European Neuropsychopharmacology (2014) 24, 1793–1797

http://dx.doi.org/10924-977X/& 2014 P

nCorresponding afax: +972 3 6358599

E-mail address: m

www.elsevier.com/locate/euroneuro

Borderline intellectual functioningis associated with poor social functioning,increased rates of psychiatric diagnosisand drug use – A cross sectional populationbased study

Karny Gigib, Nomi Werbeloffa, Shira Goldberga,Shirly Portuguesec, Abraham Reichenbergd, Eyal Fruchterc,Mark Weisera,b,n

aDepartment of Psychiatry, Sheba Medical Center, Tel Hashomer 52621, IsraelbSackler School of Medicine, Tel Aviv University, Tel Aviv, IsraelcDivision of Mental Health, Medical Corps, IDF, IsraeldDepartment of Psychological Medicine, Institute of Psychiatry, London, England, United Kingdom

Received 18 February 2014; received in revised form 27 July 2014; accepted 30 July 2014

KEYWORDSBorderline intellec-tual functioning;Social functioning;Drug use;Psychiatric diagnosis

0.1016/j.euroneurublished by Elsevi

uthor at: Departm.weiser@netvision

AbstractBorderline intellectual functioning is defined by the DSM IV as an IQ range that is between oneto two standard deviations below the mean (71oIQo84), and a considerable percentage of thepopulation is included in this definition (approximately 13.5%). The few studies performed onthis group indicate that borderline intellectual functioning is associated with various mentaldisorders, problems in everyday functioning, social disability and poor academic or occupa-tional achievement. Using data from the Israeli military, we retrieved the social and clinicalcharacteristics of 76,962 adolescents with borderline intellectual functioning and comparedtheir social functioning, psychiatric diagnoses and drug abuse with those of 96,580 adolescentswith average IQ (70.25 SD from population mean). The results demonstrated that theborderline intellectual functioning group had higher rates of poor social functioning comparedto the control group (OR=1.9, 95% CI=1.85–1.94). Individuals with borderline intellectualfunctioning were 2.37 times more likely to have a psychiatric diagnosis (95% CI=2.30–2.45) and1.2 times more likely to use drugs (95% CI=1.07–0.35) than those with average IQ. These resultssuggest that adolescents with borderline intellectual functioning are more likely to suffer frompsychiatric disorders, poor social functioning and drug abuse than those with average

o.2014.07.016er B.V.

ent of Psychiatry, Sheba Medical Center, Tel Hashomer 52621, Israel. Tel.: +972 52 666 6575;

.net.il (M. Weiser).

K. Gigi et al.1794

intelligence, and that borderline intellectual functioning is a marker of vulnerability to thesepoor outcomes.& 2014 Published by Elsevier B.V.

1. Introduction

Borderline Intellectual Functioning is defined by the DSM IV asan IQ range that is higher than that of “mental retardation” –

between one to two standard deviations below the mean(71oIQo84), encompassing 13.5% of the population. Fewinvestigators have studied the impact of borderline intellectualfunctioning on psychiatric and social outcomes, partially due tothe fear that such studies might have the unintended conse-quence of stigmatizing such individuals. Further, there are fewpopulation based databases that include IQ and such outcomes,and no longitudinal data regarding this vulnerability. None-theless, existing data support the association between border-line intellectual functioning and poor psychosocial outcome.

Hassiotis et al. (2008) examined data from the UK Wide CrossSectional Survey of 8450 adults living in private households.They found that 12.3% of the sample had borderline intellectualfunctioning, and compared to their peers with average intelli-gence, this group had increased rates of neurotic disorders,depressive episodes, phobias, substance misuse and personalitydisorders, but not psychotic disorders. Further, this group wasmore likely to receive psychiatric medications and to utilizemore community and daycare services. Seltzer et al. (2005)compared individuals with IQ scores of 85 or below with theirsiblings who obtained IQ scores above 100. They found that low-IQ individuals completed less schooling, had less prestigiousoccupations, rated themselves less physically healthy, andreported lower levels of psychological well-being.

