When reading the news on our phone, or a physical copy of a book, we usually do it automatically. We rarely, if ever, think about all the complex processes that take place in order to let us recognise that visual information. Given the amount of written content in our day-to-day life, it goes without saying that our ability to read is an incredibly useful skill.
However, for many people, such processes are disrupted. In particular, people with deficits on phonological processing, the core reading skill, will struggle to decode written symbols into sounds. In other words, they will have a hard time trying to identify and interpret the written language.
In the case of dyslexia, where this impairment is not due to cognitive dysfunction, there have been observed similarities with schizophrenia in terms of impaired reading processes (Whitford et al., 2018). These deficits have been investigated by studies showing the existence of associations between poor reading skills and increased antisocial behaviour (Maughan et al., 1996; Trzesniewski et al., 2006), with evidence for forensic populations (Baker & Ireland, 2007) as well as people with various mental illnesses (Dåderman et al., 2004). Nonetheless, it is not clear at what level this is true for both schizophrenia and other relevant mental illnesses.
This systematic review and meta-analysis by Vanova et al. (2020) aims to understand the extent to which reading deficits represent a major problem for people with different types of mental illness, and to identify the tests that best assess these impairments while taking into account differences between forensic and non-forensic groups.
Previous evidence has shown a link between reading deficits and mental illnesses, but the degree of impairment is yet to be established.
Methods
The search (up to February 2020) involved the use of Google Scholar as well as databases including Academic Search Complete, CINAHL Plus, PsycINFO, PsycARTICLES, SocINDEX, MEDLINE via EBSCO Host and Pubmed.
Inclusion criteria for observational studies (case-control, cohort and cross-sectional):
- measures assessing reading skills in adults with a range of mental health disorders;
- standardised tests also translated in other languages;
- quantitative data;
- peer-reviewed journals (in English);
- both abstract and full-text.
Exclusion criteria for the same type of studies:
- non-peer reviewed articles;
- absence of standardised tests when using experimental methods;
- single-word reading tests used to assess premorbid IQ only;
- neurodevelopmental conditions such as ADHD and autism.
Data on tests and measures, participant characteristics, main findings, language of assessment and country was extracted. If calculation of effect sizes for patient and healthy control groups was possible, studies were included in the meta-analysis, while the remaining articles were summarised in a narrative synthesis.
Statistical analyses for the meta-analysis, which was conducted using RevMan, involved the use of Hedge’s g (standardised mean difference), a random-effect model, and I2 (heterogeneity), as well as the comparison of different mental illnesses with healthy controls on reading skills and potential differences among tests. No risk of publication was observed.
Results
Overall, 34 studies (meta-analysis = 19; narrative synthesis = 15; both = 5) were included in the review.
Non-forensic populations (n = 27)
Schizophrenia was significantly associated with a number of reading deficits from an early age, including:
- phonological processing (Hedge’s g = -0.88, df = 24, p = 0.00001, CI [-1.07 to -0.70]);
- comprehension (Hedge’s g = -0.96, df = 34, p = 0.00001, CI [-1.15 to -0.78]);
- reading rate (Hedge’s g = -1.22, df = 4, p = 0.002, CI [-1.98 to -0.46]);
- fluency (Hedge’s g = -2.03, df = 4, p = 0.00001, CI [-2.82 to -1.24]);
- reading-related skills (with males showing greater deficits compared to females)
- single-word reading and speed (at a lower degree compared to the above deficits).
There was very small/nonsignificant evidence for depression (n = 2) and single-word reading, while no difference relative to healthy controls was found for bipolar disorder. Associations between psychopathy (n = 1) and medium-size deficits in phonological processing and comprehension were reported. Finally, individuals with unspecified mental illnesses (n = 3) performed below the norm, but better than those with schizophrenia.
Forensic populations (n = 7)
Individuals with personality disorders/psychopathy (n = 3) showed marked phonological processing and decoding deficits (Hedge’s g = -0.85, df = 0, p = 0.0001, CI [-1.22 to -0.47]), issues related to comprehension (Hedge’s g = -0.95, df = 0, p = 0.0003, CI [-1.48 to -0.43]), single-word reading, and speed. Among the studies focusing on various (i.e. non specified or mixed) mental illnesses, only one reported phonological processing and comprehension to be below the norm, while the others showed mixed results.
Other results
The only study that compared non-forensic and forensic populations found larger deficits for incarcerated individuals in phonological processing and decoding and comprehension. Other factors with an effect on reading deficits were positive/negative symptoms in schizophrenia and high dosage of antipsychotic medications, general verbal skills, and educational level.
Finally, only tests assessing comprehension and accuracy (NDTR and GORT-4), and rate (GORT) in schizophrenia were able to detect large deficits. However, when these reading skills were assessed using the Alouette, a comparable French test, they could not detect the presence of deficits.
