Associations between Polypharmacy and Depressive Symptoms in Patients with Chronic Diseases: A Cross-Sectional and Prospective Study in Wuhan, Hubei Province

doi:10.21203/rs.3.rs-4018947/v1

Background: Exploring the relationship between polypharmacy behaviors and depressive symptoms in patients with chronic diseases is crucial for public health. The purpose of this study was to explore the relationship between polypharmacy behaviors and depression in patients with chronic diseases.

Methods: A cluster sampling survey of patients with chronic diseases was initially conducted in March-April 2019 in 8 administrative districts of Wuhan. Depressive symptoms were measured using the Center for Epidemiological Studies-Depression Scale (CES-D10). To reduce confounding bias, the propensity score weighting method was applied, then a multivariate logistic regression model was conducted to test the relationship between polypharmacy behaviors and depressive symptoms.

Results: Inpatients with polypharmacy behaviors exhibited a higher risk of depressive symptoms (OR=1.58, 95%CI: 1.17-2.16) compared to individuals without polypharmacy behaviors. In addition, higher education level (OR=0.51, 95%CI: 0.30-0.87),physical exercise (OR=0.51, 95%CI: 0.34-0.76) and usage of Chinese medicine for treatment were associated with a lower risk of depressive symptoms. It also has been found that depression was linked to hypertension (OR=1.47, 95%CI: 1.00-2.17), poor (OR=5.15, 95%CI: 3.43-7.84) and fair (OR=2.48, 95%CI: 1.61-3.86) self-rated health status, as well as have more chronic diseases (OR=1.27, 95%CI: 1.11-1.47).

Conclusion: The occurrence of polypharmacy behaviors was significantly related to the patient's depressive symptoms. Furthermore, the patient's health information and educational level were related to depression. For chronic diseases patients who have poor mental health, interventions should be taken to reduce their inappropriate polypharmacy behaviors.

polypharmacy

depression

chronic disease

propensity score weighting

According to the report of the World Health Organization, depression, as a psychiatric disorder, is widely spreading all over the world^[1]. The prevalence of depression worldwide was estimated at 4.4% in 2015, which means more than 300 million people were affected^[2]. Characterized by symptoms of persistent feelings of hopelessness and dejection, depression is the most disabling disease in the world and the fourth leading cause of total disease burden, accounting for 40.5% of total disability-adjusted life-years caused by mental and substance use disorders^[3]. Depressive disorders often start at a young age, and they reduce people’s functioning and are often recurrent^[4]. There are also evidences that individuals with depressive symptoms can be predisposed to diseases such as HIV infection, diabetes, and even death from suicide^[5–7].

Polypharmacy behaviors, defined in various ways, including but not limited to “unnecessary drug use”, “medication use without indication” and “use five or more medications”^[8–11], along with depression were known to be associated with disability and both were defined as a geriatric syndrome^[12]. Particularly common in people with multiple chronic diseases, polypharmacy behaviors were associated with various adverse outcomes, including falls, higher mortality rate, higher hospital admissions and readmission rate, longer hospital stay^[13], and was also observed to have an association with higher levels of psychological distress^[14].

Meanwhile, a series of risk factors for depression have been found in previous studies, such as, low education level, low income, live alone rarely physical activity, drinking, and smoking^[15]. Besides, there were other risk factors that contributed to the development of chronic depression in patients diagnosed with a depressive episode, such as substance abuse^[16]. There appears to be little research on the relationship between polypharmacy behaviors and depression in patients with chronic diseases, especially in hospitalized patients, despite the fact that the relationship between multiple factors and depression has been discussed in many studies. Most of the existing studies only consider the association between polypharmacy behaviors and depression in the study design, without discussing it^[17–19]. There were also findings indicating that the amount of medication taken and emotional impairment in older adults might be influenced by the same factors^[20], which suggested that poorer emotional capability, such as depression and anxiety, might be associated with increased medication use. Polypharmacy behaviors might play a crucial role in the development of depressive symptoms in adults with chronic diseases, but its relationship with depression has been underestimated.