Some studies indicate that low IQ is associated with riskyhealth behaviors, such as alcohol abuse, and cigarette smoking(Kubicka et al., 2001; Chandola et al., 2006; Weiser et al.,2010). Other studies have demonstrated an association betweenlow intelligence and delinquency in adolescence (White et al.,1989) and criminality or violence in adulthood (Huesmannet al., 2002). Finally, in a recent study, Hassiotis et al. (2011)demonstrated that participants with borderline intellectualfunctioning were more likely to report suicide attempts orself-harm compared to those with average intellectual func-tioning. However these associations were no longer significantafter controlling for income and age.

In the current study we used a population-based datasetcollected by the Israeli Draft Board of 16–17 year old men toidentify a cohort of 76,962 with borderline intellectual func-tioning and 96,580 with average IQs. The analyses examinedclinical and social characteristics (social functioning, psychiatricdiagnosis and drug use) of subjects with borderline intellectualfunctioning compared to subjects with average IQ.

2. Experimental procedures

2.1. Draft board assessment

Israeli law requires that all adolescents between the ages of 16 and 17undergo a pre-induction assessment to determine their intellectual,

medical, and psychiatric eligibility for military service. This assessmentis compulsory and is administered to the entire unselected population ofIsraeli male adolescents. It includes individuals who are eligible formilitary service, as well as those who will ultimately be excluded fromservice for medical, psychiatric, or social reasons.

2.2. Cognitive assessment

The cognitive test battery yields a total score which is a highly validmeasure of general intelligence, equivalent to a normally distrib-uted IQ score. Tests are administered by a trained psychometrician.The cognitive assessment comprises four sub-tests: (a) Arithmetic-R, which assesses cognitive reasoning, concentration, and conceptmanipulation. This sub-test is similar to the ‘arithmetic’ sub-testfrom the Wechsler Intelligence Scales. The test is in a multiple-choice format and contains twice as many items, and thereforeincludes harder test items; (b) Verbal analogies which assess verbalabstraction and categorization (i.e. the ability to understand therelationship between words and the use of this relationship inseveral contexts). This is a multiple-choice test. This test is similarto the ‘similarities’ sub-test from the Wechsler Intelligence Scale.Unlike the Wechsler test, this sub-test is a multiple choice test andsubjects are requested not only to identify and report the semanticof causal relationships between the test items, but also to applythese relations to target items; (c) A non-verbal spatial analogiestest which measures non-verbal abstract reasoning and problem-solving abilities. This test is also a multiple-choice test; (d) OTIS-R,a modified, Otis-type verbal intelligence test adapted from the USArmy Alpha Instructions Test, which measures the ability to under-stand and carry out verbal instructions (Lezak, 1995). Tests areprogressive, beginning with relatively simple items and becomingmore difficult. Tests are group-administered and are time-limited.All scores are based on the number of correct answers. In manyvalidation studies conducted by the Draft Board, the summary scoreof the cognitive test battery has been found to be a highly validmeasure of general intelligence (Gal, 1986)

2.3. Draft board psychiatric assessment

After the cognitive assessments are performed, a semi-structured,thirty minute, interview is held. The purpose of the interview is toassess personality and behavioral traits that will lead to anestimation of the potential conscript's suitability for militaryservice, particularly service in combat units (Gal, 1986). Theinterview is administered by trained enlisted individuals (most ofthem female soldiers) who participated in a 3 month trainingcourse. The interviewers are under regular supervision by seniorinterviewers and participate in ongoing training. The behavioralassessment, administered only to males, includes a subscale asses-sing current social functioning. Based on structured questions,social functioning is then scored on a scale of 1–5: (1) Very poor:complete withdrawal, (2) Poor: weak interpersonal contacts,(3) Adequate: can form relationships with individuals and in agroup, (4) Good: good interpersonal relationships and (5) Excep-tional: superior interpersonal relatedness. The test–retest reliabil-ity of the behavioral assessment for inductees interviewed afterseveral days by different interviewers is above 0.8, and population-based norms are available (Reeb, 1968; Gal, 1986). The draft boardscreening is described in detail in other reference (Gal, 1986).

