The literature related to people with borderline intellectual functioning (BIF) was systematically reviewed in order to summarize the present knowledge. Database searches yielded 1,726 citations, and 49 studies were included in the review. People with BIF face a variety of hardships in life, including neurocognitive, social, and mental health problems. When adults with BIF were compared with the general population, they held lower-skilled jobs and earned less money. Although some risk factors (e.g., low birth weight) and preventive factors (e.g., education) were reported, they were not specific to BIF. The review finds that, despite the obvious everyday problems, BIF is almost invisible in the field of research. More research, societal discussion, and flexible support systems are needed.
People with borderline intellectual functioning (BIF) have an IQ test score that is one to two standard deviations below average, in the range of 70 to 85. If normal distribution of intelligence is considered, 13.6% of the population fits into that category. Not all of the people who score in this IQ range have problems with adaptive behavior (conceptual, social, and practical skills), nor do they all need support, but this figure can be used as a guide. Despite the high percentage of people in this category, BIF is a rarely studied topic. When it is included in studies, the focus of the research is often on people with mild intellectual disability (MID) or a specific learning disability (SLD), and the BIF group is either combined with these or treated as a control group. There seem to be two traditions examining BIF: medical and pedagogical. Generally speaking, the medical tradition concentrates on BIF as a consequence of some medical condition, and the pedagogical tradition concentrates on the difficulties of teaching children with BIF. A research tradition focusing on BIF for its own sake, however, is lacking. There is also no unanimous term for the phenomenon of BIF, and it has had numerous names in the past. The names used in the literature include, for example, borderline mental retardation, slow learner, mild cognitive impairment, and general learning disability.
Historically, there has been interest in BIF in the intellectual disability (ID) community in the late 1960s and again in the late 1990s. In the 1960s, the American Association on Intellectual and Developmental Disabilities (AAIDD; formerly called the American Association on Mental Retardation) defined those with an IQ test score of 70 to 85 as eligible for classification as mentally retarded (Heber, 1959, 1961). The report by the President's Committee on Mental Retardation (PCMR), “The Six-Hour Retarded Child,” identified the group of children who were labeled as mentally retarded during school hours based solely on an IQ test score without regard to their adaptive behavior (President's Committee on Mental Retardation, 1969). Outside the academic setting, however, these children seemed to manage reasonably well. The report was also concerned with overrepresentation of ethnic groups and poverty among those children labeled as mentally retarded. In the early 1970s, the classification system around mental retardation was modified, and BIF was removed from the diagnostic category (Grossman, 1973). Three decades later, the PCMR revisited “The Six-Hour Retarded Child” in a report and book entitled The Forgotten Generation (President's Committee on Mental Retardation, 1999; Tymchuk, Lakin, & Lucasson, 2001). The general proposition of the report was that adults with mild cognitive limitations face challenges in every aspect of life—finances, employment, housing, well-being, and family—and that they become more vulnerable as the demands of society grow increasingly complex. This proposition was supported by Fujiura (2003) when he studied a group of Americans with mild intellectual impairments and found a sizable cohort of Americans who shared many support needs and social and economic vulnerabilities with those labeled as “mentally retarded.” This cohort was distinguishable from the general population of the United States as well as from those with SLDs.
Since 1973, BIF has not been included in any diagnostic category. The 11th edition of the AAIDD manual (Schalock et al., 2010) defines intellectual disability as being characterized by significant limitations (approximately 2 standard deviations below the mean) in both intellectual functioning and adaptive behavior. Even though BIF does not meet the diagnostic criteria of ID, the AAIDD manual recognizes it by describing individuals who fall slightly above the upper ceiling for a diagnosis of ID and often face challenges in society that are similar to those faced by persons with ID who have lower IQ scores. The manual calls for nonstigmatizing, accessible, and individualized support for these people. In the Diagnostic and Statistical Manual of Mental Disorders (5th ed.; DSM-5; American Psychiatric Association, 2013), BIF is recognized in a section labeled “other conditions that may be the focus of clinical attention.” The manual advises that differentiating BIF and mild ID requires careful assessment of intellectual and adaptive functions and their discrepancies. The manual defines intellectual disability as a disorder that includes both intellectual deficits (approximately 2 standard deviations below mean; IQ test score of 65–75) and adaptive functioning deficits. DSM-5 emphasizes the role of adaptive functioning. It advises that the levels of severity should be defined based on adaptive functioning and that clinical judgment should be used when interpreting the results of IQ tests. DSM-5 also points out that IQ test scores are approximations of conceptual functioning and that, for example, a person with an IQ test score above 70 may have severe problems in adaptive functioning. The DSM-5 and the AAIDD manuals emphasize the role of adaptive behavior. In the International Classification of Diseases (ICD-10; World Health Organization [WHO], 1992), mental retardation is described as a condition of arrested or incomplete development of the mind. It is characterized by impairment of the skills that contribute to intelligence (i.e., cognitive, language, motor, and social abilities). The role of IQ is emphasized when the manual describes the degrees of mental retardation that are estimated by standardized intelligence tests and that can be supplemented by assessing social adaption.