Due to the different methods and populations selected in the study, most of the researches were unable to draw a clear relationship between polypharmacy behavior and depression. Therefore, this study aims to examine the associations between polypharmacy behaviors and depressive symptoms in adults with chronic diseases, clarify the influence of polypharmacy behaviors on depression, and pay attention to the emotional problems in hospitalized patients with polypharmacy behaviors.

2.1 Study participants and procedure

The data of this cross-sectional study was collected by a cluster sampling method from March to April 2019, We surveyed patients with chronic diseases in the community as well as inpatients with chronic diseases separately.

For community chronic disease patients, we selected Wuhan and Yichang cities as urban sample areas, and Qianjiang and Zhijiang cities as rural sample areas for the study, and randomly selected 3 streets or townships in each of the sample areas as research sites (12 in total), and took communities and villages as units, and pulled out from the health information system all their chronic disease patients' personal basic information and their drug purchases, and patients with chronic diseases who were prescribed five or more types of drugs in the last three months were taken as survey subjects until 100 people were completed in each street or township. For inpatient chronic disease patients, 8 general hospitals were randomly selected from 8 administrative districts (including Jianghan District, Jiangan District, Qiaokou District, Hongshan District, Wuchang District, Hanyang District, Caidian District and Jiangxia District) in Wuhan, Hubei Province, China, all eligible patients in the cardiology and endocrinology departments of each hospital were selected.

Exclusion criteria included those who had not been taking medicine for more than 3 months, those who were younger than 18 years, those who couldn’t express themselves clearly, those who were unwilling to cooperate with this investigation, those who were severely ill and could not complete the questionnaire, and those diagnosed with acute. A total of 1,976 questionnaires were collected, 62 patients were excluded due to poor memory or other personal reasons.

2.2 Measure

The primary predictor of interest in this study was whether a patient reported having polypharmacy behaviors. In this study, polypharmacy behavior was measured by asking the patients, “In the last three months, how many kinds of drugs have you taken to treat your chronic diseases? (excluding Traditional Chinese medicine)”. Respondents who took 5–9 drugs were considered as general polypharmacy behaviors, patients who took 10 or more drugs were considered as severe polypharmacy behaviors, and the others were regarded as “non-polypharmacy”. Confounding variables included demographic characteristics, health status, medication information, and so on. And the dependent variable was whether a patient had depression.

2.2.1 Demographic characteristics

Demographic characteristics, include the patient's age, domicile, gender, education level, annual income, medical insurance, and residence status. Residence status was classified into “living alone” and “not living alone”.

2.2.2 Health status

Health status questions include the physical exercise, disease type, the number of diseases, self-rated health status. The physical exercise was measured by asking the patients, “1. Do you take time to exercise?”. The answer is “0 = No, 1 = Yes”. The disease type was measured by asking the patients,“2. Do you suffer from any of the following diseases?”. This is a multiple choice. The answer options include: hypertension, diabetes, etc. 3. The number of diseases is calculated by the investigator after the patient answered all the types of diseases. The last question is “4. How do you feel about your health?”, the answer is: “1 = good, 2 = fair, 3 = bad”.

2.2.3 Medication information

Medication information, include the drug history, whether to take traditional Chinese medicine treatment, and adverse drug reactions (ADRs). The drug history was measured by asking the patients, “In which year did you start taking the drugs for chronic diseases?”

2.2.4 Dependent variable

The dependent variable is whether a patient had depression symptom. Depressive symptoms were measured by The Center for Epidemiological Studies-Depression Scale (CES-D10). These 10 items represent the depressive feelings and behaviors of the respondents during the last week. The answers for CES-D10 are on a four-scale metrics coding from 0 to 3 (0 = less than 1 day; 1 = 1 to 2 days; 2 = 3 to 4 days; 3 = 5 to 7 days).Thus, the total score of the scale ranged from 0 to 30, with higher scores indicating more depressive symptoms. The CES-D10 has been used in previous studies and has shown good internal reliability^[21–23]. Several studies have reported a cut-off point of 10 with good validity to identify clinically significant depression. Accordingly, we use the score of 10 as the cut-off point to describe the prevalence of depression^[24].

