Father and son attachment styles in alcoholic and non-alcoholic families

Mythili Hazarika1, Dipesh Bhagabati2

1Senior Lecturer, Clinical Psychology, Department of Psychiatry, Gauhati Medical College Hospital, Guwahati, Assam, India, 2Professor and Former Head, Department of Psychiatry, Gauhati Medical College Hospital, Guwahati, Assam, India


Background: The theory of attachment is important to understand a lot of human behaviour. Styles of attachment could be important predictors in developing dependence on alcoholism. Insecure attachment patterns could be significant risk factors for future alcohol use. Methods: Participants for this study consist of fathers with alcohol dependence syndrome (ADS) from treatment centres and fathers from the community with no dependency on alcohol, and their sons (n=200). The Michigan Alcoholism Screening Test (MAST), socioeconomic status scale were administered, and attachment styles were derived by the Attachment Style Questionnaire (ASQ). We hypothesised a prior concept reflecting theoretical predictions for the association between attachment styles and alcohol in both the generations. Results: Statistics on SPSS-16 was used to test our hypotheses. As predicted, fathers with ADS had insecure attachments styles in comparison to the control group. Substance abuse/dependence and treatment participation were at an all-time low for the secure group. Conclusion: The findings from this study identify attachment styles as an influential factor in understanding the divergence between alcohol dependence in treatment seekers. The findings further imply that differential treatment may need to be provided taking into account one’s attachment representation to promote successful recovery. It also highlights the need to develop secure ties in children of alcoholic parents to protect them from use of substances as a coping and a learned mechanism. Limitations of the study and suggestions for further research are highlighted and implications for diagnosis and treatment are discussed.

Keywords: Recovery. Coping. Learning. Diagnosis. Treatment.

Correspondence: Dr. Mythili Hazarika, MA, MPhil, PhD, Senior Lecturer, Clinical Psychology, Department of Psychiatry, Gauhati Medical College Hospital, Bhangagarh, Guwahati-781032, Assam, India. Hazarika.mythili@gmail.com

Received: 9 November 2015

Revised: 9 May 2017

Accepted: 12 May 2017

Epub: 8 July 2017

DOI: 10.5958/2394-2061.2018.00003.4


Alcohol use and abuse are health hazards, and the problems of addiction are not limited to individuals in certain social strata but appear to affect people in all levels of the society.[1] The International Classification of Diseases (ICD-10) defines “alcohol abuse as repeated use despite recurrent adverse consequences; further defining alcohol dependence as alcohol abuse combined with tolerance, withdrawal, and an uncontrollable drive to drink”.[2] Findings of the first systematic effort in India to document the nationwide prevalence of drug use suggest alcohol (21.4%) as the primary substance used and 17% to 26% of alcohol users qualified for ICD-10 diagnosis of dependence, translating to an average prevalence of about four per cent.[3] The intake of alcohol in Assam (37% in rural areas and a prevalence rate of 365/1000 population)[4] is higher  than the national prevalence, and the concerns about use and abuse of alcohol in children as well as adolescent population is a serious public health issue.[5]

Children of Alcoholics (CoA) is a general term for children with one or more alcoholic parents. The consequences of living in a family with alcoholic parents are variable but, nearly all children from alcoholic families are at risk for developing behavioural and emotional difficulties. They may live with psychophysical wound as a result of parental alcoholism.[6] Attachment styles and bond with parents have been documented as major influential factors in alcohol studies. Flores[7] reported that “addiction is an attachment disorder because for better or worse an increasing number of individuals in the present day society rely on psychoactive substances to help manage their fear and difficulties in interpersonal relationships”.

Attachment research has repeatedly shown the significant contribution of early social interactions on developing inner working models and object representations that substantially impact on bonding behaviour, subject-environment interactions, and psychopathology in later life.[8] On the other hand, the genetic and biological aspects also determine one’s pattern of attachment to a higher extent than previously thought,[9] and reports have identified insecure patterns of attachment to be much higher in alcohol addiction.[10]

But, Indian literature from this perspective is scanty. We felt the increased need for comprehensive investigation, particularly with adolescent and adult CoA as there is a dearth of exploration carried out on this issue in the Indian sociocultural context. Although statistics suggest that alcohol use and dependence is a major problem in India. Earlier research has indicated that for alcoholism or psychosocial maladaptation to develop in CoA, parent-child bond and their interaction, and typical patterns of attachment are contributory factors.