For the people with BIF, all the manuals imply the same difficulty: However severe a person's problems with adaptive behavior may be, people with BIF are not eligible to receive a diagnosis of ID and they cannot gain access to ID-related support and services because their IQ is “too high.” This is true if IQ is used as an eligibility criterion instead of needs-based access to support and services. People with BIF can also be left without appropriate support in school because they are not necessarily eligible for the special education services meant for students with SLDs. In the DSM-5, SLD is described as a “specific learning disorder” and in ICD-10 it is defined as “specific developmental disorders of scholastic skills.” Both definitions include the assumption that learning difficulties are not attributable to intellectual disabilities and that there is a discrepancy between performance in academic skills and the general level of intelligence.
The field now acknowledges, in research as well as in practice, the multiple social and functional problems as well as the lack of targeted support related to BIF. But despite this acknowledgement and the earlier interest in the topic, the research remains scant. There is still no unanimous name for the phenomenon and no norms to guide how to treat people with BIF. The purpose of the present study is, in addition to increasing knowledge about the problems and risk factors related to BIF, to examine the prevalent situation in light of current literature and to bring the topic up for societal and research discussion. First, we were interested in the general information on studies concerning BIF: publication and population details, the terms used, and the quality of the studies. Second, we wanted to know what difficulties people with BIF face across their life courses in comparison to those the general population faces. Third, we attempted to determine whether there were any risk or preventive factors visible in the studies.
In our study, we systematically collated empirical group evidence that fit prespecified eligibility criteria in order to answer research questions. We used explicit, systematic methods that were selected with a view to minimizing bias.
We originally planned to include studies if the participants in them had an IQ test score measuring 70 to 85. However, as only a few studies used this exact criterion (Table 1), we also included studies where most of the subjects in a BIF group met the criterion. No age limits for the participants were set. Group studies reporting results relevant to the study questions were included. Outcomes were considered relevant if they dealt with neuropsychological, social, mental health, independence, risk, or preventive issues.
Search Methods for Identification of Studies
In March 2012, we carried out our electronic database searches. The following databases were searched: ERIC (Educational Resources Information Center, 1960 to March 2012), ISI (Web of Science, 1945 to March 2012), MEDLINE (1950 to March 2012), and PsycINFO (1887 to March 2012). All searches were restricted to the English language. The search strategy used to search ISI was “Topic,” which searches the title, abstract, and key words of the records. This search was modified as necessary to search other databases (modifications available upon request from the first author). The search terms (combined with “or”) were as follows: borderline developmental disability, borderline intellectual functioning, borderline intellectual disability, borderline IQ, borderline learning disability, borderline mental retardation, minor intellectual disability, general learning disability, general learning disorder, gray-area children, marginal learners, slow learners, garden variety slow learners, and nonspecific learning disabilities.
In addition to these searches, we carried out two other searches in the same period. We noted the references of the studies included for review from the database searches and then conducted a manual search for the titles and abstracts in four journals (American Journal on Intellectual and Developmental Disabilities, Intellectual and Developmental Disabilities, Journal of Applied Research in Intellectual Disabilities, and Journal of Intellectual Disability Research) from the beginning of 2000 to March 2012. These journals were selected on the basis that they were the ones most commonly found in the reference lists of the studies that had already been included.
Study Selection Process
Citations identified through database searches were transferred to the reference management program RefWorks (www.refworks.com) and duplicate citations were removed. The titles and abstracts were examined and a study was excluded if the inclusion criteria were not met. The excluded records were coded into one of four categories: (a) “no IQ”: intelligence quotient was not determined in the study; (b) “wrong IQ”: intelligence quotient of most of the participants did not fall within the range of 70 to 85; (c) “not a group study”: only group studies were included in order to collate extensive information; and (d) “irrelevant results”: the results of the study were irrelevant to our study questions. If the criteria were fulfilled, or fulfillment was uncertain, the studies were included for full-text evaluation, and they were assessed for eligibility using the four categories previously described. If the reason for exclusion was “irrelevant results,” the study was marked on an extraction sheet (see “Methods for Handling Data”). The studies that fulfilled the inclusion criteria were included in the final sample for systematic review. At this point, the studies identified through additional searches were included in the sample. The study selection process was carried out by the first author (MP, a doctoral student). Reliability was confirmed by two reliability checks: random samples of 88 (7.3%) citations identified through database searches, and 10 full-text articles (5.8%) were coded by another author (VN, PhD). In both checks, the initial agreement between coders was 90% and a final consensus was reached by discussing the differences.
Methods for Handling Data
An extraction sheet was used for the studies that were excluded as irrelevant. The sheet covered author, publication year, population characteristics, aim, results and conclusions, and reasons for exclusion.
An inclusion sheet was used for the included studies to collect descriptive information. The following information was gathered: author, year, country, journal, number of participants (study/control groups), gender, age, term used, data collection year, how data was collected, IQ score and how it was measured, aims, hypotheses, main outcomes, and main conclusions.
