SciELO - Scientific Electronic Library Online

vol.21 suppl.2Non-Japanese, Japanese and Japanese descendant older adults in the Health, Wellbeing and Aging Study: functional and health conditionsSelf-medication practice trend among the Brazilian elderly between 2006 and 2010: SABE Study author indexsubject indexarticles search
Home Pagealphabetic serial listing  

Services on Demand




Related links


Revista Brasileira de Epidemiologia

Print version ISSN 1415-790XOn-line version ISSN 1980-5497

Rev. bras. epidemiol. vol.21  supl.2 São Paulo  2018  Epub Feb 04, 2019 


Survival of the elderly and exposition to polypharmacy in the city of São Paulo, Brazil: SABE Study

Nicolina Silvana Romano-LieberI 

Ligiana Pires CoronaII 

Liete Fatima Gouveia MarquesI 

Silvia Regina SecoliIII 

IDepartment of Public Health Practice, School of Public Health, Universidade de São Paulo - São Paulo (SP), Brazil.

IIDepartment of Epidemiology, School of Public Health, Universidade de São Paulo - São Paulo (SP), Brazil.

IIIDepartment of Medical-Surgical Nursing, School of Nursing, Universidade de São Paulo - São Paulo (SP), Brazil.



The use of polypharmacy may be due to the concomitant presence of chronic conditions, medical care by several doctors simultaneously and self-medication. Combined with the vulnerability of the elderly to the effects of drugs due to pharmacokinetic and pharmacodynamic changes, polypharmacy makes this population more susceptible to adverse outcomes. In Brazil, studies show that polypharmacy is a common problem among elderly people. However, few information is available on the association between polypharmacy and mortality.


It was assessed the survival of the elderly from São Paulo city exposed to the use of polypharmacy (five or more medications).


That was a population-based cohort, the Health, Well-Being and Aging Study (SABE Study), conducted from 2006 to 2010. The sample was composed of 1,258 individuals aged 60 years or more. The Kaplan-Meier method and Cox proportional risks model were used to examine the association between polypharmacy and mortality.


The probability of survival after five years of the users of polypharmacy at baseline was 77.2%, while among the non-users was 85.5%. Polypharmacy remained as a risk factor for death even after adjustment in other conditions associated with mortality, such as age, gender, income, chronic diseases and hospitalization.


The results point polypharmacy as an indicator of mortality in elderly people. The use of multiple medications by the elderly should be carefully assessed to avoid or minimize the damage to this population.

Keywords: Polypharmacy; Aged; Pharmacoepidemiology; Cohort studies; Mortality


The improper use of pharmacotherapy in the elderly has been the subject of several studies in which various types of problems are addressed, such as the ingestion of potentially inappropriate medications (PIM), the non-use of specified drugs, the use of excessive doses, or treatments lasting longer than necessary, as well as polypharmacy 1,2.

Several factors contribute to the occurrence of polypharmacy in the elderly, such as the concomitant presence of chronic conditions, medical care by several doctors simultaneously, and self-medication3. Combined with the increased vulnerability of the elderly to the effects of the drugs due to pharmacokinetic and pharmacodynamic changes caused by aging, polypharmacy makes this population more susceptible to adverse outcomes. Studies have shown the association between polypharmacy and the occurrence of adverse drug reactions, drug interactions, intoxication, use of PIM, and lack of adherence to treatment, resulting in increased health-care costs due to the need for hospitalization or emergency or outpatient care services3,4,5,6.

Additionally, polypharmacy has been a significant predictor, in statistical terms, to other outcomes of large magnitude such as hospitalizations, admissions to long-term institutions, hypoglycemia, fractures, reduced mobility, pneumonia, and poor nutrition5. Particularly in relation to elderly living in communities and institutionalized, polypharmacy is associated with higher mortality risk1,2,7,8,9,10.

In Brazil, although previous studies have shown that polypharmacy is a common problem among the elderly11,12,13,14,15, there is not much information about its association with mortality. This population-based study aimed at analyzing the survival of elderly people previously exposed to polypharmacy.