1795Borderline intellectual functioning

After this initial screening interview, potential conscripts whoare deemed to have significant behavioral problems (approximately15–20% of those screened) are referred for an in-depth psychosocialassessment. The in-depth psychosocial assessment is performed bya postgraduate (MA) level clinical social worker or psychologist.Drug abuse is systematically asked about in the initial screeninginterview only for those male adolescents who are suspected ofhaving significant behavioral problems, before being referred forthe in-depth psychosocial assessment.

The criteria for referral for the in-depth psychosocial assessmentinclude one or more of the following: (a) obtaining the lowest scoreon the rating of social functioning, (b) documentation or self-reportof present or past psychiatric symptoms, including enuresis, sleepdisturbances, drug or alcohol abuse; (c) clinical judgment of theinterviewer that the adolescent will not adapt well to the demandsof the military.

If the interviewer suspects that the adolescent has a psychiatricdisorder, he is referred to a board certified psychiatrist forevaluation and an ICD-9 diagnosis. Diagnoses during the timecovered by this study were based on ICD-9 criteria, and weredivided into non-affective psychotic disorders, and non-psychoticdiagnoses grouped for purposes of the study to: neurosis, minoraffective disorders, anxiety disorders and PTSD; Adjustment dis-order; Personality disorders; Anti-social personality disorder, sub-stance abuse; Schizophrenia spectrum; Major affective disorders.

For more detailed description of the Draft Board assessmentprocedure see (Gal, 1986).

2.4. Analytic sample

654,173 Male adolescents were consecutively assessed by the DraftBoard. Adolescents with missing IQ score (n=669, 0.1%) or withmissing SES (n=154,352, 23.5%) were excluded, leaving 499,766adolescents.

Borderline intellectual functioning was defined as an IQ in therange of 71–84; 76,962 (15.3%) adolescents were identified ashaving borderline intellectual functioning.

In order to compare the borderline intellectual functioning group toa group of subjects with the mean IQ in the population, we selectedthose with an IQ between 96 and 104 (70.25 SD) controls. This IQrange was selected in order to ensure that the lower boundary of theaverage IQ group would be significantly higher than the best function-ing individuals of the borderline intellectual functioning sample so thatthe average IQ group will represent individuals around the meanpopulation IQ (19.7% according to the normal distribution probability).Thus, 96,580 (19.3%) adolescents were included in the control group.

Table 1 The prevalence of psychiatric disorders among the bo

Variable Borderline intellectualfunctioning group (n, %)

No psychiatric diagnosis 65,441, 85.7%Schizophrenia spectrum 175, 0.2%Anti-social PD, 367, 0.5%Substance abuse 86, 0.1%Major affective disorder 210, 0.3%Neurosis, minor affective

and anxiety, PTSD1417, 1.9%

Personality disorders 7815, 10.2%Non-affective psychosis 621, 0.8%Adjustment disorder 224, 0.3%Total 76,356, 100%

2.5. Data analysis

Logistic regression was used to calculate the association between IQgroup (borderline intellectual functioning vs. average IQ) and socialfunctioning, yielding odds ratios and 95% confidence intervals. Associoeconomic status (SES) is strongly associated with cognitiveability and IQ (Noble et al., 2005), all analyses controlled for SES.

Multinomial regression was used to calculate the associationbetween IQ group and psychiatric diagnosis.

As psychiatric illness is associated with low IQ, poor socialfunctioning and drug abuse (Weiser et al., 2003, 2004, 2007;David et al., 2008), we also stratified the analyses according tothe presence of psychiatric illness at the draft board assessment.For the purpose of this analysis, social functioning was treated as adichotomous variable, using the median as a cut-off point forcoding, thus comparing those with very poor or poor socialfunctioning with those who had adequate, good or superior socialfunctioning.

Analyses were performed with SPSS 18.

3. Results

The mean IQ score of the borderline intellectual functioninggroup was 79.5 (SD=4.18) and of the control group 97.89(SD=0.96).