2.3 Patient and public involvement

None.

2.4 Statistical analysis

The statistical software SPSS 24.0 and R 4.0.4 were used to conduct data analysis. Chi-square test or T test was used to check for any association between depression and various explanatory variables. Variables with statistical significance in the previous chi-square analysis (p < 0.05) were included in the multivariate logistic regression model to estimate their effects on depression. Propensity score weighting approach was applied to adjust for the observed difference in characteristics between those with polypharmacy behaviors (including general polypharmacy behaviors and severe polypharmacy behaviors) behavior and those without, which therefore could better tease out the net influence of having polypharmacy behavior on patients’ depression. Odds ratios (ORs), Marginal Effects, and their 95% confidence intervals were reported to indicate the relationship between two variables. In all analyses, statistical significance was defined as a two-tailed p-value < 0.05.

3.1 Interviewee information

Overall, we surveyed 718 inpatients, with an average age of 74.73 years, of whom 144 (20.06%) were registered in rural areas, and 415 (57.8%) were female. Table 1 lists the demographic characteristics, health status, and medication information of the inpatients. There were 523 (72.8%) patients with hypertension, 293 (40.8%) patients with diabetes, and 612 (85.2%) patients with other chronic diseases. There were 2.88 kinds of chronic diseases and 13.95 years of drug history per capita, and the average number of drugs used was 5.87. There were 291 with depressive symptoms (40.5%) and 427 with no depressive symptoms (59.5%). In general, education level, exercise, suffering from other chronic diseases (excluding hypertension and diabetes), the number of diseases, self-rated health status, whether adverse drug reactions occurred, and multiple medications were associated with depression.

Table 1

Characteristics of inparticipants
Variables	Depression				P
	No		Yes
	Frequency/Mean (n = 427)	Weighted percentage/SD	Frequency/Mean (n = 291)	Weighted percentage/SD
Polypharmacy
No	203	47.5%	92	31.6%	< 0.001
General polypharmacy behaviros	196	45.9%	160	55.0%
Severe polypharmacy behaviros	28	6.6%	39	13.4%
Demographic characteristics
Age(years)
< 65	111	26.0%	74	25.4%	0.865
≥ 65	316	74.0%	217	74.6%
Domicile
Urban	350	82.0%	224	77.0%	0.101
Rural	77	18.0%	67	23.0%
Gender
Male	190	44.5%	113	38.8%	0.131
Female	237	55.5%	178	61.2%
Level of education
Primary school	106	24.8%	101	34.7%	0.015
Junior high school	129	30.2%	79	27.1%
High school	113	26.5%	75	25.8%
University	79	18.5%	36	12.4%
Annual income/yuan
≤ 36000	146	34.2%	113	38.8%	0.421
36001–69000	128	30.0%	84	28.9%
≥ 69001	153	35.8%	94	32.3%
Medical insurance
Employee health insurance	292	68.4%	184	63.2%	0.316
Medical Insurance for Urban and Rural	109	25.5%	89	30.6%
No medical insurance	26	6.1%	18	6.2%
Residence status
Not live alone	345	80.8%	223	76.6%	0.178
Live alone	82	19.2%	68	23.4%
Health status
Exercise
No	332	77.8%	248	85.2%	0.013
Yes	95	22.2%	43	14.8%
Hypertension
No	85	29.2%	110	25.8%	0.308
Yes	206	70.8%	317	74.2%
Diabetes
No	258	60.4%	167	57.4%	0.417
Yes	169	39.6%	124	42.6%
Other chronic diseases
No	86	20.1%	20	6.9%	< 0.001
Yes	341	79.9%	271	93.1%
Number of diseases	2.597	1.43	3.299	1.57	< 0.001
Self-rated health status
Good	173	40.5%	40	13.7%	< 0.001
Fair	132	30.9%	83	28.5%
Poor	122	28.6%	168	57.7%
Medication information
Drug history/year	13.681	12.28	14.351	11.49	0.462
Adverse drug reaction
No	360	84.3%	222	76.3%	0.007
Yes	67	15.7%	69	23.7%
Took Chinese medicine
No	270	63.2%	191	65.6%	0.509
Yes	157	36.8%	100	34.4%