This study is part of the Health, Well-Being and Aging Study (SABE Study). The complete methodology is available in the first article of this supplement.

This study used as baseline the second follow-up visit of the cohort, i.e., 1,413 elderly aged 60 or over in 2006. Because it is an investigation on the consumption of medication, all the elderly included used some kind of drug on a regular basis in 2006. Thus, after the exclusion of 155 participants who did not use drugs regularly that year, the final sample of this analysis was of 1,258 elderly people. Figure 1 summarizes the composition of the study population.

Figure 1. Composition of the sample in the follow-up period of the SABE Study (2006-2010). 

Data were obtained through home interviews conducted by trained interviewers. Thequestionnaire used consisted of sections relating to living conditions and health status of the elderly. Physical data and anthropometric variables were collected by trained nutritionists, using duly calibrated equipment16.

The SABE Study was approved by the Research Ethics Committee of the School of Public Health of Universidade de São Paulo. All participants signed an Informed Consent (IC). There are no conflicts of interest.


The use of drugs was determined by the following questions: “Can you show me the medicines that are currently using or taking” and “Can you tell me the name of the medicines that are using or taking?”. The respondents were asked to show the packaging of medicines and their prescriptions, if any. The independent variable of interest was polypharmacy, defined as the use of five or more drugs.

The analyses were conducted using the independent sociodemographic and health variables. Demographic variables included were gender, age, education (years of schooling completed), per capita income (categorized into distribution thirds), and living arrangement (living alone or living with someone). Health variables that are part of the issue were primary health services provider (public or private), self-rated health (categorized as very good/good, regular, and bad/very bad), and the presence of self-reported chronic diseases (hypertension, diabetes, cardiovascular disease, cerebrovascular disease, osteoporosis and osteoarticular disease, and the total number of self-reported chronic diseases). The occurrence of hospital admissions in the last year and the score on the modified version of the Mini Mental State Examination (MMSE), validated for the SABE Study, were also included. This scale has 13 items (maximum score of 19 points) that are less dependent on education, and the cutoff point used for positive screening is 12 or less.


For descriptive analysis, mean values and standard error were calculated for continuous variables, and the proportions for categorical variables were also calculated. Differences between groups were estimated by the Wald test for average equality and the Rao-Scott test, considering sample weights to the population estimates with population weights.

For survival analysis, all deaths in the period were seen as failure. The observation time used to estimate survival functions was the time interval between the 2006 interview and the 2010/2011 interviews for those who were monitored until the end of the period. In case of death, the observation time was the time interval between the 2006 interview and the date of death. For those lost during follow-up (refusals, institutionalized, and not found), the observation time was half the time between the 2006 interview and the average period of the 2010/2011 interviews.

The survival functions were calculated using the Kaplan-Meier method, in which curves were estimated by grouping the elderly according to the variables selected for the study. TheLog-rank test was used to compare the survival functions for each variable.

In order to assess the risk factors associated with death, the hazard ratios (HR) were calculated with 95% confidence intervals, according to the Cox proportional hazards model, including the independent variable of interest - polypharmacy (group reference: elderly who consumed one to four drugs) - and other sociodemographic and health variables mentioned above. The proportionality of the Cox model was verified based on the Schoenfeld residuals diagnostic test and through visual inspection of the curves ln(-ln(S(t))) × follow-up time.

The variables with p < 0.20 in the univariate Cox analysis for association with mortality were included in the multiple Cox model as covariates. After the estimation of the final model, the adjusted survival function was drawn according to the occurrence of polypharmacy.

The analyses were performed using the Stata® software, version 11 (ST module), taking into account the sample weights and inferences considering the effect of the design.


In 2006, in the baseline study, polypharmacy was found in 33% of the elderly, being more frequent among women than among men: 36.6 and 26.9%, respectively (p = 0.010). Polypharmacy was also more common in elderly aged over 75 years (41.3%) than in elderly aged 60-74 years (30.1%, p = 0.002) and in subjects with chronic diseases such as hypertension, diabetes, cardiovascular disease, cerebrovascular disease, and articular disease (p < 0.001). The average number of prevalent chronic diseases in the elderly who had polypharmacy at baseline was 2.4, while in the elderly who consumed between one and four drugs, the average was 1.4 (p < 0.001).