Keeping it in mind and to explore this area in alcoholism research, the focus of the current study was on styles of attachment of fathers with alcohol dependent symptoms and their sons with a comparative group of fathers with non-alcohol dependent symptoms and their sons.

Materials and methods

The Medical Institutional Review Board at the Gauhati University approved the procedures of this study. The total number of participants (n=200) provided written consent for their involvement in the study. It is a case-control study with four groups of 50 subjects in each group comprising of alcohol dependent fathers (ADF), non-alcohol dependent fathers (NADF), sons of ADF (SADF), and sons of NADF (SNADF). The ADF were from treatment centres located in Guwahati, Assam and the NADF were from the community by the snowball technique who met the inclusion and exclusion criteria. Inclusion criteria for ADF were middle age (40-60 years), middle socioeconomic status (SES), nontribal, high school educated, and with a score of ≥13 in the Michigan Alcoholism Screening Test (MAST).[11] Sons had to be minimum high school passed and 15+ years with the mentioned inclusion criteria. Exclusion criteria was no history of chronic physical illness or psychotic disorders, severe depressive disorder and affective disorder, history of treatment for any psychiatric disorder in the past, epilepsy or past history of generalised tonic-clonic seizures (excluding withdrawal seizures), history of organic brain disorder or dysfunctions or mental retardation, past and present history of other drug use, and multiple substance abuse/dependence. The community group comprised of individuals of the same criteria except that the MAST score of <13. The participants included in the study completed a questionnaire packet including inventories for demography, MAST, Kuppuswamy SES,[12] checklist of various common ailments, and Attachment Style Questionnaire (ASQ)[13] having high reliability and validity for the population under study.

Preliminary analyses and research strategy

The main objective was to examine the attachment styles of ADF and SADF, and NADF and SNADF. We hypothesised that ADF would have insecure attachment styles but their sons would have secure attachment styles in comparison to NADF and SNADF. Family studies have revealed a number of differences and dysfunctions between alcoholic and non-alcoholic family environments. Though almost all reviews support the preposition that parental alcoholism is associated with behavioural, emotional, and psychosomatic complaints in their offspring, there are some studies that have suggested that a good relationship with a non-alcoholic parent may serve as a protective mechanism. Keeping this in mind, the hypothesis was that SADF and SNADF would both have secure attachment styles. The ADF continued with the standardised inpatient treatment programmes. At the end of the second week of abstinence, they were screened based on the criteria for inclusion into the study. The group from the treatment centres continued their standard treatment regimen and necessary precautions were taken so that participation in the study did not affect their treatment process. Participation was purely on voluntary basis and they were informed that there would be no direct or indirect benefits for participating in the study. Confidentiality and anonymity was assured.

Data obtained was analysed using descriptive statistics such as mean, standard deviations, and frequencies. Where appropriate, inferential statistics such as student’s t-test, chi-square, and analysis of variance (ANOVA) were also utilised. Attachment styles among ADF and NADF constituted the major variable of interest in this investigation. ANOVA (treatment seekers and community controls) and post-hoc tests were performed across six domains, including demographic/affective measures, SES, physical parametres, alcohol dependence, and attachment styles to look for significant associations between fathers of both the groups and sons of both the groups. Bonferroni corrections were applied within each domain.


The total sample was 200 (ADF=50, NADF=50, SADF=50, and SNADF=50). The sociodemographic profile did not differ statistically between ADF and NADF. Forty three ADF studied till graduation and 42 from NADF were graduates, seven from ADF and eight from NADF studied till post-graduation with no significant difference (chi-square=0.078 and p-value >0.779). There was no significant difference between the groups on the variables of occupation, family type, number of children, and SES, and the p-value was more than 0.05 level. Similarly, between SADF and SNADF the demographic variables matched well. The age ranges were from 15-20 and 21-25 with no significant difference (chi-square=4.059; p-value=0.255) between the groups. Fifty two per cent of both the groups’ students studied till graduation. Statistics on SPSS-16 was used to test the hypotheses. Descriptive data and comparison of attachment styles’ variables among alcohol dependent fathers (ADF), non-alcohol dependent fathers (NADF), sons of alcohol dependent fathers (SADF), and sons of non-alcohol dependent fathers (SNADF) are shown in Tables 1/1(A)-3/3(A), and Figures 1 and 2. They indicate the secure and insecure attachment styles used by ADF and NADF, and their sons. In ANOVA, the calculated value of F (i.e. 88.132) was greater than the tabulated value of F (i.e. 2.63) at (3,196) df for five level of significance between the mean effect of secure attachment styles, calculated value of F (i.e. 31.056) was greater than tabulated value of F (i.e. 2.63) at (3,196) df for five level of significance between the mean effect of avoidant attachment style, and the calculated value of F (i.e. 17.061) was greater than tabulated value of F (i.e. 2.63) at (3,196) df for five level of significance between the mean effect of anxious-ambivalent attachment style between and within ADF, NADF, SADF, and SNADF. Thus, there was significant difference in attachment styles between the groups. Turkey’s post-hoc analyses indicated significant difference between and within the groups in secure, avoidant, and anxious-ambivalent attachment styles.


