• CNV
• socioeconomic status

Socioeconomic status (SES) is a measure of an individual’s or family’s economic and social status based on factors, such as income, education and occupation. SES is a major determinant of health and related outcomes. Lower SES confers increased risk for multifactorial disorders like stroke and cardiovascular disease and plays a key role in child health and development.1–4 Early life adversity negatively impacts child health and produces lasting and deleterious effects on developmental outcomes.5 Biological factors including genetic variants that may influence SES are only beginning to be understood. Recently, an association between pathogenic CNVs and lower SES in clinically unaffected adults from the UK Biobank was described.6 However, this study examined a limited spectrum of common CNVs and there was no information available on the inheritance of these CNVs.6 The aim of our study was to investigate the correlation of a wide range of unselected pathogenic and likely pathogenic CNVs, and their inheritance pattern, with SES.

We interrogated an anonymised departmental database of results from >17 000 postnatal, mostly paediatric (98%), clinical array comparative genomic hybridisations performed at the Manchester Centre for Genomic Medicine between 2010 and 2017. This database included information on each identified CNV, its clinical pathogenic classification, size, type (loss or gain), where available inheritance status (de novo or inherited from a parent), and postcode of the patient (online supplementary methods).

We identified 473 unique unrelated cases of pathogenic or likely pathogenic autosomal CNVs (table 1 and online supplementary table S1) with known inheritance status (n_inherited=218 ; n_{de novo}=255) and available postcode information (online supplementary methods). We then obtained Index of Multiple Deprivation Rank (IMDR) (and the seven constituent domains) associated with each of these postcodes (https://www.gov.uk/government/statistics/english-indices-of-deprivation-2015). English Indices of Deprivation is a widely used measure of SES in health research7 8 and reflects the SES of the households of the affected individuals in this study (online supplementary table S1).

We found that IMDR composition for families with pathogenic or likely pathogenic CNVs was significantly different when compared with the IMDR of the general population of the North West England (p_x ² _{goodness of fit}=1.8×10^{− 8}). Furthermore, families with inherited pathogenic and likely pathogenic CNVs were significantly more likely to be living in areas of higher deprivation when compared with families that have individuals with de novo pathogenic and likely pathogenic CNVs (p_x ² _trend=2.1×10⁻⁶) or with the general population of North West England (p_x ² _trend=4.3 ×10⁻¹⁴) (figure 1 and online supplementary figure S1) (online supplementary results). This difference was significant across the following deprivation domains of IMDR—income; employment; health; education, skills and training (online supplementary figures S2–S6). There were no significant differences in the following domains: barriers to housing and services, and living environment (online supplementary figures S7 and S8). We also performed the Jonckheere-Terpstra test in both de novo and inherited CNVs to see if there was any correlation between SES and age of diagnosis. No significant trends were identified (p-value=0.0615 for de novo CNVs and p-value=0.1615 for inherited CNVs) (online supplementary figure S9). IMDR comparisons according to CNV type did not reveal any significant difference between CNV losses and gains (p=0.52) (online supplementary results) (online supplementary figure S10). There was no evidence of an effect of CNV size on IMDR (online supplementary results) (p-value Kruskal-Wallis rank-sum test=0.48) (online supplementary figure S11). Our data show that the presence of pathogenic and likely pathogenic CNVs is correlated with SES. Notably in our cohort, this correlation seems to be driven by partially penetrant inherited CNVs of low effect size rather than de novo CNVs that are usually associated with more severe and more penetrant phenotypes.9

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Presence of pathogenic or likely pathogenic CNVs is correlated with socioeconomic status and is primarily driven by inherited CNVs. An anonymised Departmental database of results from over 17 000 postnatal clinical array comparative genomic hybridisation testing performed at Manchester Centre for Genomic Medicine between 2010 and 2017 was curated for CNVs, clinical classification, size, loss or gain status, the inheritance status (de novo or inherited from a parent) where available, and postcode. From the postcodes associated with the CNVs, we retrieved the Index of Multiple Deprivation Rank (IMDR) and its seven constituent domains, using the English indices of deprivation 2015. The distribution of IMDR deciles was compared for inherited (dashed line) versus de novo (dotted line) CNVs against the general population of the North West of England. χ² tests of independence and trend were performed on the IMDR deciles. The significance value for the χ² test was set at p<0.05.

The vast majority of the probands in our cohort are children (98%), and therefore, the IMDR dataset reflects the SES of the household in which they are growing up. The parents of children with de novo CNV do not carry the CNV and are not affected by the condition of their children. On the other hand, at least one of the parents of individuals with inherited CNVs will be carrying the same CNV. Almost 51% (n=112) of inherited CNVs in our cohort can be classed as recurrent (table 1 and online supplementary table S1).9 The penetrance of these recurrent CNVs has been estimated to range between 10% and 62%.9 Based on reported estimated penetrance of these recurrent CNVs, majority of the carrier parents of our index cases are likely to be classed as medically unaffected. The high level of deprivation observed in our cohort suggests that being a carrier of a low penetrant negatively impacts SES even in the absence of a medical phenotype. Our observations indicate that there are likely subclinical effects in individuals who are medically non-penetrant carriers of milder pathogenic and likely pathogenic CNVs. This agrees with the recent findings from the UK Biobank study whereby carriers of pathogenic CNVs had lower levels of household income and higher deprivation in the absence of neurodevelopmental disorders.6 Lower SES in families with medically relevant inherited pathogenic and likely pathogenic CNVs with milder phenotype could therefore be due to cumulative multigenerational consequences of these subclinical effects.

These results demonstrate that in addition to the primary phenotypes of the pathogenic and likely pathogenic CNVs, their secondary socioeconomic and resultant medical consequences need to be studied, especially in families with inherited CNVs. The combined frequency for a subset of these pathogenic CNVs in the general population is estimated to be at least 3.8%10 and therefore are significant in the context of public health. The correlation of SES with inheritance patterns of pathogenic and likely pathogenic CNVs, therefore, provides further unique insights into biological determinants of SES and has important implications for the planning of medical and social services.