Table 1 shows the characteristics of the elderly according to their follow-up status at the end of the period between 2010 and 2011. The elderly who died in the meantime showed higher frequency of polypharmacy than the elderly group that was monitored up to the end the study (p < 0.001). Deaths were also more common in elderly men, older and presenting the chronic diseases mentioned. The occurrence of hospitalization in the year prior to the baseline interview was also higher in the group of patients who died.

Table 1. Distribution of the elderly according to sociodemographic and health variables at baseline and status at follow-up. SABE Study, São Paulo (2006-2010). 

Variables Elderly individuals in follow-up (%) Death (%) p-value
Number of medications used < 0.001
1 - 4 68.0 53.1
5 and more 32.0 46.9
Sex < 0.001
Male 33.9 48.1
Female 66.1 51.9
Age in years < 0.001
60 - 74 80.4 39.3
75 or more 19.6 60.7
Education (years of study) 0.003
8 and over 15.3 25.0
4 - 7 27.1 19.5
1 - 3 38.2 41.4
None 19.4 14.1
Per capita income 0.079
1.º Tertile 30.2 37.7
2.º Tertile 33.2 35.0
3.º Tertile 36.6 27.3
Living arrangements
Lives alone 12.6 14.3 0.591
Lives with company 87.4 85.7
Primary health services provider 0.499
Public 53.5 50.7
Private 46.5 49.3
Self-rated health < 0.001
Very good/good 43.1 25.9
Regular 48.3 54.5
Bad/very bad 8.7 19.7
Hypertension 69.5 64.9 0.333
Diabetes 22.5 30.6 0.04
Cardiovascular disease 24.1 36.2 0.003
Cerebrovascular disease 7.1 16.4 < 0.001
Osteoarticular disease 37.9 29.9 0.090
Number of chronic diseases - mean (SE) 1.7 (0.05) 2.0 (0.09) 0.018
Hospitalization in the last year < 0.001
None 91.3 74.9
One or more 8.7 25.1
Score on the adapted MMSE* - mean (SE) 16.4 (0.15) 12.4 (0.53) < 0.001

SE: mean, standard error; MMSE: Mini Mental State Examination.

The probability of survival after 5 years for patients with polypharmacy at baseline was 77.2%, while in individuals taking up to four medications, it was 85.5%. The survival curves during follow-up showed a significant difference (p < 0.001 in the Log-rank test).

The crude and adjusted Cox proportional hazards regression models are shown in Table2. Polypharmacy remained a risk factor for death even after adjusting for other conditions associated with mortality, such as age, gender, income, chronic diseases, and hospitalization, with a 57% higher HR in comparison to those who consumed up to four drugs. The survival curve, adjusted after the final model (Figure 2), shows that the probability of death is greater (and proportional) in elderly patients with polypharmacy throughout the follow-up period, more markedly from the second year on.

Table 2. Cox proportional hazards models for mortality in the period. SABE Study, São Paulo (2006-2010). 

Variables Crude analysis Adjusted analysis
HR (95%CI)** p-value HR (95%CI)** p-value
Polypharmacy* 1.84 (1.33-2.54) < 0.001 1.57 (1.05-2.34) 0.026
Female 0.60 (0.43-0.83) 0.002 0.47 (0.33-0.66) < 0.001
Age (years) 1.11 (1.09-1.13) < 0.001 1.09 (1.06-1.11) < 0.001
Per capita income
1.º Tertile 1.00 1.00
2.º Tertile 0.85 (0.58-1.25) 0.416 0.86 (0.57-1.28) 0.449
3.º Tertile 0.62 (0.41-0.96) 0.031 0.72 (0.45-1.67) 0.186
Diabetes 1.48 (1.03-2.11) 0.034 1.58 (0.95-2.63) 0.077
Cardiovascular disease 1.71 (1.22-2.39) 0.002 1.25 (0.80-1.94) 0.326
Cerebrovascular disease 2.19 (1.42-3.34) < 0.001 1.01 (0.63-1.62) 0.966
Number of chronic diseases 1.22 (1.04-1.43) 0.014 0.87 (0.68-1.11) 0.259
Hospitalization in the last year 3.21 (2.2-4.70) < 0.001 1.68 (1.10-2.57) 0.015
Score on the adapted MMSE 0.86 (0.84-0.89) <0.001 0.92 (0.90-0.95) <0.001

