• complete cross-sectional survey;
  • depressive symptoms;
  • middle age;
  • social support;
  • stress buffering effect;
  • suicide prevention


  1. Top of page
  2. Abstract

Abstract  Little is known about the association between depression and the buffering effects of social support in mid-life crisis. The aim of this study is to determine the buffering effects of social support on depression concerning middle-aged individuals, while also taking reciprocity and gender differences into careful consideration. A cross-sectional survey of all middle-aged individuals (40–69 years of age) using a large sample (n = 4558) from a community-living population, who resided in Rokunohe town, Aomori prefecture in northern Japan (response rate = 69.8%), was undertaken. This town recently had a lot of suicides. Two-way anova was used to analyze the effects of stressor and social support on the Center for Epidemiologic Studies Depression scale scores. The authors found a stress buffering effect of social support on the depressive symptoms occurring in middle age, however, a significant difference in the stress buffer effect was only observed in male subjects. Moreover, when the authors take reciprocity into account, the effect of the buffer on depression was found not only in males receiving support but in males providing support as well. In conclusion, pertaining to males, social support reduces depressive symptoms under stressful circumstances in middle age, not only when they receive such support but also when they provide it. Therefore, these findings suggest that reciprocal social support is important for males in relation to community mental health.


  1. Top of page
  2. Abstract

It is said that low levels of social connections increase the risk of death. In an Alameda County Study, California, USA, the most isolated group of men was found to have an age-adjusted relative mortality rate 2.30-fold higher than the men with the most connections, and the rate was 2.8 times higher than for the women.1 In addition, the lack of social ties is also considered to have a large influence on mental health.2

Numerous studies have shown a positive relationship between social support and mental health, especially depression.3–6 As a result, as for Public Health, it is said that social support can improve the standard of health, however, there are still many unclear factors regarding how social support influences mental health.

Two models have been used to explain how social relationships influence health,7 namely the main effect model and the stress buffering model by Cohen and Wills.8,9

According to the main effect hypothesis, people with relatively strong social support have better health than those with weak social support, irrespective of their exposure to stressors. In other words, a lack of social support or isolation tends to negatively influence health by stressors. In contrast, according to the indirect effect (buffer) hypothesis, people with strong social support tend to have better health than those with weak social support, but only with respect to exposure to stressors. In other words, support buffers protect persons from the potentially pathogenic influence of stressful events.10

The study of the stress buffering effect by social support has been researched in earnest since about 1980.11–14 There have been some community-based investigations about middle-aged individuals, however, several of them were limited studies, such as for African–American subjects, which took religion into account,15 or consisting only of married men,16 or married women.17 Moreover, research on the elderly,10,18–21 or a specific group (e.g. twin women,22,23 HIV patient,24 pregnant or postpartum women,25,26 students27,28) have been reported after the 1990s when Cohen’s theory was popular, and few such studies have been made on middle-aged individuals.29

In contrast, the reciprocity of social support should also be considered.30–32 The balance between help provided (in other words, giving or donated) and help received (e.g. instrumental support, emotional support) correlated negatively with any influence on health.33 Because, when an individual receives social support, the person feels a burden, indebtedness, or load. Recently, the study of the reciprocity has been performed, not only concerning well-being of persons,33,34 but also respecting the buffering effect. In the latter case, only older adults18,35 or students36 have so far been studied.

However, little research has previously been done for the middle-aged in a community-residing population studying the stress buffering effect, particularly in a type of social support based on reciprocity. In this article, the authors should clarify the stress buffering effect of social support on depressive symptoms in a large community sample of middle-aged individuals (40–69 years of age), by comparing males and females, and then separating them into instrumental, emotional and providing (giving) support.

In addition, these days in Japan, an increased suicide rate among middle-aged individuals remains a huge problem.37 As a result, a region with a high suicide mortality rate in Japan was selected for this investigation.38–40 To understand the actual conditions concerning mental health, establishing suicide prevention systems for farm villages was another purpose of this research. Since it is a factor in improving mental health, the social supports for middle-aged individuals in local areas were investigated.