The data was generated from four groups of participants (ADF, NADF, SADF, and SNADF; total n=200). The groups matched well in demographics. Middle SES and only non-tribal were included to reduce the sampling bias. SADF used more alcohol though not statistically significant and earlier studies have reported the co-occurrence of father’s alcoholism and son’s alcohol dependence.[14] The current study proposed to take fathers and sons to evaluate their associations as it is reported to be greater, and mothers with alcohol dependence could not be a focus due to lack of available female population in hospital setting and underreporting from the population in study.

The current findings suggest that fathers in both the groups differ significantly in attachment styles with ADF using more insecure (avoidant and anxious-ambivalent) and NADF using more secure attachment styles which was hypothesised in the current study. This was consistent with earlier findings in which people with alcoholism reported insecure attachment, characterised by fearful-avoidant and dismissed-avoidant styles.[15] Some authors have also reported that individuals with a secure attachment style seek social support to cope with emotional stress, whereas individuals with an insecure attachment style tend to seek other means, such as use of alcohol or illicit drugs as a coping mechanism for emotional self-regulation,[16] and reported fear of intimacy, lower secure attachment, and differentiation of self.[17] Kassel and his colleagues[18] in their research have found that both drug and alcohol is related to anxious-ambivalent attachment. Earlier researchers theorised that anxiously attached individuals drink alcohol to decrease their negative emotions, especially feelings of abandonment by others which is popularly known as the self-medication theory.[19] In individuals having alcohol use, different styles of attachment could exert control on one’s emotions and relationships with others.[20]

Indian studies with SADF are scarce and in India, the family interaction patterns and ties are strong. Thus, affect expression are significantly seen among family members compared to the western world. Hence, mother plays a significant role in bonding. So, the hypothesis framed was that there would be no difference in attachment styles between SADF and SNADF. But, the findings indicate that when alcoholism was present, the children in a very consistent manner failed to develop emotional ties with either parent. Hence, the insecure styles of attachment existed. Thus, supporting the research about the absence of “compensatory relationship” with both parents when normal family interaction was disrupted by inappropriate parental behaviours. Some earlier studies were consistent with the present study[21] that children of fathers with alcohol-related problems had higher risk of attachment insecurity and more parental rejection. Few authors have theorised that parents help children to modulate emotional states and reduce internal tension,[22] and when confronted by insensitive and damaging care, children will generally manage homeostasis and regulation on their own, rather than with the extra emotional support that parents can provide. The current study suggests that more disruptive the parental alcoholic behaviour during the childhood, the less likely the children would identify with the secure patterns of attachment with the alcoholic parent or either parent. Hence, the type of attachment and bonding most characteristic of adult CoA was either insecure/anxious-avoidant or insecure/ambivalent.

Few limitations are self-report measures; so, high susceptibility to social desirability biases. The sampling was a purposive sample (non-tribal, middle SES, and male gender); therefore, the generalisation of the findings is limited to those who share the sample characteristics. Sampling bias makes it difficult to generalise the findings to alcoholic population at large. The alcoholic group was from the clinical setting which means they are very severe and that offspring’s status also is likely to be very severe. Most individuals with alcohol use disorders do not seek treatment and they are considered different from those currently on treatment. The study is cross-sectional in nature, due to which the cause-effect relationship between attachment styles and alcoholism cannot be commented upon. Hence, a longitudinal design will be useful to draw inferences on the various variables contributing to alcoholism.


The findings from this study identify attachment styles as an influential factor in understanding the divergence between alcohol dependence in treatment seekers and non-dependence in the community. The findings further imply that differential treatment may need to be provided taking into account one’s attachment representation to promote successful recovery. It also highlights the need to develop secure ties in children of alcoholic parents to protect them from use of substances as a coping and a learned mechanism.