The methodological quality of the included studies was evaluated using the criteria created by Dalemans, De Witte, Wade, and Van Den Heuvel (2008). They based their list on different criteria lists for nonrandomized studies (Downs & Black, 1998; Prins, Blanker, Bohnen, Thomas, & Bosch, 2002). The criteria list has later been used also by Verdonschot, de Witte, Reichrath, Buntinx, and Curfs (2009). The list consists of 15 items: six items describe aspects of informativity, four items describe external validity, and five items describe internal validity. Two authors (MP and VN) assessed the quality of all the studies independently, reaching initial agreement of 91% (Cohen's kappa 0.78). Final consensus was reached by discussing the differences.
Figure 1 shows the study selection process. The database searches produced 1,726 citations. On the basis of the title and abstract evaluations, full texts of 203 studies were obtained, of which 45 were included in the final sample. Four studies were found by the additional searches, bringing the final number of studies to 49.
General Information on Studies
Publication and population details
Table 1 shows publication and population details of all 49 studies. Of these, 33 were published after 2000, 10 in the 1990s, 5 in the 1980s, and 1 in the 1970s. Thirty-nine percent took place in the United States. Other locations were the Netherlands (18%), United Kingdom (12%), Germany (8%), India (6%), Australia (4%), Canada (4%), and one study each from Italy, Norway, Finland, New Zealand, Sweden, and Iran. The sample sizes varied between 11 and 8,450 subjects (mean 667). The total of all subjects was 32,663. Of the studies, 48% focused on children, 23% on children and adolescents, 10% on adolescents, 12% on adults, and the rest spanned multiple age groups. In 26 studies, the majority of subjects was men; in only two studies was the majority composed of women.
Of the original 16 terms included in the database searches, seven were used in the included studies: borderline intellectual disability (in seven studies), borderline intellectual functioning (15), borderline intelligence (eight), borderline IQ (six), borderline learning disability (one), borderline mental retardation (four), and slow learners (five).
Methodological quality of the studies
The methodological quality rating of the studies varied between five and 15 (maximum possible 15), with the mean being 10.9 (see Table 2). In general, informativity and internal validity were better than external validity.
Only 8.2% of the studies were conducted with general population samples. Subjects with BIF were often recruited from clinics or hospitals (30.6%), and from special schools or special education classes (30.6%; Table 1).
Academic and cognitive skills
Eighteen studies concentrated on the academic or cognitive skills of people with BIF. The studies, recruitment processes of the subjects, tests used, and outcomes are described in detail in Table 3. Limitations for generalization and comparison of results are set by the different subject recruitment processes among the studies. The place of recruitment varied between mainstream schools (six studies), special schools (nine), and clinics (three). In addition to below-average IQ, criteria for recruiting the subjects also included reading difficulties (in 3 studies), mathematical difficulties (1), and poor school performance (3). It is thus possible that participants with difficulties in academic achievement were overrepresented in the BIF samples. The methodological quality scores of the studies (Table 2) varied from 5 to 14 (mean 9.7). The number of subjects varied between 45 and 411 (mean 134).
An assessment of memory skills was included in eight studies (Alloway, 2010; Birch, 2003; Bonifacci & Snowling, 2008; Henry, 2001; Kortteinen, Närhi, & Ahonen, 2009; Maehler & Schuchardt, 2009; Schuchardt, Gebhardt, & Maehler, 2010; Swanson, 1994). Overall, children with BIF were outperformed by peers of the same age with average intelligence. Two exceptions were reported for visual memory by Henry (2001) and Kortteinen et al. (2009). They found no differences between the BIF group and the control groups. Henry also studied more complex memory functions, in which the performance was at the same level in both groups, although Henry believed a ceiling effect for the average control group was possible. In general, a deficit in memory functions of people with BIF seems clear. In 51 tasks, performance was poorer than that of average intelligence peers, and in only six tasks was performance at the same level for both groups.
Schuchardt et al. (2010) and Schuchardt, Maehler, and Hasselhorn (2011) studied whether the deficits in working memory functions reflect a developmental lag or a qualitative deviation from normal development by matching the mental ages of children with BIF with those of average intelligence. In the first study (2010), they reported that children with BIF show structural abnormalities in their phonological store, and developmental lags in their visuospatial and central executive subsystems. After the second study (2011), however, they concluded that, because the performance of children with BIF corresponded with that of a control group matched for mental age, the findings indicated a developmental lag. Hence, these children's working memory functions seemed to develop in line with their general intellectual abilities.
Alloway (2010) and Hartman, Houwen, Scherder, and Visscher (2010) studied the executive functions of children with BIF by comparing them with a control group of children of the same age. Hartman et al. (2010) studied planning skills, and Alloway examined shifting attention, cognitive inhibition, problem solving, planning, and response inhibition. Poorer results were detected systematically in all presented tasks. A few other studies also examined skills related to executive functions. Napora-Nulton (2003) studied processing speed in familiar and novel situations, and reported all reaction times as being slower in the BIF group than in the average children. In addition, Bonifacci and Snowling (2008) studied processing speed. Four reaction-time tests showed poorer performance (in speed and in accuracy) for children with BIF than for the children in the control group, as did the test of sustained attention. However, van der Meer and van der Meere (2004) reported impulse control skills of children with BIF as being similar to those of children in the control group with average intelligence.