  1. Top of page
  2. Abstract

Study population

Rokunohe town, a small town in Aomori prefecture, located in the northern part of Japan, was selected for investigation in this study. There are 11 099 (5402 male, 5697 female) current residents as of 2003 and this town is also in an agricultural region (with rice cultivation, pig farms and root vegetables). The study population compared middle-aged subjects ranging from 40 to 69 years of age.

The number of samples totaled 4558 persons consisting of 2333 males and 2225 females, who lived in Rokunohe town as of 1 September 2003. The research method used was the placement method using a self-administered questionnaire that did not have to be signed. Volunteers (the public health supporters were appointed by the town office) distributed and collected the questionnaires to all the homes in the town. The investigation period was from 7–15 September 2003 and the response rate was 69.81% (3182 persons). A total of 3132 persons questionnaire was used, due to 50 persons who did not answer the gender. The authors investigated the overall mental health of the subjects (e.g. stress coping, suicidal ideation). This paper mainly reports on the stress and social support available for these individuals.

Ethical protocol

The questionnaire was anonomous and no connection could be made to each resident. The investigation was approved by the town mayor and as part of the prefecture and town enterprise. Informed consent had to be given at the front of the questionnaire. Moreover, the volunteers explained that the participants were able to refuse to answer the questions if they so desired.


The demographic characteristics on each resident’s age, gender, marital status, occupation, living arrangements, communication with friends, hospital visits, sleep and subjective health sense was recorded. In addition, the range of illnesses included 15 diseases.

Depression was measured based on the 20-item Center for Epidemiologic Studies Depression scale (CES-D).41 And the Social Support scale used 10-items of the Measurement of Social Support-Elderly (MOSS-E).42 Both scales were certified as to reliability and validity.43,44 Although this Social Support scale was originally established for the elderly, it has also been applied to middle-aged individuals.45

The full Social Support scale consists of three subscales: Instrumental support (three items, e.g. cleaning and cooking, shopping, gardening), Emotional support (four items, e.g. on the side, consultation, encouraged, careful), and Providing (giving) support (three items, e.g. do the housework, nursing, shopping substitute for someone). All the items in these scales are shown in Table 1.

Table 1.  Measurement of Social Support-Elderly
  1. (1)–(3) instrumental support, (4)–(7) emotional support, (8)–(10) providing support. These items are scored in yes = 1, no = 0.

(1) Is there anyone you can ask to help you with cooking and shopping four your daily needs?
(2) Is there anyone you can ask to help you with your gardening, cleaning, washing etc.?
(3) Is there anyone you can easily ask to help you with other chores?
(4) Is there a person who cares for you and by your side when you are in difficulty?
(5) Is there someone you can talk to or ask for advice when you are worried?
(6) Is there someone who encourages you when you are down or feeling depressed?
(7) Is there someone who is concerned about you and is interested in your welfare?
(8) Is there someone you often help or do housework for?
(9) Is there someone you shop for or often help?
(10) When your friend or neighbor is sick in bed, do you care for and nurse them?

Providing support is briefly explained as follows: The reciprocity of social support is the balance of received support and provided support. If the balance or reciprocity of transfer cannot be achieved, then a feeling of burden and influence on a persosn well-being occurs.32,33

The social support was analyzed by dividing support into various types, namely instrumental, emotional and providing support. An answer of yes was scored as 1 point, and a perfect score was 10 points (instrumental, 3 points; emotional, 4 points; providing, 3 points; and total, 10 points). The question is as follows: ‘Ask about people (a neighbor, a friend, separated child) except for family members who live together’.

Furthermore, regarding stressors, a question was asked with four answers: ‘Did you experience any dissatisfaction, worries, hardships, or stressors in your daily life in the past month?’ (1, much; 2, sometimes; 3, not so much; 4, nothing).

Data analysis

There were not many valid answers using the scale of CES-D and MOSS-E. Both scales used the full response without any blanks. Especially with respect to CES-D, any wrong answers on reverse items were deleted. In addition, both scales were evaluated by Cronbach’s alpha reliability coefficient and by the Kolmogorov–Smirnov test of normality.

The χ2-test was used to describe the statistics between males and females. The Kruskal–Wallis test and Mann–Whitney U-test were used for comparisons between the demographic characteristics groups to analyze the related factors of each scale. At this time, the items concerning the degree of stressors and of the communication with friends were used to analyze CES-D. In contrast, the items of related health were used to analyze MOSS-E as social support. Moreover, the Bonferroni correction was applied when multiple comparisons were made. On that occasion, not the median, but the average value was shown in the list.