Schizophrenia in non-forensic samples and personality disorders & psychopathy in forensic samples appear to be particularly linked to reading deficits.
Conclusions
Most studies showed that individuals with a range of mental health disorders are characterised by reading skill deficits. This is particularly true for schizophrenia in non-forensic populations, and for personality disorders/psychopathy in forensic groups.
These findings suggest a link between several mental illnesses and reading impairment.
Strengths and limitations
The authors have not discussed the strengths and limitations of their review. However, it is worth mentioning that the use of meta-analysis alongside the narrative synthesis represents a powerful tool to combine data and obtain more accurate estimates of the overall effect size. Moreover, one of the potential issues related to systematic reviews is the risk of incurring publication bias, which was avoided through the use of Egger’s and Begg’s tests and funnel plots.
Nonetheless, some limitations also apply. The analyses did not include a quality assessment, which is a significant omission as this would increase the degree of confidence we have in the findings and enhance the robustness of the review as a result. In addition, potentially relevant demographic information such as participants’ ethnicity was not reported, thus undermining the generalisability of these findings.
The authors did provide us with a synthesis of the findings, which has been clearly reported in both the text and the tables, but not knowing the quality of the studies they have included is a problem.
The use of meta-analysis increased the accuracy of this review, but the lack of a quality assessment of the included studies means that we don’t know the overall robustness of the evidence.
Implications for practice
The review highlighted important clinical implications related to the presence of reading deficits in people with different mental illnesses. The authors emphasise the issues in diagnosing reading impairments such as dyslexia in people with mental illnesses, and push to acknowledge the need for more tailored programmes aimed at improving reading skills of these individuals. In doing this, they suggest a balance between improving less affected aspects of the reading process (which could compensate for the existing deficits) and the implementation of interventions based on already tested approaches.
When reading this review, I was particularly struck by two things. First, there appears to be a difference between men and women with mental illnesses in their reading skills, with men performing worse. It would be interesting for future studies to try and understand why that seems to be the case. Second, it was highlighted that these deficits might be present in the English language more often or intensively compared to other languages, such as in the case of developmental dyslexia (Ziegler et al., 2003). One possible explanation for this might be that the English language tends to be more inconsistent and less orthographically transparent compared to other languages, leading to further difficulties in decoding written words (Spencer, 2000). Understanding the differences among several languages might prove key to overcoming reading barriers.
In conclusion, I believe that this review makes some interesting points, and that these should be taken into consideration by researchers and policy makers in order to promote a real improvement on the reading skills (and, as a consequence, the general well-being) of people with mental health conditions.
This review recommends the use of a balanced approach where both affected and less affected reading skills of people with mental illnesses are targeted.
Statement of interests
None.
Links
Primary paper
Vanova, M., Aldridge-Waddon, L., Jennings, B., Puzzo, I., & Kumari, V. (2020) Reading Skills Deficits in People with Mental Illness: A Systematic Review and Meta-analysis. (PDF) European psychiatry: the journal of the Association of European Psychiatrists, 1-70.
Other references
Baker, S. F., & Ireland, J. L. (2007). The link between dyslexic traits, executive functioning, impulsivity and social self-esteem among an offender and non-offender sample. (PDF) International journal of law and psychiatry, 30(6), 492-503.
Burkholder, R. A., & Pisoni, D. B. (2003). Speech timing and working memory in profoundly deaf children after cochlear implantation. (PDF) Journal of experimental child psychology, 85(1), 63-88.
Dåderman, A. M., Lindgren, M., & Lidberg, L. (2004). The prevalence of dyslexia and AD/HD in a sample of forensic psychiatric rapists. (PDF) Nordic Journal of Psychiatry, 58(5), 371-381.
Maughan, B., Pickles, A., Hagell, A., Rutter, M., & Yule, W. (1996). Reading problems and antisocial behaviour: Developmental trends in comorbidity. (PDF) Journal of Child Psychology and Psychiatry, 37(4), 405-418.
Spencer K. (2000). Is English a dyslexic language? (PDF). Dyslexia (Chichester, England), 6(2), 152–162. https://doi.org/10.1002/(SICI)1099-0909(200004/06)6:2<152::AID-DYS158>3.0.CO;2-P
Trzesniewski, K. H., Moffitt, T. E., Caspi, A., Taylor, A., & Maughan, B. (2006). Revisiting the association between reading achievement and antisocial behavior: New evidence of an environmental explanation from a twin study. (PDF) Child development, 77(1), 72-88.
Whitford, V., O’Driscoll, G. A., & Titone, D. (2018). Reading deficits in schizophrenia and their relationship to developmental dyslexia: A review. (PDF) Schizophrenia research, 193, 11-22.