The arithmetic skills of children with BIF were constantly observed to be poorer than those of the children in the average control group (Claypool, Marusiak, & Janzen, 2008; Kortteinen et al. 2009; MacMillan, Gresham, Bocian, & Lambros, 1998). The results for reading, spelling, and related skills were not as clear. MacMillan et al. (1998) found reading and spelling skills to be poor, as did Kortteinen et al., regarding their participants with BIF and reading disabilities (RD; 47% of their participants with BIF also had RD). However, the reading and spelling skills of the participants with BIF and no RD were at the level of average controls. Bonifacci and Snowling (2008) as well as Claypool et al. (2008) found reading skills to be better in the average control group. Although Atkinson (1984) did not compare the adults in his study directly with the average control group, the results indicate poor reading skills among the adults, because 17% were defined as semiliterate and 12% as nonreaders. Regarding the other skills related to reading and spelling that were studied, performance in most of the others tasks was poor (auditory processing, phonological and syntactic skills, and reading comprehension; Birch, 2003; Kortteinen et al., 2009). For participants with BIF and no RD, rapid naming performance was at the level of those in the average control group, but for those with BIF and RD, task outcomes were poorer (Birch, 2003; Kortteinen et al. 2009).
A study on acquiring abstract theories (Mähler, 2005) reported that students with BIF used theories that were more immature than the ones used by their peers of average intelligence. They preferred intentional mechanism as a relevant causal explanation (e.g., “If I wish my child to have blue eyes, she will have them.”) over more sophisticated mechanisms. Children with BIF were also found to use less advanced learning strategies than other children used. For example, they preferred an “isolation strategy” over a “global strategy”; that is, they focused on a single element of a task rather than on its global element (Swanson, 1994). Swanson also found that students with BIF benefit less from hints or clues than students in control groups do.
In their follow-up study, Fernell and Ek (2010) reported that pupils with BIF, assessed in Grade 4, received grades that were significantly lower when finishing compulsory school than those received by their peers with average intelligence.
Three studies (Hartman et al., 2010; Vuijk, Hartman, Scherder, & Visscher, 2010; Westendorp, Houwen, Hartman, & Visscher, 2011) reported on the motor skills of children with BIF. Vuijk et al. (2010) found that 40% of children with BIF showed no problems with motor skills, 17% showed borderline problems, and 43% showed definite motor problems when their performance was compared with the norms of the Movement Assessment Battery for Children. Both Hartman et al. and Westendorp et al. studied gross motor skills. They similarly found children with average intelligence outperforming students with BIF on both subscales. Westendorp et al. reported that, of the 12 skills studied, only three (hop, jump, throw) were at the same level of performance in the two groups.
Comparison of mild intellectual disability (MID) and specific learning disabilities (SLDs)
Some of the above studies also reported results for people with MID or SLDs. Six studies (see Table 3) compared subjects with MID and BIF, mainly covering areas of memory, motor, and academic functioning (Hartman, et al. 2010; Henry, 2001; MacMillan et al., 1998; Schuchardt et al., 2010; Vuijk et al., 2010; Westendorp et al., 2011). Out of 17 measured skills, the BIF group outperformed the MID group in 11 skills and was at the same level in six skills. The difference between groups was clear in memory and motor functioning, because the results mainly showed better performance among people with BIF than among those with MID. Five studies compared subjects with SLDs and BIF, covering areas of memory, executive functioning, and reading (Bonifacci & Snowling, 2008; Birch, 2003; Maehler & Schuchardt, 2009; Napora-Nulton, 2003; Swanson, 1994). The SLD group outperformed the BIF group in 10 out of 15 measured skills. In five measures, the performance of both groups was at the same level. Differences between groups were particularly clear regarding executive functioning because all the measures showed poorer performance among the BIF group.
Studies concerning peer interaction showed more solitary play and less peer- or group-play behavior among children with BIF than among those in the average control groups (Guralnick & Groom, 1987; Roberts, Pratt, & Leach, 1991). Peers seem to have a great impact on behavior: When children with BIF were paired with same-age average control children, twice as much positive interaction was observed as when they were compared with pairs of two children with BIF (Guralnick & Groom, 1987).
Fenning, Baker, Baker, and Crnic (2007) found that mothers of children with BIF exhibited less positive and less sensitive parenting than the mothers of children in the groups with average or significantly below average intelligence (IQ test score < 70). These mothers were also the least likely to display a style of positive engagement. Although children with BIF did not behave more problematically than children with average intelligence in observational situations did, their mothers reported that they had more externalizing symptoms. According to the authors, the findings suggest that parental understanding of the problems of children with BIF is inadequate and is not on the same level as, for example, the parental understanding of a child with mental retardation. The authors concluded that children with BIF are at risk for poor parenting.