The buffering-effects were analyzed separately in males and females, and they were reported as significant when interactions were observed between the stressor and the three categories of social support. The two-way anova was used when analyzing the effects of stressors (high-low) and social support (high-low) on CES-D scores as the dependent variable.8 The main effect model was a case in which the main effect was significant in two-way anova. In contrast, the buffering effect model was a case in which the interaction was significant.

In addition, the authors distributed the individuals among those with ‘high social support, low social support’, and with ‘high stress, low stress’. For each social support subscale the scores were recoded as a dichotomous variable ‘high’ and ‘low’ by a median (‘high’ are instrumental, 3 points; emotional, 4 points; providing, 3 points; and ‘low’ are instrumental, 0–2 points; emotional, 0–3 points; providing, 0–2 points). Furthermore, stressor also recoded ‘high (heavy and sometime)’and ‘low (not much and nothing)’.

Two-way anova was used to analyze the answers, namely all the items (stressor, depressive scale, social support scale) including instrumental: 1481 (male high 341, low 336; female high 426, low 378); emotional: 1466 (male high 397, low 286; female high 536, low 247); and providing: 1378 (male high 317, low 322; female high 406, low 336).

All analyses utilized the spss windows software package version 10.0 (SPSS, Chicago, IL, USA).


  1. Top of page
  2. Abstract

The mean age of the residents was 54.9 years (SD = 8.4). Female subjects comprised 53.6% of the overall subjects. Other backgrounds were as follows: married (81.5%), living alone (4.8%), farmers including as a side-job (38.9%). The grades of stressor were classified into the following four steps (n[%]); ‘heavy’ 428 (14.8%), ‘sometimes’ 1423 (49.2%), ‘not much’ 731 (25.3%) and ‘nothing’ 311 (10.8%).

The factors causing the most stressors were as follows: ‘economical problems’ 879 (31.2%), ‘work’ (superior official, contents, responsibility etc.) 595 (21.1%), ‘health and illness’ 509 (18.1%), and ‘interpersonal relations’ 495 (17.6%). Particularly, ‘work’ (male 28.2%, female 14.9%; χ2 = 74.1, P < 0.001) and ‘interpersonal relations’ (male 12.2%, female 22.3%; χ2 = 48.2, P < 0.001) showed a significant difference between males and females. (This question had multiple answers for a few items for both genders.)

The states of the CES-D score were as follows: effective answers were 55%, Cronbach’s alpha was 0.82, and the scores of light (≥16) were 33.5% male, 39.0% female; moderate (≥21) were 19.4% male, 20.8% female; and heavy (≥31) were 4.1% male, 4.6% female.

Table 2 shows the means and standard deviations (SD) of CES-D score by Kruskal–Wallis test, each item about age group, marital status, stressors, and communication with friends. The more an individual suffered from stressors, and the fewer relationships they tended to have with friends of both genders, the higher CES-D score levels. However, the CES-D score did not show a significant difference in relationship to the age group or the marital status of females.

Table 2.  Related factors of Center for Epidemiologic Studies Depression scale score Thumbnail image of

Moreover, the social support score (MOSS-E) findings were as follows: effective answers were 73%, Cronbach’s alpha was 0.91 (subscale; Instrumental = 0.85, Emotional = 0.90 and Providing = 0.78). In addition, the average score (Mean ± SD total and three subscale) among genders were: male 6.60 ± 3.44 points (institute 1.98 ± 1.25, emotional 3.03 ± 1.50, providing 1.58 ± 1.24), and female 6.96 ± 3.17 points (institute 2.05 ± 1.21, emotional 3.24 ± 1.34, providing 1.65 ± 1.23).

Table 3 presents the means and SD of MOSS-E score by non-parametric test showing each item about subjective health sense, sleep, hospital visits, and the number of illnesses. Statistically significant associations were found between social support and subjective health sense, the number of illness, in both genders. However, a significant association was found for sleep in males.

