current state of the romantic relationship issignificantly related to seven items of health status (Q.'s 1, 10, 13, 18,19, 22, and 28). We find a significant relationship between currentstate of romantic interest and number of days missed from work orschool because of illness or injury (Q. 10). A romantic relationshipthat is breaking up (groups 5 and 6) is most likely to break into thenormal pattern of work or school attendance, due to ill health. Thosein group 6 report the most impairment from common colds, fiu,virus, etc. (Q. 1).

Those who are in a good relationship (group 3) are significantly lessdiscouraged (Q. 13) than group 6, those who have just broken up, andgroup 2, for whom the relationship is at a very early stage. We under-stand why group 6 feels bad. The group 2 findings also could be anti-cipated because these are the same people who are in an early stage ofthe relationship (between two-and-six-months) and are feeling dis-couraged about it, as we found in the analysis of the duration questionabove.

Additionally, people in group 3 are significantly happier with theirpersonal lives (Q. 28) than any other group. Similarly group 3 differssignificantly from all others in reporting that discussing problems withtheir current romantic interest 'helps' (Q. 22).

It is also no surprise to learn that those who are in good relationships(group 3) find life interesting (Q. 18) and are not down-hearted and blue(Q. 19). It is surprising, however, to find that those who have brokenup with a mild romantic interest (group 5) are even more interested inlife and less blue. Though this group misses work or school due toillness, it also appears to be responding with lively interest to new-foundfreedom and release from a not very stimulating relationship.

In general, poor health status seems to follow from the process ofbreak-up or when the relationship is uncertain or troubled.

What do you get when you fall in love? Some health status effects 83

Those who were placed in groups 5 and 6, indicating an actual orimminent break-up of their relationship, were asked three additionalquestions, analyzed below. Because there were so few cases in thesegroups (31 in all), the results must be treated with extreme caution.While they are not necessarily unreliable, they are potentially so due tounrepresentative sampling in some of the cells.

Q. 4. Who Caused the Break-UpFor romantic relationships that are breaking up or have just broken up.Table 2 shows that perception of who caused the break-up is signifi-cantly related to seven health status items (Q.'s 2, 4, 5, 7, 8, 9, and 12).Group 5, those who claim that the other person exclusively caused thebreak-up, is significantly worse than the other groups on all the healthitems: falling asleep or staying asleep (Q. 4); taking non-prescriptionmedication (Q. 6); suffering from headaches (Q. 7); and being in firmcontrol over one's behaviour, thoughts, emotions or feehngs (Q. 12).However, we must be cautious in interpreting these findings as group 5has an N of only 2.

These results, taken at face value, raise a question about the healtheffects of locus of responsibility for loss. Dohrenwend (1974) reportsthat loss events for which the person felt responsible were more likelyto lead to psychopathology. Here we see some reversal of this, namelythat ill effects result from attributing the loss to the other. The emotionthis would entail is anger rather than depression as is implied byDohrenwend's finding (Kemper, 1978). Indeed, none of the healthstatus items refiecting depression are significant for this romanticinvolvement question. Thus, ill health effects may stem from excess ofanger or depression following a significant life change.

We found, also, four significant interactions between sex and romanticinvolvement as this affected health status. Digestive upset (Q. 2) wasmost prevalent in group 2, those who were themselves mainly respons-ible for the break-up, but this was due to females in this group. Malesin group 4, attributing the break-up mainly to the other, were especiallyunlikely to be taking non-prescription medication (Q. 6). Maleswho attributed the break-up exclusively to the other, were not solucky; they contributed to an excess in likelihood of staying in bedbecause of illness or injury (Q. 9). Females who allocated responsibilityabout equally (group 3) were more likely to suffer from skin prob-lems.

These interactions must be treated gingerly, due to small N's in somecells, but are suggestive of the sex-linked aspects of how health statusis affected by romantic involvement. In this instance, the attribution ofblame differs somewhat according to sex, and this produces some

84 Kemper and Bologh

different emotions, which in tum may be assumed to entail differentforms of organismic stress and illness.