Social information processing seems to differ among children with BIF when they were compared with their peers. Embregts and van Nieuwenhuijzen (2009) and van Niuwenhuijzen, Vriens, Scheepmaker, Smit, and Porton (2011) studied children's responses to video vignettes describing demanding social situations. Both studies reported mainly more passive and/or aggressive and less assertive responses in various social situations (e.g., generating spontaneous responses to problems, evaluating social situations, choosing from different ways to behave) for children with BIF than for children in the control group.
McAlpine, Kendall, and Singh (1991) studied the recognition of facial expressions in children with normal intelligence, borderline intelligence, and intellectual disability (ID). They found that children with BIF recognized all facial expressions of emotion more frequently than their peers with ID did, but less frequently than their peers with average intelligence. Van Nieuwenhuijzen et al. (2011) studied facial expressions as well and found no difference between the groups in the recognition of sad, happy, and angry expressions. However, fear was less recognized by the children with BIF than by their peers.
Seltzer et al. (2005) found no differences in the frequency of meeting friends and relatives between high school graduates with BIF and their average-intelligence siblings in adulthood. Additionally, no differences were found in the proportions of people who had the opportunity to turn to friends, parents, siblings, or children in times of trouble. The only difference between groups was observed in the number of organizations participated in, with the BIF group participating in fewer organizations than their siblings.
Three studies showed the prevalence rates of BIF in different criminal populations: 30% among adults in pretrial detention (Crocker, Gote, Toupin, & St-Onge, 2007), 33% among juvenile criminal offenders (Schuster & Guggenheim, 1982), and 47% among incarcerated “hardcore” juveniles (Hollander & Turner, 1985). People with BIF seem to be overrepresented in populations of criminal offenders, because the presumed prevalence of BIF would be about 14%.
Douma, Dekker, Ruiter, Tick, and Koot (2007) studied antisocial and delinquent behaviors in 526 youths (aged 11 to 24 years) with BIF. Each type of behavior (mean to others, physical aggression, theft/arson, property destruction, authority avoidance) was exhibited by roughly 10–20% of youths. Of these, physical aggression, theft/arson, and property destruction were seen more in the BIF group (19%, 12%, and 13%, respectively) than in the general population group (13%, 6%, and 8%, respectively). Boys with BIF, but not girls, displayed antisocial behavior.
Among the children with poor school performance, the prevalence of behavioral disturbances was 94% for children with average intelligence and 86% for children with BIF (Thompson, Lampron, Johnson, & Eckstein, 1990).
The prevalence of mental health problems among people with BIF was reported in four studies (Chen, Lawlor, Duggan, Hardy, & Eaton, 2006; Dekker & Koot, 2003; Emerson, Einfeld, & Stancliffe, 2010; Hassiotis et al., 2008). Whether comparing already existing diagnoses (Chen et al., 2006) or symptoms of the mental health problems (Dekker & Koot, 2003; Emerson et al., 2010; Hassiotis et al., 2008), mental health problems were more prevalent among subjects with BIF than among the general population (Table 4). Although Dekker and Koot did not include a control group in their study, they concluded that the prevalence of most mental disorders among subjects with BIF exceeded the prevalence that was observed in the general population. Chen et al. found the current emotional problems of people with BIF to be at the same level as those with ID. Emerson et al. also reported a similar prevalence of emotional and conduct problems among children with BIF and with ID. In their study, children with ID had higher levels of hyperactivity and peer problems than the children with BIF did, but they also showed more prosocial behavior.
A different observation was made by Douma, Dekker, Verhulst, and Koot (2006) when they studied self-reports on mental health problems of youth with BIF. When they compared the emotional and behavioral problem scores on the Youth Self-Report Scale of youths with BIF with the scores of their average peers, they found that the youths with BIF did not differ.
Hassiotis, Tanzarella, Bebbington, and Cooper (2011) studied rates of suicidal behavior (thoughts and acts) among a large general population sample of people with BIF. When people with BIF were compared with the general population, the people with BIF were more likely to have attempted suicide or to have harmed themselves. However, these observations were likely not to be related specifically to BIF, because differences were no longer significant after controlling for income and age.
Mental health care
Hassiotis et al. (2008) found that individuals with BIF were proportionally less likely to receive treatment for mental health problems than the general population. When they did receive treatment, they were more likely to be treated with medication, and less likely to be treated with counseling. Dekker and Koot (2003) examined children and adolescents with BIF and found 27% of those with diagnoses receiving professional help. The study did not include a control group, but authors concluded the prevalence to be similar to rates found in the general population of the same age.
Employment and Marriage
The reliability of results of five studies dealing with employment and marriage were hampered by the small sample sizes (see Table 1) and the age of the studies, three of which were over 2 decades old. Recent figures on employment were available from only one study.
