Which phonological memory measure distinguishes children with from children without auditory processing disorders?

The aim of this study was to compare performance in phonological working memory across children with auditory processing disorders (APD), partly associated with different developmental disorders, and to examine which memory measure distinguishes children with from children without APD. Data from a health care research database were retrospectively analyzed. All children completed three measures of phonological serial recall: one-syllable digits, one-syllable words, and multi-syllabic nonwords of increasing length (2-6 syllables). The study included 48 subjects with APD only (mean age: 87.9 months; SD 12.9 months); 35 with co-occurring APD and developmental language disorder (DLD) (mean age: 82.2; SD 13.4); 16 with co-occurring APD and dyslexia (mean age: 109.2; 18.7); 10 with co-occurring slight intelligence reduction and APD (mean age: 102.4; SD 12.5), while 13 were non-affected, typically forming controls (non-APD; mean age: 86.0; SD 16.1). The mean intelligence (T-score) of the first three mentioned APD groups and the controls were significantly different (p < 0.0001) from the children with lowered intelligence and APD (35.4; SD 3.1). Mean performance in all three verbal memory measures was significantly different from the controls in children with DLD+APD (digit span: p=0.0009; word span: p=0.0012; nonwords: p=0.05) and children with APD only (digit span: p=0.0037; word span: p=0.0006; non words: p=0.05). The set of memory measures (excluding the small and older group with lowered intelligence) went into a stepwise discriminate analysis. A significant group separation APD vs. non-APD was made by the subvocal rehearsal component of the phonological loop, as indicated by word span (F=14.40; p=0.0002) and digit span (F=2.87; p=0.0932). The rate of misclassification was 23%. The present results raise the question of the validity of APD diagnostics which includes the phonological working memory on principle. It remains controversial whether the assessment of the cognitive function "phonological memory" should remain part of (C)APD diagnosis. Further research should aim to replicate the study with subjects of the same age and supplementary studies of the "central executive" and processing precision of the phonological store.