Table 3.  Related factors of Measurement of Social Support-Elderly Thumbnail image of

Table 4 and Fig. 1 show two-way anova results of stressor (high-low) and social support (high-low). An apparent stress-buffering effect was found only in males, as shown by the significant interactions between stressors and all of the three categories of social support such as instrumental F (1673) = 7.17 P = 0.008, emotional F (1679) = 7.21 P = 0.007 and providing support F (1635) = 12.57 P < 0.001.

Table 4.  Result of two-way anova of stressor (high–low) and social support (high–low) divided by subscales
  1. Dependent variable is depression score.

  2. emo.sup, emotional support; Institutesup, instrumental support; pro.sup, providing support; SS, sum of square.

Instrumental support MaleInstrumental support Female
Stressor5 475.215475.2103.08<0.001Stressor3 710.313710.366.25<0.001
Stressor * Institutesup.380.91380.97.170.008Stressor * Institutesup.108.01108.01.930.165
Error35 748.367353.1  Error44 802.680056.0  
Total179 951.0677   Total234 531.0804   
Emotional support MaleEmotional support Female
Stressor5 201.815201.8100.60<0.001Stressor2 925.212925.255.96<0.001
Stressor * emo.sup.372.81372.87.210.007Stressor * emo.sup.131.01131.02.510.114
Error35 107.967951.7  Error40 719.777952.3  
Total180 731.0683   Total224 760.0783   
Providing support MaleProviding support Female
Stressor5 117.515117.596.73<0.001Stressor3 533.013533.065.55<0.001
Stressor * pro.sup.664.91664.912.57<0.001Stressor * pro.sup.21.2121.20.390.531
Error33 596.463552.9  Error39 614.273553.9  
Total168 489.0639   Total209 734.0739   

Figure 1. Stress-buffering effects of social support on depressive symptoms divided into subscales and genders (a,c,e, men; b,d,f, women). (a,b) Instrumental support; (c,d) emotional support; (e,f) providing support. (◆), low; (●), high.

Download figure to PowerPoint

Moreover, the main effect was observed to have a significant direct effect on depressive symptoms only for emotional support. However, neither instrumental support nor providing support was found to have a main effect in both genders.


  1. Top of page
  2. Abstract

The social support was related to health based on the findings of this investigation (Table 3). Especially, the relationship between the social support and mental health was analyzed by those who received and provided support (Fig. 1).

The authors have shown that the stress buffering effect of social support on depressive symptoms in a community sample of middle-aged individuals, showed a significant difference and a stress buffering effect only in males. Therefore, in situations with a high amount of stress, and a high degree of social support, the depression score was lower than a low social support. This buffer was effective in male, but not in female subjects, in addition to all subscales. Moreover, when analyzing reciprocity, a buffer effect on depression was found in males by providing support. As a result, not only receiving but also giving support, is important for males. These findings are very unique, because many recent studies did not examine gender differences.18,34,36

For all that, Schieman and Meersman researched for the elderly person in both genders.35 There was neither a main effect nor a buffering effect for the depressive symptoms when the male and female provided support.

There have been several studies on gender differences concerning stressors and social support, although about stress-buffer, no significant differences have been observed. While these results are consistent with recent studies, studies on Caucasian female twins and adolescents students showed no buffering effect for female subjects.22,27 In contrast, however, few previous studies have shown that a stress-buffering effect of social support may be somewhat stronger among women than among men.46 Therefore, women’s interpersonal ties tend to be more emotional than those of men.47 Nevertheless, as for the examination of gender difference of this study, the analysis is considered to have been too crude because of very weak correlations.

Why do females not have a ‘buffering effects of social support on depression’?

This finding may be related to differences in the ‘social networks’ among the genders.

Regarding these gender differences, Deborah Belle observed that there is plentiful evidence that males tend to participate in more activity-focused relationships than do females, while female subjects at all ages maintain more emotionally intimate relationships than do male subjects.48

What is most striking about men’s mobilization of support is that it is so heavily focused on one support provider, namely the wife.48

However, social ties also bring additional stresses into women’s lives, and in many relationships women appear to provide more support to others than they receive in return.49

Furthermore, Kawachi and Berkman summarize their interpretation of an opinion of the Belle findings: (i) women maintain more emotionally intimate relationships than men; (ii) they mobilize more social supports during periods of stress than men; and (iii) provide more frequent and more effective social support to others than do men.7

As an aside, as to the fact that females did not show any relationship to the ‘buffering effects of social support on depression’, readers should think about any potentially confounding variables. About 40% of the middle-aged individuals were engaged in agriculture in Rokunohe town. However, it seemed that interpersonal relationships were very close. As a result, the frequency of interchange with friends among females was higher than in males in this study (Table 2). Possibly, there might be a ‘negative support’ in a tight interpersonal relationship.