We surveyed 1196 community patients with a mean age of 68.75 years, of whom 618 (51.7%) were rural household origin and 693 (57.9%) were female. Demographic characteristics, health status, medication information, and medical behaviors of community patients are presented in Table 2. 324 had depressive symptoms (27.1%) and 872 had no depressive symptoms (72.9%). Overall, household registration, gender, education, type of health insurance, annual income, exercise, medication adherence, adverse drug reactions, hospitalization in the past year, trust in physicians, severity of illness, type of decision making, number of illnesses, medication knowledge, and multiple medication using were associated with depressive symptoms.

Table 2

Characteristics of community participants
Variables	Depression				P
	No		Yes
	Frequency/Mean (n = 872)	Weighted percentage/SD	Frequency/Mean (n = 324)	Weighted percentage/SD
Polypharmacy
No	709	81.3%	235	72.5%	0.001
Yes	163	18.7%	89	27.5%
Demographic characteristics
Age(years)
< 65	249	28.6%	98	30.2%	0.841
≥ 65	463	53.1%	169	52.2%
Domicile
Urban	480	55.0%	98	30.2%	< 0.001
Rural	392	45.0%	226	69.8%
Gender
Male	387	44.4%	116	35.8%	0.008
Female	485	55.6%	208	64.2%
Level of education
Primary school	350	40.1%	193	59.6%	< 0.001
Junior high school	276	31.7%	85	26.2%
High school	186	21.3%	37	11.4%
University	60	6.9%	9	2.8%
Annual income/yuan
0-9999	206	23.6%	137	42.3%	< 0.001
10000–50000	295	33.8%	101	31.2%
≥ 50001	371	42.5%	86	26.5%
Medical insurance
Employee health insurance	405	46.4%	82	25.3%	< 0.001
Medical Insurance for Urban and Rural	467	53.6%	242	74.7%
Residence status
Not live alone	83	9.5%	41	12.7%	0.135
Live alone	789	90.5%	283	87.3%
Health status
Exercise
Never	88	10.1%	64	19.8%	< 0.001
Occasionally	145	16.6%	73	22.5%
Regularly	174	20.0%	69	21.3%
Always	465	53.3%	118	36.4%
Hypertension
No	105	12.0%	42	13.0%	0.666
Yes	767	88.0%	282	87.0%
Diabetes
No	594	38.1%	215	66.4%	0.563
Yes	278	31.9%	109	33.6%
Number of diseases
1	387	44.4%	82	25.3%	< 0.001
2	315	36.1%	91	28.1%
> 3	170	19.5%	151	46.6%
Disease severity
Light risk	393	45.1%	88	27.2%	< 0.001
Medium risk	361	41.4%	136	42.0%
High risk	118	13.5%	100	30.9%
Medication information
Adverse drug reaction
No	713	81.8%	241	74.4%	0.005
Yes	159	18.2%	83	25.6%
Medication adherence
Poor	331	38.0%	151	46.6%	0.007
Good	541	62.0%	173	53.4%
Medication knowledge
Low	405	46.4%	202	62.3%	< 0.001
High	467	53.6%	122	37.7%
Medical behavior
Whether trust the doctor
Very trust	526	60.3%	173	53.4%	0.002
More trust	250	28.7%	99	30.6%
Commonly	59	6.8%	19	5.9%
Don't trust	30	3.4%	27	8.3%
Very mistrust	07	0.8%	6	1.9%
Whether be in hospital in the past year
No	603	69.2%	161	49.7%	< 0.001
Yes	269	30.8%	163	50.3%
Regular blood pressure measurement
No	146	16.7%	65	20.1%	0.181
Yes	726	83.3%	259	49.9%
Type of decision
Personal decision-making	58	6.7%	34	10.5%	< 0.001
Shared decision-making	103	11.8%	14	4.3%
Doctor decision-making	711	81.5%	276	85.2%

3.2 Propensity score weighting (PSW) results

Standardized mean difference (SMD) is an indicator used to evaluate the balance between the experimental group and the control group before and after weighting, using a threshold of 0.10 to indicate an imbalance. Table 3 and Table 4 show the SMD values before and after the PSW for inpatients and community patients, respectively. Covariates with large deviations in the original data were balanced after PSW, and the overall distribution of the data is more balanced.