In the most recent study, Seltzer et al. (2005) studied 201 pairs of siblings, those with BIF and their siblings with average intelligence, at ages 35 and 52. The employment rates did not differ between the groups. However, occupational prestige at both ages and total earnings at age 52 were significantly lower in the BIF group. Subjects with BIF had also been employed longer at their current jobs than siblings with average intelligence. Kinge (1979) studied former pupils of classes for slow learners at age 30. He reported employment rates of 81% for men and 8% for women in the BIF group, and 94% for men and 41% for women in the average intelligence group. Of the married women, the proportion of housewives was 83% in the BIF group and 62% in the average intelligence group. Of the working men, the proportion working as unskilled workers was 50% in the BIF group and 19% in the average intelligence group. In regards to the occupational status of the subjects with BIF, 19% were low-income earners. These proportions were larger than among the subjects with average intelligence (6% of men and 33% of women). Kinge did not report statistical analysis on the differences between the groups. Atkinson (1984) studied former male pupils of a special school in their fifties and found 85% of them to be employed. Of them, 62% had held the same job for over 20 years and 70% worked unskilled or semiskilled jobs. Their average weekly wage was lower than that of the general population. Zetlin and Murtaugh (1990) reported an observation that differed from the above results when they followed adolescents with BIF before and after high school graduation. Although 80% of adolescents had held at least one job during a period of two years before graduation and one year after it, they had difficulty maintaining a job. They had held one to five jobs per person, with 1 month to 3 years being the longest time spent at one job. All jobs were unskilled or semiskilled.
Seltzer et al. (2005) found that, at age 35, the rate of marriage was somewhat lower among subjects with BIF than it was among their siblings of average intelligence (89% vs. 94%, respectively). Fifteen years later, such a difference was not found. There were also no differences in the number of children and in the proportion of those married to their first spouse. In Kinge's study (1979), 92% of women with BIF were married, which did not differ from women with average intelligence. Atkinson (1984) found 70% of former male pupils with BIF to be married. Hassiotis et al. (2008) found 56% of people with BIF and 70% of people with average intelligence were living as couples. The difference was likely to be due to the BIF group having been considerably younger than the group with average intelligence.
Risk and Preventive Factors
The methodological quality of 11 studies dealing with risk or preventive factors varied greatly, with scores from 6 to 14 points (mean 11.0; see Table 2). The sample sizes also varied widely, from 11 to 1681 (mean 397; see Table 1).
Potential risk factors for BIF were identified from seven studies. Based on three studies, the risk of BIF somewhat increased with low birth weight (LBW). The prevalence of LBW in the adult BIF population was higher (24%) than in the average intelligence population (13%) in the study of hospital birth cohorts (Chen et al., 2006). The prevalence of BIF in the group of 3-year-old children with LBW was 19% (Ramey et al., 1992). Another study reported 13% of 6-year-old children with LBW as having BIF, which is the expected prevalence of BIF in the general population (Chaudhari, Bhalerao, Chitale, Pandit, & Nene, 1999). However, the same children were followed to age 12 and the prevalence of BIF was found to increase to 24% (Chaudhari, Otiv, Chitale, Pandit, & Hoge, 2004).
Within a group of 15 4-year-olds with BIF, poor family environment (including factors such as poverty, being a member of a minority ethnic group, or living with a single parent) were present in 80% of the cases (O'Brien, Rice, & Roy, 1996). A low level of education for the mother was present in 76% of adults with BIF, compared with 68% in the average intelligence population (Chen et al., 2006). Exposure to toxic metals was detected as a potential risk for BIF in randomly selected school children with BIF, of whom 36% presented elevated levels of lead and 17% elevated levels of cadmium in comparison to 7% and 6%, respectively, in the control groups (Marlowe, Errera, & Jacobs, 1983). Farhadifar et al. (2011) reported mother's illiteracy, a familial history of ID, and maternal drug use during pregnancy as being more common for children with BIF (78%, 25%, and 21%, respectively) than for children in the control groups with average intelligence (39%, 4%, and 7%, respectively).
It seems that early developmental delays are connected to BIF. O'Brien et al. (1996) reported 73% of children with BIF as having delayed mental or motor development, as compared with 42% in children in the control group. Because the prevalence in the control group was so high, very mild developmental delays must have been included in the study. Another study reported 58% of children with BIF to have either delayed walking or delayed talking. Furthermore, 62% of children with BIF showed some form of soft neurologic signs (Karande, Kanchan, & Kulkarni, 2008. The study did not include a control group.
Three preventive factors that seem to have a positive effect on the life of those with BIF were detected: education, social contacts, and personal qualities. The positive influence of education was reported in three studies comprising favorable school records (Atkinson, 1984), education beyond high school (Seltzer et al., 2009), and more years of education (Vaillant & Davis, 2000). The same studies also reported the positive effects of social contacts, including supportive parents (Atkinson, 1984), role models for achievement (Seltzer et al., 2009), and warm relationships (Vaillant & Davis, 2000). Two studies reported personal qualities, including flexibility to change with situational demands (Atkinson, 1984), and childhood competence (including ego strength, perseverance, and relationship skills; Vaillant & Davis, 2000). It is not clear, however, whether the personal qualities have caused the more successful life paths or vice versa.