Q. 5. Did You Want the Break-Up?No significant findings relating health and romantic involvement wereobtained for this question. This suggests that the degree of control overthe break-up (romantic question 4), rather than wanting it or not is thedecisive aspect with respect to health.

Q. 6. New Romantic InterestFor relationships that have just broken up, health effects are mitigatedor exacerbated depending on degree of involvement with a new romanticinterest which is related significantly to eight health items (Q.'s 6, 7, 9,18, 21, 23, 27, and 28). In general, a mild interest in someone new(group 2) is preferable to either no interest (group 1) or a very stronginterest (group 3). Group 2 is less likely to have taken medication (Q.6), suffered from headaches (Q. 7), stayed in bed (Q. 9), felt discouraged(Q. 13), and was more likely to find life interesting (Q. 18) and satisfy-ing (Q. 28). In fact, while keeping in mind that the N for group 3 isvery small, it looks as if becoming very interested in someone newimmediately after breaking up with a previous romantic interest canmake one feel very tired (Q. 21).

There are two significant sex-linked interaction effects here. Malesare more likely to be among those who stay in bed (Q. 9) if they havenot found anyone new (group 1). And females are more likely to feelrelaxed (Q. 25) when they have a mild interest in someone new (group2).

Conclusion

We began this research into the effects of romantic involvement onhealth status with three questions from the literature on the healthconsequences of life events: (1) whether only undesirable events pro-duce negative health effects; (2) whether single life events or a com-posite of events have this consequence; and (3) whether certaintheoretically indicated events are especially consequential for health atparticular life stages. In choosing romantic involvement as our focalevent, we intended to examine an experience that is widely reputed tohave both desirable as well as undesirable elements. In this sense it is amixed event. When it is going well, as our data show, the romanticrelationship clearly produces no negative health effects. On the otherhand, when it is going badly, a variety of health effects are reported. An

What do you get when you fall in love? Some health status effects 85

unanticipated finding here was the curvilinear effect of duration ofinvolvement on health status, with the two-to-six month relationshipproving to be especially troublesome.

On the question of desirability-undesirability, our research supportsthe view that specifically undesirable events have negative healthassociations and that desirable life events have positive associations withgeneral well being. However, the research also suggests that life eventssuch as romantic involvement cannot be judged as desirable or un-desirable in themselves. We find, for example, that within the contextof a life experience generally considered to be desirable, such as aromantic relationship, there may be undesirable stages and shadings.

This ties in closely with the question of whether single or aggregatelife events produce health consequences. Although much of the researchon life events and their health effects has focused on aggregates of lifeevents, we have found that a single life event, that of romantic involve-ment, does have an impact on health status. Our research supports fheview that a single specific, long term event or condition may besufficient to affect health. This finding indicates the need for moreresearch on the health effects of single life events.

However, a strictly singular life event may be a misconception. Forexample, if losing a job is a singular event it ramifies nonetheless intoa number of different relationships e.g., with spouse, children, parents,friends, et al. Also, it will signify something different for a singleperson with few emotional or financial dependents than for a marriedperson supporting both a nuclear family and parents. The impact ofjob loss also varies depending on its duration, since this is related to un-employment compensation. Thus, strictly speaking, as a single event itis quite complex. Similarly, romantic involvement is singular, butcomplex, proceeding through stages of commitment and resolution ofrelational problems. Our data show that stable, reciprocated, romanticfeelings have no ill health effects. But stability is a matter of durationover time, allowing for the cumulation of good feelings and the success-ful handling of inevitable challenges, even with time-out for argumentsand anger. Thus, in an important sense, a stable romantic relationshipis a composite of experiences that are mainly but not exclusivelypositive. It may, perhaps, be found, that virtually all single events —whether desirable or undesirable — are better understood as compositesin their own right.