Table 3

The SMD values before and after the PSW for inpatients
Variables	Standardized mean difference
Variables	Before Weight	After Weight
Age	0.13	-0.05
Domicile	-0.18	-0.05
Gender	-0.13	0.04
Education	-0.04	0.07
Annual income	0.21	0.05
Medical insurance	-0.11	0.02
Residence status	-0.05	-0.01
Exercise	0.11	0.05
Hypertension	0.09	-0.01
Diabetes	0.29	-0.02
Other chronic diseases	0.55	-0.01
Self-rated health status	0.24	0.00
Adverse disease reaction	0.02	0.05
Took Chinese medicine	0.14	0.05

Table 4

The SMD values before and after the PSW for community patients
Variables	Standardized mean difference
Variables	Before Weight	After Weight
Age	0.16	0.05
Domicile	-0.20	-0.02
Gender	0.00	-0.03
Education	0.14	-0.05
Annual income	0.05	-0.03
Medical insurance	-0.18	0.00
Residence status	-0.08	-0.03
Exercise	-0.20	0.03
Hypertension	0.13	0.02
Diabetes	0.55	0.10
Number of diseases	1.15	0.06
Disease severity	0.70	0.14
Adverse disease reaction	-0.36	0.06
Medication adherence	-0.10	-0.05
Medication knowledge	0.03	-0.01
Whether trust the doctor	-0.03	-0.01
Whether be in hospital in the past year	-0.52	0.00
Regular blood pressure measurement	-0.16	0.04
Type of decision	0.00	-0.02

3.3 Regression analysis results

Table 5 shows the odds ratios and p value for multivariate logistic regression and PSW regression for inpatients. Due to the small number of patients with severe polypharmacy behaviors, we have combined general polypharmacy behaviors, and severe polypharmacy behaviors into polypharmacy behaviors for comparison. The results of the PSW regression are different from the results of the multivariate logistic regression. The multivariate logistic regression overestimates the impact of polypharmacy behaviors on the depression due to crowd confounding factors.

Table 5

**Multivariate logistic regression before and after the PSW for inpatients**
Variables	Before weighting OR (95%CI)	P-value	After weighting OR (95%CI)	P-value
Polypharmacy behaviros
No	0.00(ref)		0.00(ref)
Yes	1.63(1.12–2.36)	0.010	1.58(1.17–2.16)	0.003
Demographic characteristics
Age(years)
< 65	0.00(ref)		0.00(ref)
≥ 65	0.84(0.56–1.28)	0.420	1.01(0.68–1.49)	0.966
Domicile
Urban	0.00(ref)		0.00(ref)
Rural	1.34(0.85–2.12)	0.202	1.25(0.82–1.90)	0.304
Gender
Male			0.00(ref)
Female	1.04(0.73–1.49)	0.813	1.29(0.94–1.77)	0.110
Level of education
Primary school	0.00(ref)		0.00(ref)
junior high school	0.73(0.47–1.14)	0.167	0.68(0.46–1.02)	0.061
High school	0.69(0.43–1.12)	0.134	0.80(0.53–1.21)	0.293
University	0.53(0.29–0.95)	0.034	0.51(0.30–0.87)	0.013
Annual income/yuan
≤ 36000	0.00(ref)		0.00(ref)
36001–69000	1.12(0.72–1.74)	0.609	1.17(0.79–1.75)	0.440
≥ 69001	1.13(0.72–1.78)	0.590	1.42(0.95–2.14)	0.087
Medical insurance
Employee health insurance	0.00(ref)		0.00(ref)
Medical Insurance for Urban and Rural Residents	1.24(0.81–1.90)	0.315	1.14(0.77–1.68)	0.508
No medical insurance	0.78(0.37–1.63)	0.518	0.71(0.34–1.43)	0.339
Residence status
Not live alone	0.00(ref)		0.00(ref)
Live alone	1.40(0.91–2.17)	0.129	1.23(0.83–1.84)	0.302
Health status
Exercise
No	0.00(ref)		0.00(ref)
Yes	0.47(0.29–0.74)	0.001	0.51(0.34–0.76)	0.001
Hypertension
No	0.00(ref)		0.00(ref)
Yes	1.70(1.12–2.56)	0.013	1.47(1.00-2.17)	0.049
Diabetes
No	0.00(ref)		0.00(ref)
Yes	0.87(0.59–1.29)	0.497	0.96(0.68–1.35)	0.811
Other chronic diseases
No	0.00(ref)		0.00(ref)
Yes	1.61(0.84–3.15)	0.158	1.47(0.71–3.23)	0.315
Number of diseases	1.31(1.12–1.54)	< 0.001	1.27(1.11–1.47)	0.001
Self-rated health status
Good	0.00(ref)		0.00(ref)
Fair	2.82(1.78–4.55)	< 0.001	2.48(1.61–3.86)	< 0.001
Poor	5.34(3.44–8.43)	< 0.001	5.15(3.43–7.84)	< 0.001
Medication information
Drug history/year	1.00(0.98–1.01)	0.958	0.99(0.98-1.00)	0.188
Took Chinese medicine
No (Ref)			0.00(ref)
Yes	0.72(0.50–1.04)	0.081	0.66(0.48–0.91)	0.013
Adverse drug reaction
No			0.00(ref)
Yes	1.48(0.97–2.27)	0.066	1.22(0.82–1.80)	0.327
Constant	0.11	< 0.001	0.12	< 0.001

