Common patterns of morbidity and multi-morbidity and their impact on health-related quality of life: evidence from a national survey

Abstract

Background: There is limited evidence about the impact of specific patterns of multi-morbidity on health-related quality of life (HRQoL) from large samples of adult subjects.

Methods: We used data from the English General Practice Patient Survey 2011-2012. We defined multi-morbidity as the presence of two or more of 12 self-reported conditions or another (unspecified) long-term health problem. We investigated differences in HRQoL (EQ-5D scores) associated with combinations of these conditions after adjusting for age, gender, ethnicity, socio-economic deprivation and the presence of a recent illness or injury. Analyses were based on 831,537 responses from patients aged 18 years or older in 8,254 primary care practices in England.

Results: Of respondents, 23 % reported two or more chronic conditions (ranging from 7 % of those under 45 years of age to 51 % of those 65 years or older). Multi-morbidity was more common among women, White individuals and respondents from socio-economically deprived areas. Neurological problems, mental health problems, arthritis and long-term back problem were associated with the greatest HRQoL deficits. The presence of three or more conditions was commonly associated with greater reduction in quality of life than that implied by the sum of the differences associated with the individual conditions. The decline in quality of life associated with an additional condition in people with two and three physical conditions was less for older people than for younger people. Multi-morbidity was associated with a substantially worse HRQoL in diabetes than in other long-term conditions. With the exception of neurological conditions, the presence of a comorbid mental health problem had a more adverse effect on HRQoL than any single comorbid physical condition.

Conclusion: Patients with multi-morbid diabetes, arthritis, neurological, or long-term mental health problems have significantly lower quality of life than other people. People with long-term health conditions require integrated mental and physical healthcare services.

Fig. 1

Comorbidity prevalence (%) by medical condition index conditions (rows), comorbidities (columns). The area of each bubble is proportional to the percentage of respondents with the condition identified by the row label who also reported the comorbidity identified in the columns. The range of comorbidity prevalence runs from 47 % for high blood pressure in respondents with kidney/liver disease to 1 % for epilepsy in respondents with high blood pressure. Notes: angina = angina or long-term heart problem, arthritis = arthritis or long-term joint problem, asthma = asthma or long-term chest problem, cancer = cancer in the last five years, deaf = deafness or severe hearing impairment, hbp = high blood pressure, kidney = kidney or liver disease, back = long-term back problem, mental = long-term mental health problem, neuro = long-term neurological problem, other = another long-term problem

Fig. 2

Quality of life score (EQ-5D × 365) declines with increasing morbidity count. Unadjusted nonparametric regression of EQ-5D scores against number of long-term physical conditions, separately for each of the four subgroups formed by age (under 65 vs. older) and the presence of mental health condition (yes vs. no) categories

Fig. 3

Three condition combinations with the largest interaction effects (depicted with 95 % CI bars) on full health equivalent days per year (EQ-5D × 365), GPPS 2011–12. Note: Out of 220 triplet combinations, these are 25 whose 95 % CI did not include any of the values in the range [−11, 11]. The full list of combinations and their interaction effects is in Appendix, Table A5. Model 2 with dyads & triads versus dyads only F = 19.3 (286, 823121), p < 0.0001. For a description of the labels, see footnote to Fig. 1

Fig. 4

Adjusted mean utility scores of 55–64-year-old White woman of average level of material deprivation, by chronic condition. Based on predictions from Model 1, for a White, female, age 55–64 person of mean socioeconomic level, using a linear fixed-effects model adjusting for age, gender, ethnicity, index of multiple deprivation, recent illness or injury, and interactions of age and chronic conditions (see Appendix Tables A4 and A5 for detailed analyses of the impact of combinations of conditions on quality of life). For a description of condition labels, see footnote to Fig. 1