On the question of whether particular events are particularly con-sequential at specific life stages, our data do not permit an answer,since we did not compare the effects of romantic involvement betweenwidely different age groups. We did, however, choose to investigate thequestion in a young adult group, since it is widely accepted that the

86 Kemper and Bologh

establishment of a stable love relationship is one of the developmentaltasks of that life stage.

In general, for this age group, we found significant relationships tohealth with: number of times in love; duration of the romantic interest;current state of the romantic interest; and for those who have justbroken up: who caused the break-up and whether or not one wasromantically interested in somebody new. A relationship of longduration that is going well appears to have a positive effect on one'shealth status.

The results also show that, at least for this sample, the adverse healtheffects of love relationships can extend to physical ailments thatrequire bedrest or medication. In some instances this occurs through aninteraction effect with sex. The strongest adverse health reactions occuramong those whose relationships are in the two-to-six month period ofachieving commitment, and among those whose relationships havebroken down. Since most young adults experience one or more brokenrelationships, and since even relationships that last must go through thetwo-to-six month period, it is likely that many persons experiencehealth effects of falling in love, even when the relationship grows intostability. Hence the song lyrics with which we began may overstate thecase only a little: perhaps not often pneumonia, but, according to ourresults, colds, virus, fiu, digestive disorder, headache, skin rash, fatigue,nerves, discouragement, and the like.

Department of Sociology,St. John's University

New York

Notes

1. We are aware that when respondents answer such questions as 'Since age 15,how many times have you felt yourself to be in love?' they may not all meanthe same thing by 'love'. For some, a relatively light emotional involvementmight mean love, while for others it requires a very heavy emotional invest-ment. This is a general problem in social research, the solution to which canbecome methodologically quite complex. We have adopted the view - accept-ing the proper caveats — that there is some value in investigating the con-sequences of a romantic state defined by the actor. When research on objectivestates of actors' romantic involvements is available, the appropriate com-parisons can be made.

2. The GWBS was designed to provide indicators of the presence and severity orfrequency of selected symptoms considered important with respect tosubjective well-being or distress (Dupuy, 1974). It has been administered to asample of 6,931 noninstitutionalized adults (ages 25-74) in the United Statesas part of the National Health and Nutrition Examination Survey and a 22-item

What do you get when you fall in love? Some health status effects 87

version was used by Mueller, et al. (1977) in their study of stressful life eventsand psychiatric symptomatology. The GWBS was found to correlate as highlywith other, more extensive, measures (the Minnesota Multiphasic PersonalityInventory and the College Health Questionnaire) as these other measures didamong themselves (Vital and Health Statistics Series 2, No. 73, ValidationalStudy of the GWBS Schedule).

3. The one-way and two-way analyses are not fully compatible since the sums ofsquares and degrees of freedom are partitioned differently.

References

Peter Blau, Exchange and Power in Social Life. New York: Wiley, 1964.George W. Brown, 'Meaning, measurement, and stress of life events', in B.S.

Dohrenwend and B.P. Dohrenwend (eds). Stressful Life Events: Their Natureand Effects. New York: WUey, 1974, pp. 217-43.

Ernest Burgess, 'The romantic impulse and family disorganization.' Survey, vol.57, 1926, pp. 290-4.

Richard C. Centers, 'Attitude similarity-dissimilarity as a correlate of heterosexualattraction and love.' Journal of Marriage and Family, vol. 37, 1975, pp. 305-12.

Barbara S. Dohrenwend et al. 'Exemplification of a method for scaling Ufe events:the PERI life events scale.' Journal of Health and Social Behaviour, vol. 19,1978, pp. 205-229.

Bruce P. and Barbara S. Dohrenwend (eds). Stressful Life Events: Their Nature andEffects, New York: Wiley, 1974.

Bruce P. Dohrenwend, 'Overview and prospects for research on stressful life events,in B.S. Dohrenwend and B.P. Dohrenwend (eds), op. cit., pp. 313-33.