Only two studies on interventions for people with BIF were identified. Early training with 3-year-old at-risk children showed a prevalence of BIF from 5% to 8%, compared with 19% for the control group without training (Ramey et al., 1992). Adolescents with BIF showed improvement in acquiring community skills through the method of combining specific written instructions with feedback (Cuvo, Davis, O'Reilly, & Mooney, 1992).
The review shows that many people with BIF face major difficulties across their life courses. When they were compared with the general population, neuropsychological, social, and mental health problems were evident. Although employment rates were relatively high, wages were lower than average. Low birth weight, negative family environment, mothers with low education levels, and toxic metals can be seen as risk factors for BIF. Preventive factors included education, social connections, and some personal qualities.
The purpose of the review was to collect literature concerning BIF in order to increase knowledge about the problems related to BIF, to examine the prevailing situation in the field in the light of current literature, and to bring the topic up for discussion. Of the 1,726 original citations yielded from database searches, only 49 studies made it to the final sample of the review. This is an astonishingly low number considering the nearly 14% prevalence of people with an IQ score of 70 to 85. Fortunately, research interest seems to have grown in recent years. Since 2000, studies have been carried out in 14 different countries, with most of them occurring in the United States and the Netherlands. Age groups from 2 to 74 years were represented. Among those studied, males were more represented than females. The quality of the studies varied greatly.
In general, the information on BIF is fragmentary. Based on the review, cognitive and academic difficulties are clear. Of the 52 skills measured, the performance by people with BIF was at the level of their average intelligence peers in only seven of them. With the exception of memory, only a few studies in each area were found, and studies concerning many relevant cognitive functions (e.g., conceptual learning) were unavailable.
The difficulties in neurocognitive functioning were consistently reported in relation to same-age peers. It is only in the area of memory functions that people with BIF were compared with others with a similar mental age, and the conclusion was that memory functions were at the level of their general cognitive abilities (Schuchardt et al., 2010; Schuchardt et al., 2011). Using designs that utilize control groups with a similar mental age in other areas of neurocognitive functioning would provide important information on the similarities and differences between the development of subjects with BIF and typically developing children. The design might help to identify which aspect is differentiating subjects: a developmental lag or qualitatively different functioning. These kinds of studies are important for the sake of theoretical formulations and for intervention planning.
People with BIF seem to differ from the average population in terms of their social behavior. Differences were seen in play behavior, social information processing, recognition of emotions, social participation, and antisocial behavior. Children with BIF also seem to be at risk for poor parenting. In addition, subjects with BIF were found to be overrepresented in criminal samples. As compromised learning skills put children and adolescents with BIF at risk for poor school performance, they are at a higher risk for alienation or antisocial behavior than their peers.
The prevalence of mental health problems among people with BIF was repeatedly reported as being higher than in the general population. This finding is in line with several studies that have reported mental health problems as being more common among people with mild levels of ID (Koenen et al., 2009; Linna et al., 1999; Wallander, Dekker, & Koot, 2003). The results of one study on mental health contradicted the other reviewed studies. The study found self-reports of psychiatric symptoms in youths with BIF to be similar to those with average intelligence (Douma et al., 2006). The authors concluded that the results were in contrast to the commonly found higher rates of psychiatric symptoms in people with lower intelligence, but addressed that usually the results rely on parent or teacher reports, not on self-reports. However, of the other reviewed studies, Hassiotis et al. (2008; 2010) also used self-reports and, thus, the use of a self-report does not seem to be the sole reason for the results reported by Douma et al. Age differences between the adult subjects in the studies by Hassiotis et al. and the adolescent subjects in the study by Douma et al. might partly explain the difference, because it is possible that adolescents do not recognize their problems as readily as adults.
Four studies on employment consistently reported that people with BIF, when they were compared with the general population, held lower-skilled jobs, earned lower wages, and had longer careers in the same job. Although the adults studied had mostly found their places in the job market, adolescents struggled to find and keep a job. This difference may imply that the transition from adolescence to adulthood is problematic for young people with BIF. It may also reflect changes in the labor market, at least in Western societies. One possible assumption is that, as the requirements for education have increased and the number of lower-skilled jobs has decreased, it has become more difficult for young adults with compromised intellectual functioning and related poor educational attainment to find employment.
In general, the studies reported high employment rates for subjects with BIF. Kregel (2001) reported contrary unemployment rates of 70%–80% for individuals with mild cognitive limitations (IQ test score of 55–84). The rather encouraging high employment rates in the present review may prove too optimistic, because current studies with representative samples were unavailable. Two studies examined high school graduates (Seltzer et al., 2005; Zetlin & Murtaugh, 1990), and the samples in these studies are not likely to be representative of the whole population with BIF. Instead, they represent only those with strong scholastic aptitude. Two other studies were relatively old (Atkinson, 1984; Kinge, 1979), and it is questionable how well their results generalize to modern job markets. Kinge reported in 1979 that 92% of Norwegian women with BIF were housewives. This is likely not today's reality, at least not in Scandinavia, where equality in the job market has been established and being a housewife is no longer the norm.
