Summary
Systematic review of 41 studies (MEDLINE 1990–2010) on the occurrence, causes, consequences, and care of multimorbidity in the elderly. Prevalence of multimorbidity (2+ concurrent chronic diseases) ranges from 55 to 98% in persons 60+, with older age, female sex, and lower socioeconomic status (SES) as consistent risk factors. Major consequences are disability, poor quality of life, and high health care costs; the effect on mortality is controversial. The evidence base for care of multimorbid patients is essentially absent — clinical guidelines are built for single diseases and may be harmful when applied simultaneously.
Key Claims
- Comorbidity vs. multimorbidity: Comorbidity (Feinstein 1970) designates additional diseases relative to an index condition; multimorbidity removes the index condition, refocusing on the whole person. Three operational definitions exist: (1) simple count (2+ or 3+ chronic diseases), (2) weighted severity indices (Charlson Comorbidity Index, Index of Coexisting Disease [ICED], Cumulative Illness Rating Scale [CIRS]), and (3) diseases plus functional/cognitive limitations.
- Prevalence (elderly): 55–98% in persons 60+ across studies; 20–30% across all ages. Variability is driven by thresholds (2 vs. 3 vs. 4+ diseases), number of conditions evaluated, age structure, setting (community vs. institution), and ascertainment method (self-report, general practitioner [GP] records, clinical examination).
- Risk factors: Consistently: older age, female sex, lower education, lower SES. Protective: large social network. No study evaluated genetic background, biological causes, lifestyle (smoking, diet, physical activity), or environmental factors.
- Functional decline: Near-unanimous — more diseases → greater risk of disability and 3–4 year functional decline. One exception (Hudon 2008 found no association with physical activity level).
- Mortality — controversial: Three studies find more diseases → more mortality (Menotti 2001 — 10-year mortality in 65–84 year-old men; Deeg 2002 — 3-year mortality; Byles 2005 — 2-year mortality). Two do not (Marengoni 2009b — 3-year mortality same for 1 disease vs. 4+ diseases; Landi 2010 — multimorbidity predicted mortality only when combined with disability). Implication: disease count without severity or functional status is an inadequate mortality predictor.
- Quality of life (QoL): Consistently poor; associated with depression and distress. One exception: Fortin 2006b found no association with psychological distress when using a simple count (but found it when severity was incorporated). Physical QoL (SF-36/SF-12 [Short Form Health Survey] physical component) consistently impaired; mental component less consistently so.
- Health care utilization: All studies find increasing disease count → increasing prescriptions, referrals, hospitalizations, and expenditures. Wolff et al. 2002 (US Medicare, n=1.2M): 65% of 65+ with multimorbidity; higher utilization and preventable complications. Schneider et al. 2009 (n=2.2M): per-beneficiary expenditures rise with each additional condition.
- Care quality paradox: Min et al. (2007) found quality-of-care indicators satisfied increased from 47% (0 conditions) to 59% (5–6 conditions) — multimorbid patients may receive more intensive attention. But Boyd et al. (2005) showed adhering to disease-specific guidelines simultaneously for a typical multimorbid elderly patient could produce adverse drug–disease interactions — guidelines designed for single conditions are not safe when combined.
- Disease clustering: Beyond coincidental co-occurrence (product of individual disease probabilities), some pairs co-occur at higher-than-expected rates. Mechanisms include detection bias, shared risk factors (associative multimorbidity), and common biological substrates (inflammation, oxidative stress, metabolic syndrome). The distribution and combination of diseases in multimorbid patients remain largely unstudied.
Concepts Introduced or Extended
Entities Mentioned
Quotes
"Multimorbidity affects more than half of the elderly population with increasing prevalence in very old persons, women and people from lower social classes."
"We did not find a single study evaluating genetic background, biological causes (such as cholesterol, blood pressure, obesity), life styles (smoking, drinking, nutrition, physical activity), or environmental factors in relation to the development of multimorbidity."
"Adhering to disease-specific clinical guidelines in caring for elderly with multimorbidity may have undesirable effects such as adverse interactions between drugs and diseases."
My Take
A landmark state-of-the-field review that established multimorbidity as a distinct research domain. The most scientifically interesting finding is the controversial mortality result: disease count alone does not reliably predict mortality; disability status mediates or moderates the relationship. This implies that simple count-based multimorbidity is epidemiologically useful for prevalence but clinically insufficient for prognosis — severity and functional status need to be incorporated. The care-quality paradox (more conditions → higher quality scores) is counterintuitive and important: it suggests that the standard quality-indicator framework, built around process compliance with disease-specific guidelines, may be measuring the wrong thing for the most clinically complex patients. The review's explicit exclusion of severity-weighted definitions means all its prevalence and consequence estimates rest on counts, limiting clinical translation. The near-complete absence of evidence on risk factors for multimorbidity — despite its ubiquity — is a striking gap; it reflects the field's youth in 2011 and was a call to action that motivated subsequent work on biological mechanisms.