Bruce P. Dohrenwend, 'Problems in defining and sampling the relevant populationof stressful life events', in B.S. Dohrenwend and B.P. Dohrenwend (eds), op. cit.,pp. 275-310.

Barbara S. Dohrenwend, 'Life events as stressors: A methodological inquiry.'Journal of Health and Social Behaviour, yo\. 14, 1973, pp. 167-75.

H.J. Dupuy, 'Utility of the National Center for Health Statistics' General Well-beingSchedule in the assessment of self-representations of subjective well-being anddistress.' Paper presented at the National Conference on the Evaluation of Drug,Alcohol, and Mental Health Programs (April 1974).

William W. Eaton, 'Life events, social supports and psychiatric symptoms: a re-analysis of the New Haven data.' Journal of Health and Social Behaviour, vol.19, 1978, pp. 230-233.

Erik Erikson, Childhood and Society. New York: W.W. Norton & Co., 1973.Erich Fromm, The Art of Loving. New York: Harper and Row, 1956.Gersten, Yvonne C. et al, 'An evaluation of the etiologic role of stressful life-

change events in psychological disorders.' Journal of Health and So:ialBehaviour, vol. 18, 1977, pp. 228-244.

Gersten, Yvonne C. et al., 'Child behaviour and life events: Undesirable change orchange per se?' in B.S. Dohrenwend and B.P. Dohrenwend (eds), pp. 159-170.

Thomas H. Holmes and Richard R. Rahe, 'The social readjustment rating scale.'Journal of Psychosomatic Research, vol. 11, no. 2, 1967, pp. 3-18.

Thomas H. Holmes and Minoru Masuda, 'Life changes and illness susceptibility,' pp.45-72 in B.S. Dohrenwend and B.P. Dohrenwend (eds), op. cit., pp. 45-72.

88 Kemper and Bologh

Karen Homey, Growth Through Love and Sex. New York: Auxiliary Council tothe Association for the Advancement of Psychoanalysis, 1946.

Sheppard G. Kellam, 'Stressful life events and illness: A research area in need ofconceptual development,' in B.S. Dohrenwend and B.P. Dohrenwend (eds),op. cit., pp. 207-216.

Theodore D. Kemper, >1 Social Interactional Theory of Emotions. New York:Wiley, 1978.

William M. Kephart, 'The "dysfunctional" theory of romantic love: A researchreport.' Journal of Comparative Family Studies, vol. 1, 1970, pp. 26-36.

George Levinger and J. Snoek, j4ffracfion in Relationships: A New Look at Inter-personal Attraction. New York: General Leaming Press, 1972.

Robert E. Markush and Rachel V. Favero, 'Epidemiologic assessment of stressfullife events, depressed mood, and psychophysiological symptoms — a preliminaryreport', in B.S. Dohrenwend and B.P. Dohrenwend (eds), op. cit., pp. 171-190.

Abraham H. Maslow, 'Love in healthy people', in A. Montagu (ed.). The Meaningof Love. New York: Julian Press, 1953, pp. 57-93.

Daniel P. Mueller et al, 'Stressful life events and psychiatric symptomatology:change or undesirabUity?' Journal of Health and Social Behaviour, vol. 18,1977, pp. 307-317.

J.G. Rabkin and E.L. Streuning, 'Life events, stress and illness.' Science, vol. 194,1976, pp. 1013-20.

Richard H. Rahe, 'The pathway between subjects' recent life changes and theirnear-future illness reports: Representative results and methodological issues.'Pp. 73-86 in B.S. Dohrenwend and B.P. Dohrenwend (eds), op. cit., pp. 73-86.

Zick Rubin, The Social Psychology of Romantic Love. Unpublished Ph.D.Dissertation, University of Michigan, 1969.

A. Vinokur and M.L. Seltzer, 'Desirable versus undesirable life events: Theirrelationship to stress and mental distress.' Jourrml of Personality and SocialPsychology, vol. 32, 1975, pp. 329-37.