The causes of BIF are unknown. There need not even be an actual cause, because BIF can be seen as part of the normal variation, with IQ being at the lower range of the normal distribution. The reported risk factors for BIF (low birth weight, mothers with low education, negative family environment), as well as those factors reported to protect people from adverse outcomes (education, social connections, and some personal qualities), were general and certainly not specific to people with BIF.
The results of studies of cognitive skills comparing BIF and MID or BIF and SLD show that BIF seems to fall somewhere between the two categories, because people with BIF mostly had a better performance level than those with MID and a worse level than those with SLD. It is possible that, due to the subject recruitment process, learning difficulties are overrepresented in BIF groups studied, which may affect the results. There are no comparisons between these three groups that consider the amount of support needed or received in real life. It can, however, be speculated that the BIF group is in a worse situation than the other two. Because the problems with BIF are not as visible as those in MID and not as specific as those in SLDs, they often go unrecognized and, consequently, no support is offered.
In general, the findings of the review showed that the performance level of people with BIF is typically worse than that of their average intelligence peers and better than that of people with MID. The relationship between these groups can be understood as one of being on different points of a continuum of intellectual abilities. However, social support and services do not form a similar continuum, but are delivered on a yes-or-no basis. In order to be eligible for services, certain criteria have to be fulfilled. The critical criterion in the area of compromised intellectual abilities is receiving a diagnosis of ID. If that criterion is met, services are readily provided, income (at least in Finland) is guaranteed with a pension, and expectations for excelling at school and in the labor market are lowered. For people with BIF who do not qualify for those services, the expectations for learning and independent living are similar to those of others, yet their lower intellectual abilities and related learning problems and less-than-optimal adaptive skills hamper their success in these areas. Social support lies behind cut-off lines, in this case behind a diagnosis of ID. Those just above the cut-off line do not have access to help and face high performance expectations.
The current literature shows that there is a large group of people that need support in multiple areas of life. The decision made a few decades ago to exclude borderline intellectual disability from ID diagnoses has had, and continues to have, a huge impact on the lives of people with BIF. The decision was affected by the possible stigma related to ID and concerns about overrepresentation of ethnic groups and poverty in the category of those with a mild level of ID. However good the intentions were, the decision excluded people with BIF from existing services and, since then, no official substitutive service has been introduced. For current policy makers, an important first step would be to acknowledge the existence of this group. The second step should be to decide what actions to take concerning services targeted at BIF. At the moment, there are no official positions regarding BIF or resources appointed for it. This gap leaves local officials to deal with the issue as best as they can and, in the worst case, people with BIF are left without any support. From a societal point of view, people with BIF are a large group of capable people who, when they are compared with the general population, struggle with many aspects of life. To enable the full inclusion and participation of persons with BIF in society, it is crucial to recognize the difficulties that lower-than-average intellectual abilities produce in, for example, vocational education and job training, and in developing services to accommodate these needs. With the right kind of timely support, people with BIF could find their place in society and labor markets.
Limitations of the Study
It is possible that some relevant studies were not included in the review due to the wide range of terms in the field used to describe the phenomenon. This possibility was minimized by searching through relevant journals and through the reference lists of the included studies.
BIF was defined solely on the basis of IQ test score, without taking into account other functional aspects. Although this definition is far from perfect, a systematic review would have been impossible to conduct without clear and globally used criteria.
Despite the obvious, everyday problems that people with BIF face, the issue seems to be almost invisible in the field of research. There is a need for longitudinal and population-based studies focusing on people with BIF. More research on the nature of BIF is needed (e.g., is it a qualitatively different functioning or a developmental lag that explains the differences between subjects?). The critical life periods to study are the transitions from compulsory school to secondary school and from secondary school to the labor market, because these are the points at which adolescents seem to struggle.
In addition, societal discussions of BIF are needed. People with BIF are left without official services in society because they often do not meet the criteria for special services. In reality, they often struggle with the demand to be “normal.” There is a major need for flexible support systems that are based on real needs.
This study was funded by the Finnish Slot Machine Association and Finnish Cultural Foundation, Häme Regional Fund. The authors wish to thank Heli Numminen and Hanna Kortteinen for their earlier work with the topic, Pekka Kuikka for his ideas concerning the topic, Tuuli Kiljala for her assistance in the acquisition of the articles, and Anneli Sintonen for her help with organizing the data.
Minna Peltopuro, Timo Ahonen, and Jukka Kaartinen, University of Jyväskylä, Jyväskylä, Finland; Heikki Seppälä, FAIDD, Helsinki, Finland; and Vesa Närhi, University of Eastern Finland, Joensuu, Finland and Niilo Mäki Institute, Jyväskylä, Finland.