*Five drugs or more; **hazard ratio (95% confidence interval); Schoenfeld residuals diagnostic test: p = 0.117.

Figure 2. Adjusted survival* in the period according to the number of drugs used. SABE Study, São Paulo (2006-2010). 


Polypharmacy proved to be an important risk factor for deaths in the 4-year follow-up period in elderly people, regardless of other factors associated with mortality such as age, gender, income, chronic diseases, and hospitalization. The HR was 57% higher compared to those who took up to four drugs.

Population-based studies conducted previously also found association between polypharmacy and mortality, although the comparison of the findings is hampered by methodological aspects. The definitions of polypharmacy, the ages of the elderly at baseline, the follow-up time, and the adjustments vary widely.

A cohort study in Finland assessed the association between polypharmacy and mortality among the elderly aged over 75 years, in a follow-up period divided into two 5-year phases, the first with an average follow-up of 3.62 years, and the second, 3.79 years. The authors considered polypharmacy as using six to nine drugs and excessive polypharmacy as ten ormore drugs. The members of the group not exposed to polypharmacy, i.e., using fiveor less drugs, showed a higher survival rate. The mortality rate in the group with excessive polypharmacy was higher (55% in the first phase and 61% in the second phase) than in the group with polypharmacy (33 and 40%, respectively), while the group without polypharmacy has 27 and 23%, respectively. In the second phase of the study, there was an association between excessive polypharmacy and mortality (HR = 2.23) in the multivariate model, after adjusting for demographic variables and functional and cognitive status. However, there was no adjustment for comorbidity8.

A study with Mexican-American elderly individuals aged 65-99 years, being monitored for 8 years, considered polypharmacy as the use of more than 4 drugs and found an association with mortality after adjustments, including comorbidities. The risk of mortality associated with polypharmacy was 27% higher among individuals exposed to polypharmacy2.

In Australia, a cohort of about 4,000 men aged between 65 and 83 years was monitored for 4½ years. Polypharmacy, understood as the intake of five or more drugs, was used as an indicator of suboptimal drug use. Research has shown that the number of drugs used was independently associated with mortality (HR = 1.04), after the adjustment of comorbidities1.

An 18-year cohort held with individuals aged over 65 years living in the UK demonstrated association between polypharmacy (consumption of five or more medications) and short-term mortality (the first 2 years) for female and male elderly individuals. This association remained, although attenuated, in the medium and long term for women, but became insignificant for men in the long term. The HR of the total period was 1.42 for men and 1.30 for women, after adjustments, including comorbidities7.

Also in the UK, a follow-up period of 1 year with individuals aged 65 years or older living in the community or institutionalized observed positive association of mortality with the use of three drugs or more-the association increased with the number of drugs used. In comparison with the use of 0-2 drugs by the elderly in the community and the institutionalized elderly, respectively, the HR was 1.12 and 1.16 for the use of 3-5 drugs; HR = 1.48 and 1.34, respectively, for the use of 6-10 drugs; and HR = 2.32 and 1.59, respectively, for the use of 11 or more medications, after adjustments, including comorbidity9.

An investigation carried out in eight European countries has shown that polypharmacy, considered as the use of ten drugs or more, is associated with increased mortality in institutionalized patients with cognitive impairment at the end of life, after a follow-up period of 1 year10.