The reason why this study is extremely important is as follows. This is a study on the stress-buffering effect about the social support for middle-aged (40–69 years) individuals in a large community sample, which was divided into genders and subscales to take reciprocity into consideration. In addition, the authors clearly demonstrated gender differences relating to the effects of social support. Moreover, these results suggest that males who not only receive but also provide social support tend to show a decrease in depressive symptoms under stressful circumstances in the middle aged. As a result, social support may be one possible prevention method in high suicide areas. Recently, a prospective cohort study of the general population in Japan confirmed that isolation (living alone, being divorced) increased the risk of suicide in men.50

Some limitations in these findings include some important factors. The measurement of stressors remains difficult to evaluate. This measure might not show the stressful life events or daily hassles, and the findings may be confounded with stress responses or strain. It is also difficult to evaluate the quantity, quality and strength of social support. In addition, the response rate of the depression scale was not high with 55% (1639 persons), therefore, analytical object of two-way anova was scarcely 1400 persons. This number of people analyzed in the study is considerably small in comparison to the study population (4558 persons). Nevertheless, both analytical objects of two-way anova (about 1400 persons) and investigation participants (3132 persons) did not differ substantially concerning backgrounds. Third, a cross-sectional survey showed a weak causal relationship between social support and depressive symptoms. Finally, these findings also are related to sociocultural factors, for example, the women’s continuing low status in farming areas or a tight interpersonal relationship. As a result, these sociocultural factors are considered to be a confounding variable. Moreover, these results may be limited to just the region in north Japan.

Readers should, therefore, keep these points in mind concerning the findings of this study, that is to say, carefully generalize the investigation result.

In conclusion, the authors herein showed a positive relationship between social support and mental health. However, when analyzing reciprocity, the authors only identified a significant difference and a stress buffering effect in males.


  1. Top of page
  2. Abstract

This study was supported by Research on Health Services of Health Labour Science Research Grant. (The research theme is ‘Establishment of a system of suicide prevention in Aomori Prefecture’.) The authors wish to express their gratitude to Chikako Tsujiura and Ritsu Sasaki, public health nurses in the Rokunohe Town Office, for their valuable support in this investigation; to Dr Mallet of Ryukyu University, professor Brian Quinn of Kyusyu University, and Dr Thomas R. D. Hadden of Hachinohe Junior College for their kind advice on the English usage; and to associate professor Hidemi Todoriki of Ryukyu University, for valuable advice on the epidemiological investigation.