3.1. Social impairment

Logistic regression analyses controlling for SES revealed thatthe borderline intellectual functioning group had higherrates of poor social functioning compared to the controlgroup (31% in the borderline intellectual functioning groupas compared to 18% in the control group; OR=1.9, 95%CI=1.85–1.94). When removing individuals with psychiatricillness from the analyses, the association between border-line intellectual functioning and poor social functioning wasmostly unchanged (OR=1.78, 95% CI=1.73–1.83).

3.2. Psychiatric diagnosis

Logistic regression analyses controlling for SES revealed thatindividuals with borderline intellectual functioning were2.37 times more likely to have any psychiatric diagnosisthan those with average IQ (95% CI=2.30–2.45). The results

rderline intellectual functioning group and control group.

Average IQ group (n, %) Odds ratio (CI)

89,828, 93.2% –

86, 0.1% 2.498 (1.916–3.255)125, 0.1% 3.784 (3.073–4.658)85, 0.1% 1.209 (0.888–1.645)119, 0.1% 2.016 (1.601–2.539)1202, 1.2% 1.663 (1.535–1.802)

4455, 4.6% 2.508 (2.411–2.609)212, 0.2% 3.382 (2.883–3.967)244, 0.3% 1.281 (1.063–1.546)96,356, 100%

K. Gigi et al.1796

varied only minimally between significant diagnostic groups,but nonetheless the highest association was for anti-socialpersonality disorder and non-affective psychotic diagnosis(Table 1).

3.3. Drug use

Only a minority of subjects were screened for drug abuse.The current analyses left us with 12,785 (16.6%) cases withborderline intellectual functioning and 10,528 (10.9%) con-trols with average IQ who were screened for drug abuse.

Logistic regression controlling for SES demonstrated thatindividuals with borderline intellectual functioning weresignificantly more likely to use drugs (OR=1.2, 95%CI=1.07–1.35).

4. Discussion

In this study a unique, population-based database was used,which included a detailed assessment of all Israeli 16–17years old male adolescents, in order to examine therelationship between borderline intellectual functioningand psychosocial variables. We found that compared tothose with average intellectual functioning, young men withborderline intellectual functioning had increased rates ofpoor social functioning, psychiatric diagnoses, and drugabuse with the higher rates of poor social functioningpresent even in those subjects without any psychiatricdiagnosis.

These data are supported by findings of other investiga-tors using different designs and different populations(Rajput et al., 2011). Particularly, the results of Hassiotiset al. (2008) are remarkably similar to the results of thecurrent study, as their findings show that people withborderline intellectual functioning suffer from a socialdisadvantage, increased rates of neurotic disorders, moredepressive episodes, phobias, substance misuse and person-ality disorders, as compared to their peers of normalintelligence.

The finding that individuals with borderline intellectualfunctioning were more than twice as likely to have anypsychiatric diagnosis as those with average IQ is in line withprevious literature which indicates that cognitive impair-ment appears to be associated with the entire spectrum ofpsychiatric disorders (Weiser et al., 2004; David et al.,2008). The finding that individuals with borderline intellec-tual functioning were more likely to have poor socialfunctioning, psychiatric disorders and drug use might indi-cate that low IQ is at least in part a risk factor for generallypoor functional outcome.

According to the Flynn effect (Flynn, 1987) almost allcultures gained 20 points in their IQ scores since 1930 to thepresent day. Flynn (1987) indicated that this IQ gain is dueto an increase in the fluid IQ component of the IQ test, acomponent that measures mainly abstraction abilities andon the spot problem solving. In a world that demands theseabilities in everyday life, people with borderline intellectualfunctioning might experience many difficulties and adjust-ment problems.

Our data combined with previous data by other groupsemphasize the vulnerability of individuals with borderline

intellectual functioning. Taking into account the high pre-valence of this group in the population (approximately13.5%) and their poor outcome, this issue should beconsidered as a public health issue. This attitude isreflected in a recently published paper (Salvador-Carullaet al., 2013) stating that borderline intellectual functioningis a “health meta-condition that requires specific publichealth, education and legal attention”.