Polypharmacy has been linked to several negative consequences, particularly high costs with health, increased risk of adverse drug events, drug interactions, non-adherence to drug treatment, reduced functional capacity, and multiple geriatric syndromes17. In this sense, polypharmacy (six or more drugs) proved to be an important predictor of mortality in elderly Japanese patients (aged over 85 years) monitored for 1 year after discharge, even after adjusting for comorbidities. It has been argued that, since it is well established that polypharmacy increases the possibility of adverse drug events, such events might be involved in the mortality of the elderly subjects, since the risk and severity of events are directly proportional to the amount of drugs used and age18.

In addition, studies show that elderly individuals who receive five or more drugs are three times more likely to get a PIM than those who use fewer drugs19. Also, adverse drug events resulting from inappropriate drug prescriptions can lead to increased hospitalizations and mortality20.

Polypharmacy could also have a direct effect on mortality through the cumulative effect of multiple drugs in the elderly’s kidney and liver systems, initiating a cascade of interactions in these individuals, who already suffer from multimorbidity2.

This study’s main limitation is the large number of losses in the follow-up period. Polypharmacy was less frequent in this group (26.9%), and, if the losses were considered, estimates exposed here could be changed. However, cohort studies show this intrinsic limitation, especially those made with individuals living in communities, who often move or refuse to continue the follow-up. In the case of the elderly, there is a portion that is institutionalized or dies without the knowledge of researchers.

On the other hand, the strength of this study was that it contains a large representative sample of the population studied and is the only Brazilian study to evaluate the survival of elderly with polypharmacy. Still, as already mentioned, it was possible to adjust the analyses for several of the most prevalent diseases among the elderly, as well as the average number of chronic diseases, eliminating them as confounding factors and emphasizing the independent effect of polypharmacy as a risk factor.


In this study, polypharmacy was independently associated with mortality. Thus, the use of multiple medications should be carefully monitored, and especially in the older, male population, with lower scores on the adapted MMSE and with a history of hospitalization.

Seeking a balance between the benefits and risks of drug therapy is a challenge. Polypharmacy can be considered as iatrogenic when adverse events caused by drugs are not recognized as such, resulting in new prescriptions. Polypharmacy can also result from the unnecessary use through repeated prescriptions and lack of monitoring8. Reducingthe number of prescription drugs, therefore, requires a multidisciplinary approach21.

Thus, the results of this study reinforce the need to develop strategies to avoid or minimize negative outcomes related to polypharmacy, contributing to the security of the elderly in medication use.


1. Beer C, Hyde Z, Almeida OP, Norman P, Hankey GJ, Yeap BB, et al. Quality use of medicines and health outcomes among a cohort of community dwelling older men: an observational study. Br J Clin Pharmacol 2011; 71(4): 592-9. [ Links ]

2. Espino DV, Bazaldua OV, Palmer RF, Mouton CP, Parchman ML, Miles TP, et al. Suboptimal medication use and mortality in an older adult community-based cohort: results from the Hispanic EPESE Study. J Gerontol A Biol Sci Med Sci 2006; 61(2): 170-5. [ Links ]

3. Heppner HJ, Christ M, Gosch M, Mühlberg W, Bahrmann P, Bertsch T, et al. Polypharmacy in the elderly from the clinical toxicologist perspective. Z Gerontol Geriat 2012; 45: 473-8. [ Links ]

4. Fulton MM, Allen ER. Polypharmacy in the elderly: a literature review. J Am Acad Nurse Pract 2005; 17(4): 123-32. [ Links ]

5. Frazier SC. Health outcomes and polypharmacy in elderly individuals: an integrated literature review. J Gerontol Nurs 2005; 31(9): 4-11. [ Links ]

6. Shah BM, Hajjar ER. Polypharmacy, adverse drug reactions, and geriatric syndromes. Clin Geriatr Med 2012; 28: 173-86. [ Links ]

7. Richardson K, Ananou A, Lafortune L, Brayne C, Matthew FE. Variation over time in the association between polypharmacy and mortality in the older population. Drugs Aging 2011; 28(7): 547-60. [ Links ]