  1. Top of page
  2. Abstract
  • 1
    Berkman LF, Syme SL. Social network, host resistance, and mortality: a nine-year follow-up study of Alameda County residents. Am. J. Epidemiol. 1979; 109: 186204.
  • 2
    Caplan G. Kondo K, Mashino H, Miyata Y (Trans.). Support Systems and Community Mental Health. Behavioral Publications, New York, 1974 (in Japanese).
  • 3
    Oxman TE, Berkman LF, Kasl S, Freeman DH Jr, Barrett J. Social support and depressive symptoms in the elderly. Am. J. Epidemiol. 1992; 135: 356368.
  • 4
    Fuhrer R, Dufouil C, Antonucci TC, Shipley MJ, Helmer C, Dartigues JF. Psychological disorder and mortality in French older adults: do social relations modify the association? Am. J. Epidemiol. 1999; 149: 116126.
  • 5
    Bazargan M, Hamm-Baugh VP. The relationship between chronic illness and depression in a community of urban black elderly persons. J. Gerontol. B. Sci. Soc. Sci. 1995; 50: 119127.
  • 6
    Zunzunegui MV, Beland F, Otero A. Support from children, living arrangements, self-rated health and depressive symptoms of older people in Spain. Int. J. Epidemiol. 2001; 30: 10901099.
  • 7
    Kawachi I, Berkman LF. Social ties and mental health. J. Urban Health 2001; 78: 458467.
  • 8
    Cohen S, Willis TA. Stress, social support, and the buffering hypothesis. Psycho. Bull. 1985; 98: 310357.
  • 9
    Cohen S, Underwood LG, Gottlieb BH. Social Support Measurement and Intervention. A Guide for Health and Social Scientists. Oxford University Press, New York, 2000.
  • 10
    Dalgard OS, Bjork S, Tambs K. Social support, negative life events and mental health. Br. J. Psychiatr. 1995; 166: 2034.
  • 11
    Dean A, Lin N. The stress-buffering role of social support. Problems and prospects for systematic investigation. J. Nerv. Ment. Dis. 1977; 165: 403417.
  • 12
    Wilcox BL. Social support, life stress, and psychological adjustment: a test of the buffering hypothesis. Am. J. Community Psychol. 1981; 9: 371386.
  • 13
    Aneshensel CS, Stone JD. Stress and depression: a test of the buffering model of social support. Arch. Gen. Psychiatry 1982; 39: 13921396.
  • 14
    Brandt PA. Stress-buffering effects of social support on maternal discipline. Nurs. Res. 1984; 33: 229234.
  • 15
    Brown DR, Gary LE. Stressful life events, social support networks, and the physical and mental health of urban black adults. J. Hum. Stress 1987; 13: 165174.
  • 16
    Syrotuik J, D’Arcy C. Social support and mental health: direct, protective and compensatory effects. Soc. Sci. Med. 1984; 18: 229236.
  • 17
    Krause N. Stress, control beliefs, and psychological distress: the problem of response bias. J. Hum. Stress 1985; 11: 1119.
  • 18
    Wallsten SM, Tweed DL, Blazer DG, George LK. Disability and depressive symptoms in the elderly: the effects of instrumental support and its subjective appraisal. Int. J. Aging Hum. Dev. 1999; 48: 145159.
  • 19
    Liang J, Bennett JM, Krause NM et al. Stress, social relationship, and old age mortality in Taiwan. J. Clin. Epidemiol. 1999; 52: 983985.
  • 20
    Murrell SA, Norris FH. Differential social support and life change as contributors to the social class-distress relationship in older adults. Psychol. Aging 1991; 6: 223231.
  • 21
    Hashimoto K, Kurita H, Haratani T, Fujii K, Ishibashi T. Direct and buffering effects of social support on depressive symptoms of the elderly with home help. Psychiatry Clin. Neurosci. 1999; 53: 95100.
  • 22
    Wade TD, Kendler KS. Absence of interactions between social support and stressful life events in the prediction of major depression and depressive symptomatology in women. Psychol. Med. 2000; 30: 965974.
  • 23
    Kessler RC, Kendler KS, Heath A, Neal MC, Eaves LJ. Social support, depressed mood, and adjustment to stress: a genetic epidemiologic investigation. J. Pers. Soc. Psychol. 1992; 62: 257272.
  • 24
    Wight RG, Aneshensel CS, LeBlanc AJ. Stress buffering effects of family support in AIDS caregiving. AIDS Care 2003; 15: 595613.
  • 25
    Ritter C, Hobfoll SE, Lavin J, Cameron RP, Hulsizer MR. Stress, psychosocial resources, and depressive symptomatology during pregnancy in low-income, inner-city women. Health Psychol. 2000; 19: 576585.
  • 26
    Collins NL, Dunkel-Schetter C, Lobel M, Scrimshaw SC. Social support in pregnancy: psychosocial correlates of birth outcomes and postpartum depression. J. Pers. Soc. Psychol. 1993; 65: 12431258.