Patients have polypharmacy behaviors were significantly associated with higher odds of depression (OR = 1.58, 95%CI: 1.17–2.16, P = 0.003), indicating that having polypharmacy behaviors, increased the likelihood of depression. In the propensity score weighting regression, patients taking the traditional Chinese medicine were significantly associated with a lower likelihood of depression (OR = 0.66, 95%CI: 0.48–0.91, P = 0.013), but in the multivariate logistic regression model, there was no significance.

With regards to other covariates of the propensity score weighting regression, we found that patients with a high level of education (OR = 0.51, 95%CI: 0.30–0.87, P = 0.013), who were willing to exercise were less likely to suffer from depression (OR = 0.51, 95%CI: 0.34–0.76, P = 0.001). However, patients with hypertension (OR = 1.47, 95%CI: 1.00-2.17, P = 0.049), more diseases (OR = 1.27, 95%CI: 1.11–1.47, P = 0.001), and those with poor self-rated health status were more likely to suffer from depression (OR = 2.48, 95%CI: 1.61–3.86, P < 0.003; OR = 5.15, 95%CI:3.43–7.84, P < 0.003).

Table 6 shows the odds ratios and p value for multivariate logistic regressions and propensity scores weighting regressions for community patients. Propensity score-weighted logistic regression results showed no significant association between depressive symptoms and the occurrence of multiple medication use behaviors in community patients.