Our data indicate that 30% of the borderline intellectualfunctioning subjects from our sample suffered from poorsocial functioning at age 17. This sub-group should beconsidered at-risk, and public health strategies should bedeveloped to mitigate this vulnerability.

Clinicians and psychiatrists might take these factors intoconsideration when treating individuals with borderlineintellectual functioning in a manner similar to the approachto treating individuals with somatic risk factors such asobesity and hypertension which are associated with poormedical outcome (WHO, 2013a, 2013b).

4.1. Limitations

The interpretation of the findings reported here should beviewed in light of two main limitations:

First, the results are not based on the entire populationof adolescents, but only on 17 years old males.

Second, the results regarding the drug use rates arebased on a subsample of male adolescents with poor socialadjustment or other behavioral problems, both of whom aremore prone to drug use (Sussman et al., 2000) and atincreased risk of later schizophrenia diagnosis (Davidsonet al., 1999; Weiser et al., 2001). However this is by far thelargest group of borderline intellectual functioning subjectsthat has been studied.

In summary, borderline intellectual functioning is asso-ciated with poor psycho-social functioning and increasedrisk for psychopathology. Clinicians should be aware of thepsychopathological vulnerability and poor social outcome ofthis group.

Role of funding source

This research received no specific grant from any funding agency,commercial or not-for-profit sectors.

Contributors

Karny Gigi performed the statistical analysis and drafted themanuscript. Nomi Werbeloff and Shira Goldenberg commented onthe statistical analysis and on the manuscript. Shirly Portuguese andEyal Fruchter were responsible for the data collection. AbrahamReichenberg: advised on the statistical analyses. Mark Weiserconceived and designed the study, supervised the statistical ana-lyses and writing.

All authors contributed to and have approved the finalmanuscript.

Conflict of interest

All authors declare that they have no conflicts of interest.

1797Borderline intellectual functioning

Acknowledgment

None.

References

Chandola, T., Deary, I.J., Blane, D., Batty, G.D., 2006. Childhood IQin relation to obesity and weight gain in adult life: the NationalChild Development (1958) Study. Int. J. Obes. 30, 1422–1432.

David, A.S., Zammit, S., Lewis, G., Dalman, C., Allebeck, P., 2008.Impairments in cognition across the spectrum of psychiatricdisorders: evidence from a Swedish conscript cohort. Schizophr.Bull. 34, 1035–1041.

Davidson, M., Reichenberg, A., Rabinowitz, J., Weiser, M., Kaplan, Z.,Mark, M., 1999. Behavioral and intellectual markers for schizo-phrenia in apparently healthy male adolescents. Am. J. Psychiatry156, 1328–1335.

Flynn, J.R., 1987. Massive IQ gains in 14 nations: what IQ test reallymeasure. Psychol. Bull. 101, 171–191.

Gal, R., 1986. The Selection, Classification and Placement Process:A Portrait of the Israeli Soldier. Greenwood Press, Westport,Connecticut.

Hassiotis, A., Strydom, A., Hall, I., Ali, A., Lawrence-Smith, G.,Meltzer, H., Head, J., Bebbington, P., 2008. Psychiatric morbidityand social functioning among adults with borderline intelligenceliving in private households. J. Intellect. Disabil. Res. 52, 95–106.

Hassiotis, A., Tanzarella, M., Bebbington, P., Cooper, C., 2011.Prevalence and predictors of suicidal behaviour in a sample ofadults with estimated borderline intellectual functioning: resultsfrom a population survey. J. Affect. Disord. 129, 380–384.

Huesmann, L.R., Eron, L.D., Dubow, E.F., 2002. Childhood predic-tors of adult criminality: are all risk factors reflected in child-hood aggressiveness? Crim. Behav. Ment. Health 12, 185–208.

Kubicka, L., Matejcek, Z., Dytrych, Z., Roth, Z., 2001. IQ andpersonality traits assessed in childhood as predictors of drinkingand smoking behaviour in middle-aged adults: a 24-year follow-up study. Addiction 96, 1615–1628.