8. Jyrkkà J, Enlunà H, Korhonen MJ, Sulkava R, Hartikainen S. Polypharmacy status as an indicator of mortality in an elderly population. Drugs Aging 2009; 26(12): 1039-48. [ Links ]

9. Shah SM, Carey IM, Harris T, De Wilde S, Cook DG. Mortality in older care home residents in England and Wales. Age Ageing 2013; 42: 209-15. [ Links ]

10. Onder G, Liperoti R, Foebel A, Fialova D, Topinkova E, van der Roest HG, et al. Polypharmacy and mortality among nursing home residents with advanced cognitive impairment: results from the Shelter Study. JAMDA 2013; 14: 450e7-12. [ Links ]

11. Carvalho MFC, Romano-Lieber NS, Bergsten-Mendes G, Secoli SR, Ribeiro E, Lebrão ML, et al. Polifarmácia entre idosos do município de São Paulo - Estudo Sabe. Rev Bras Epidemiol 2012; 15(4): 817-27. [ Links ]

12. Galato D, Silva ES, Tiburcio LS. Estudo de utilização de medicamentos em idosos residentes em uma cidade do sul de Santa Catarina (Brasil): um olhar sobre a polimedicação. Ciênc Saúde Colet 2010; 15(6): 2899-905. [ Links ]

13. Loyola Filho AI, Uchoa E, Lima-Costa MF. Estudo epidemiológico de base populacional sobre uso de medicamentos entre idosos na região metropolitana de Belo Horizonte, Minas Gerais, Brasil. Cad Saúde Públ 2006; 22(12): 2657-67. [ Links ]

14. Lucchetti G, Granero AL, Pires SL, Gorzoni ML. Fatores associados à polifarmácia em idosos institucionalizados. Rev Bras Geriatr Gerontol 2010; 13(1): 51-8. [ Links ]

15. Rozenfeld S, Fonseca MJM, Acurcio FA. Drug utilization and polypharmacy among the elderly: a survey in Rio de Janeiro City, Brazil. Rev Panam Salud Publ 2008; 23(1): 34-43. [ Links ]

16. Lebrão ML, Duarte YAO. Sabe - Saúde, Bem-Estar e envelhecimento - O projeto Sabe no município de São Paulo: uma abordagem inicial [Internet]. Brasília: Athalaia Bureau; 2003 [cited 2006 Feb 20]. Available from: ]

17. Maher Jr RL, Hanlon JT, Hajjar ER. Clinical consequences of polypharmacy in elderly. Expert Opin Drug Saf 2014; 13(1): 1-11. [ Links ]

18. Iwata M, Kuzuya M, Kitagawa Y, Suzuki Y, Iguchi A. Underappreciated predictors for postdischarge mortality in acute hospitalized oldest-old patients. Gerontology 2006; 52: 92-8. [ Links ]

19. Gallagher PF, Barry PJ, Ryan C, Hartigan I, O'Mahony D. Inappropriate prescribing in an acutely ill population of elderly patients as determined by Beers' criteria. Age Ageing 2008; 37: 96-101. [ Links ]

20. Lau DT, Kasper JD, Potter DEB, Lyles A, Bennett RG. Hospitalization and death associated with potentially inappropriate medication prescriptions among elderly nursing home residents. Arch Intern Med 2005; 165: 68-74. [ Links ]

21. Medeiros-Souza P, Santos-Neto LL, Kusano LTE, Pereira MG. Diagnóstico e controle da polifarmácia no idoso. Rev Saúde Públ 2007; 41(6): 1049-53. [ Links ]

Financial support: Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP), protocol no. 2005/54947-2 and no. 2009/53778-3.

Received: July 01, 2014; Revised: August 20, 2014; Accepted: October 15, 2014

Corresponding author: Nicolina Silvana Romano-Lieber. Departamento de Prática de Saúde Pública, Faculdade de Saúde Pública da Universidade de São Paulo. Avenida Dr. Arnaldo, 715, CEP: 01246-904, São Paulo, SP, Brasil.E-mail:

Conflict of interests: nothing to declare

Creative Commons License Este é um artigo publicado em acesso aberto sob uma licença Creative Commons