  • 27
    Ystgaard M, Tambs K, Dalgard OS. Life stress, social support and psychological distress in late adolescence: a longitudinal study. Soc. Psychiatry Psychiatr. Epidemiol. 1999; 34: 1219.
  • 28
    Roos PE, Cohen LH. Sex roles and social support as moderators of life stress adjustment. J. Pers. Soc. Psychol. 1987; 52: 576585.
  • 29
    Landerman R, George LK, Campbell RT, Blazer DG. Alternative models of the stress buffering hypothesis. Am. J. Community Psychol. 1989; 17: 625642.
  • 30
    Antonucci TC, Israel BA. Veridicality of social support: a comparison of principal and network members’ responses. J. Consult. Clin. Psychol. 1986; 54: 432437.
  • 31
    Krause N. Chronic financial strain, social support, and depressive symptoms among older adults. Psychol. Aging 1987; 2: 185192.
  • 32
    Ingersoll-Dayton B, Antonucci TC. Reciprocal and nonreciprocal social support: contrasting sides of intimate relationships. J. Gerontol. 1988; 43: 6573.
  • 33
    Lu L. Social support, reciprocity, and well-being. J. Soc. Psychol. 1997; 137: 618628.
  • 34
    Yumi N, Maki A. A study on the elderly in mutual-help system: social support and its Relation to well-being. Soc. Econ. Syst. Stud. 2000; 19: 100107.
  • 35
    Schieman S, Meersman SC. Neighborhood problems and health among older adults: received and donated social support and the sense of mastery as effect modifiers. J. Gerontol. B. Psychol. Sci. Soc. Sci. 2004; 59: 8997.
  • 36
    Jou YH, Fukuda H. Stress, health, and reciprocity and sufficiency of social support: the case of University students in Japan. J. Soc. Psychol. 2002; 142: 353370.
  • 37
    Yamashita S, Takizawa T, Sakamoto S et al. Suicide in Japan: present condition and prevention measures. Crisis 2005; 26: 1219.
  • 38
    Ono Y, Tanaka E, Oyama H et al. Epidemiology of suicidal ideation and help-seeking behaviors among the elderly in Japan. Psychiatry Clin. Neurosci. 2001; 55: 605610.
  • 39
    Sakamoto S, Tanaka E, Neichi K, Ono Y. Where is help sought for depression or suicidal ideation in an elderly population living in a rural area of Japan? Psychiatry Clin. Neurosci. 2004; 58: 522530.
  • 40
    Oyama H, Watanabe N, Ono Y et al. Community-based suicide prevention through group activity for the elderly successfully reduced the high suicide rate for females. Psychiatry Clin. Neurosci. 2005; 59: 337344.
  • 41
    Radloff LS. The CES-D scale: a self-report depression scale for research in the general population. Appl. Psychol. Meas. 1977; 1: 385401.
  • 42
    Sakihara S, Harada S. Revision and prediction validity of the measurement of social support-elderly (MOSS-E). In : SakiharaS (ed.). Longitudinal-Study on the Social Environment and a Long Life in Okinawa. University of the Ryukyus, Okinawa, 2000; 812 (in Japanese).
  • 43
    Shima S, Shikano T, Kitamura T, Asai M. New self-rating scales for depression. Clin. Psychiatry 1985; 27: 717723 (in Japanese).
  • 44
    Harada S, Shu-Chuan T, Sakihara S, Takakura M. The relationship between social support with depressive symptoms and life satisfaction among the elderly in rural community. Ryukyu Med. J. 2001; 20: 6166.
  • 45
    Watanabe N, Tujiura T, Takizawa T. Report on the Mental Health of Rokunohe Town. Aomori Prefectural Mental Health and Welfare Center, Aomori, 2004 (in Japanese).
  • 46
    Wethington E, Kessler RC. Perceived support, received support, and adjustment to stressful life events. J. Health Soc. Behav. 1986; 27: 7889.
  • 47
    Wethington E, McLeod JD, Kessler RC. The importance of life events for explaining sex differences in psychological distress. In : BarnettRC, BienerL, BaruchGK (eds). Gender and Stress. Free Press, Collier Macmillan Publishers, New York, 1987; 144156.
  • 48
    Deborah B. Gender difference in the social moderator of stress. In : BarnettRC, BienerL, BaruchGK (eds). Gender and Stress. Free Press, Collier Macmillan Publishers, New York, 1987; 257277.
  • 49
    Deborah B. Social ties and social support. In : DeborahB (ed.). Lives in Stress: Women and Depression. Sage Publications, California, 1982; 133144.
  • 50
    Fujino Y, Mizoue T, Tokui N, Yoshimura T. Prospective cohort study of stress, life satisfaction, self-rated health, insomnia, and suicide death in Japan. Suicide Life Threat. Behav. 2005; 35: 227237.