Lezak, M.D., 1995. Neuropsychological Assessment. Oxford Univer-sity Press, New York.

Noble, K.G., Norman, M.F., Farah, M.J., 2005. Neurocognitivecorrelates of socioeconomic status in kindergarten children.Dev. Sci. 8, 74–87.

Rajput, S., Hassiotis, A., Richards, M., Hatch, S.L., Stewart, R.,2011. Associations between IQ and common mental disorders:the 2000 British National Survey of Psychiatric Morbidity. Eur.Psychiatry 26, 390–395.

Reeb, M., 1968. Construction of questionnaire to replace a validstructured interview in the Israeli Defense Forces. MegamotBehav. Sci. Q. 16, 69–74.

Salvador-Carulla, L., Garcia-Gutierrez, J.C., Ruiz Gutierrez-Colosia,M., Artigas-Pallares, J., Garcia Ibanez, J., Gonzalez Perez, J.,Nadal Pla, M., Aguilera Ines, F., Isus, S., Cereza, J.M., Poole, M.,Portero Lazcano, G., Monzon, P., Leiva, M., Parellada, M.,Garcia Nonell, K., Martinez, I.H.A., Rigau, E., Martinez-Leal,R., 2013. Borderline intellectual functioning: consensusand good practice guidelines. Rev. Psiquiatr. Salud Ment. 6,109–120.

Seltzer, M.M., Floyd, F., Greenberg, J., Lounds, J., Lindstromm, M.,Hong, J., 2005. Life course impacts of mild intellectual deficits.Am. J. Ment. Retard. 110, 451–468.

Sussman, S., Dent, C.W., Leu, L., 2000. The one-year prospectiveprediction of substance abuse and dependence among high-riskadolescents. J. Subst. Abuse 12, 373–386.

Weiser, M., Reichenberg, A., Rabinowitz, J., Kaplan, Z., Caspi, A.,Yasvizky, R., Mark, M., Knobler, H.Y., Nahon, D., Davidson, M.,2003. Self-reported drug abuse in male adolescents with beha-vioral disturbances, and follow-up for future schizophrenia. Biol.Psychiatry 54, 655–660.

Weiser, M., Reichenberg, A., Rabinowitz, J., Kaplan, Z., Mark, M.,Bodner, E., Nahon, D., Davidson, M., 2001. Association betweennonpsychotic psychiatric diagnoses in adolescent males andsubsequent onset of schizophrenia. Arch. Gen. Psychiatry 58,959–964.

Weiser, M., Reichenberg, A., Rabinowitz, J., Knobler, H.Y., Lubin,G., Yazvitzky, R., Nahon, D., Gur, R.C., Davidson, M., 2004.Cognitive performance of male adolescents is lower than con-trols across psychiatric disorders: a population-based study.Acta Psychiatr. Scand. 110, 471–475.

Weiser, M., van Os, J., Reichenberg, A., Rabinowitz, J., Nahon, D.,Kravitz, E., Lubin, G., Shmushkevitz, M., Knobler, H.Y., Noy, S.,Davidson, M., 2007. Social and cognitive functioning, urbanicityand risk for schizophrenia. Br. J. Psychiatry 191, 320–324.

Weiser, M., Zarka, S., Werbeloff, N., Kravitz, E., Lubin, G., 2010.Cognitive test scores in male adolescent cigarette smokerscompared to non-smokers: a population-based study. Addiction105, 358–363.

White, J.L., Moffitt, T.E., Silva, P.A., 1989. A prospective replica-tion of the protective effects of IQ in subjects at high risk forjuvenile delinquency. J. Consult. Clin. Psychol. 57, 719–724.

World Health Organization, 2013a. Obesity. ⟨http://www.who.int/topics/obesity/en/⟩.

World Health Organization, 2013b. Cardiovascular diseases (CVDs).⟨http://www.who.int/mediacentre/factsheets/fs317/en/⟩.