Dr. Linda Cottler and the FOGARTY team, INDO-US program in chronic non-communicable diseases (CNCDs) #1D43TW009120 (M Hazarika, Fellow)


1. Murthy P, Manjunatha N, Subodh BN, Chand PK, Benegal V. Substance use and addiction research in India. Indian J Psychiatry. 2010;52(Suppl 1):S189-99.

2. World Health Organization. The ICD-10 classification of mental and behavioural disorders: clinical descriptions and diagnostic guidelines. 10th rev. Geneva: World Health Organization; 1992.

3. Ray R, editor. The extent, pattern and trends of drug abuse in India: national survey. Ministry of Social Justice and Empowerment, Government of India & United Nations Office on Drugs and Crime, Regional Office for South Asia; 2004.

4. Hazarika NC, Biswas D, Phukan RK, Hazarika D, Mahanta J. Prevalence and pattern of substance abuse at Bandardewa, a border area of Assam and Arunachal Pradesh. Indian J Psychiatry. 2000;42:262-6.

5. Bhagabati D, Das B, Das S. Pattern of alcohol consumption in underage population in an Indian city. Dysphrenia. 2013;4:36-41.

6. Seixas J. Children from alcoholic homes. In: Estes N, Heinemann E, editors. Alcoholism: development, consequences and interventions. St. Louis, MO: Mosby; 1982:193-201.

7. Flores PJ. Addiction as an attachment disorder: implications for group therapy. Int J Group Psychother. 2001;51:63-81.

8. Ainsworth MD. Attachments across the life span. Bull N Y Acad Med. 1985;61:792-812.

9. Barry RA, Kochanska G, Philibert RA. G x E interaction in the organization of attachment: mothers’ responsiveness as a moderator of children’s genotypes. J Child Psychol Psychiatry. 2008;49:1313-20.

10. De Rick A, Vanheule S. Attachment styles in alcoholic inpatients. Eur Addict Res. 2007;13:101-8.

11. Selzer ML. The Michigan Alcoholism Screening Test: the quest for a new diagnostic instrument. Am J Psychiatry. 1971;127:1653-8.

12. Kuppuswamy B. Manual of socioeconomic status (Urban). Delhi: Manasayan; 1981.

13. Feeney JA, Noller P, Hanrahan M. Assessing adult attachment. In: Sperling MB, Berman WH, editors. Attachment in adults: clinical and developmental perspectives. New York: Guilford Press; 1994:128-52.

14. Gabrielli WF Jr, Mednick SA. Intellectual performance in children of alcoholics. J Nerv Ment Dis. 1983;171:444-7.

15. Hankin BL, Kassel JD, Abela JR. Adult attachment dimensions and specificity of emotional distress symptoms: prospective investigations of cognitive risk and interpersonal stress generation as mediating mechanisms. Pers Soc Psychol Bull. 2005;31:136-51.

16. Brennan KA, Shaver PR, Tobey AE. Attachment styles, gender, and parental problem drinking. J Soc Pers Relat. 1991;8:451-66.

17. Thorberg FA, Lyvers M. Attachment, fear of intimacy and differentiation of self among clients in substance disorder treatment facilities. Addict Behav. 2006;31:732-7.

18. Kassel JD, Wardle M, Roberts JE. Adult attachment security and college student substance use. Addict Behav. 2007;32:1164-76.

19. Allan CA. Alcohol problems and anxiety disorders--a critical review. Alcohol Alcohol. 1995;30:145-51.

20. Wedekind D, Bandelow B, Heitmann S, Havemann-Reinecke U, Engel KR, Huether G. Attachment style, anxiety coping, and personality-styles in withdrawn alcohol addicted inpatients. Subst Abuse Treat Prev Policy. 2013;8:1.

21. Eiden RD, Edwards EP, Leonard KE. Mother-infant and father-infant attachment among alcoholic families. Dev Psychopathol. 2002;14:253-78.

22. Cicchetti D, Rogosch FA, Lynch M, Holt K. Resilience in maltreated children: processes leading to adaptive outcome. Dev Psychopathol. 1993;5:629-47.

Hazarika M, Bhagabati D. Father and son attachment styles in alcoholic and non-alcoholic families. Open J Psychiatry Allied Sci. 2017;9:15-9. doi: 10.5958/2394-2061.2018.00003.4. Epub 2017 Jul 8.

Source of support: Nil. Declaration of interest: None.

Creative Commons License
This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.

Nach oben