Table 6

**Multivariate logistic regression before and after the PSW for community patients**
Variables	Before weighting OR (95%CI)	P-value	After weighting OR (95%CI)	P-value
Polypharmacy behaviours
No
Yes	0.96(0.64–1.44)	0.851	0.80(0.48–1.33)	0.399
Demographic characteristics
Age(years)
18–65
66–75	0.91(0.63–1.34)	0.645	0.71(0.37–1.34)	0.293
>75	0.77(0.46–1.29)	0.327	0.44(0.19–1.02)	0.059
Domicile
Urban
Rural	2.80(1.60–4.94)	< 0.001	4.21(1.63–11.34)	0.003
Gender
Male
Female	1.21(0.79–1.86)	0.392	1.03(0.54–1.97)	0.924
Level of education
Primary school
junior high school	0.89(0.60–1.33)	0.576	0.61(0.32–1.14)	0.125
High school	0.72(0.41–1.25)	0.244	0.63(0.27–1.46)	0.290
University	0.64(0.25–1.54)	0.337	1.10(0.34–3.37)	0.866
Annual income/yuan
0-9999
10000–50000	0.76(0.52–1.12)	0.163	0.96(0.51–1.82)	0.902
>50000	0.81(0.51–1.29)	0.378	0.92(0.43–1.99)	0.828
Medical insurance
Employee health insurance
Medical Insurance for Urban and Rural Residents	1.12(0.64–1.95)	0.679	0.98(0.38–2.42)	0.961
Residence status
Live alone
Not live alone	0.86(0.53–1.40)	0.53	0.60(0.29–1.24)	0.166
Health status
Exercise
Never
Occasionally	0.86(0.52–1.42)	0.544	0.76(0.36–1.59)	0.471
Regularly	0.58(0.35–0.96)	0.033	0.48(0.22–1.02)	0.057
Always	0.49(0.31–0.78)	0.002	0.28(0.14–0.56)	< 0.001
Hypertension
No
Yes	0.46(0.27–0.80)	0.005	0.49(0.20–1.18)	0.108
Diabetes
No
Yes	0.61(0.41–0.89)	0.012	0.62(0.36–1.06)	0.082
Number of diseases
1
2	1.28(0.86–1.90)	0.222	1.70(0.74–4.07)	0.219
> 3	2.93(1.87–4.60)	< 0.001	3.67(1.52–9.25)	0.005
Disease severity
Light risk
Medium risk	1.37(0.95–1.96)	0.088	1.71(0.89–3.37)	0.114
High risk	2.23(1.42–3.51)	0.001	3.80(1.82–8.16)	< 0.001
Medication information
Adverse drug reaction
No
Yes	0.84(0.59–1.22)	0.366	0.73(0.43–1.23)	0.238
Medication adherence
Poor
Good	0.87(0.64–1.18)	0.369	0.94(0.56–1.56)	0.800
Medication knowledge
Low
High	0.78(0.56–1.09)	0.148	0.97(0.57–1.66)	0.915
Medical behavior
Whether trust the doctor
Very trust
More trust	1.07(0.26–4.55)	0.925	0.67(0.04–8.55)	0.765
Commonly	0.40(0.10–1.70)	0.210	0.06(0.00-0.87)	0.046
Don't trust	0.50(0.13–1.93)	0.306	0.28(0.02–3.13)	0.317
Very mistrust	0.33(0.09–1.27)	0.099	0.13(0.01–1.49)	0.114
Whether be in hospital in the past year
No
Yes	0.59(0.43–0.81)	0.001	0.64(0.39–1.06)	0.084
Regular blood pressure measurement
No
Yes	1.14(0.76–1.69)	0.523	0.69(0.33–1.40)	0.313
Type of decision
Personal decision-making
Shared decision-making	0.22(0.10–0.50)	< 0.001	0.16(0.04–0.56)	0.005
Doctor decision-making	0.54(0.31–0.95)	0.031	0.48(0.18–1.26)	0.138
Constant	2.39(0.39–14.34)	0.339	10.41(0.49-261.17)	0.135

After weighting the community patients data, patients' household registration, exercise status, severity of illness, type of involvement in medication decisions, and number of illnesses were significantly associated with whether patients were depressed. Patients in rural areas were more likely to be depressed relative to those in urban areas (OR = 4.21, 95% CI: 1.63–11.34, p = 0.003), patients who always exercised were less likely to be depressed relative to those who never exercised (OR = 0.28, 95% CI: 0.14–0.56, p < 0.001), patients with severe disease were more likely to be depressed relative to those without severe disease (OR = 3.80, 95% CI: 1.83–8.16, p < 0.001), patients with shared decision making were less likely to be depressed relative to those who made their own decisions (OR = 0.16, 95% CI: 0.04–0.56, p = 0.005), and patients with three or more conditions were more likely to be depression (OR = 3.67, 95% CI: 1.52–9.25, P = 0.005).

This study was conducted to investigate the prevalence of depression in hospitalized patients with chronic diseases in Wuhan, Hubei, China, and to explore its association with polypharmacy behaviors. We found that 40.5% of the patients had depressive symptoms, which was consistent with existing study^[23], but higher than the rate that Canoui-Poitrine had found^[25]. The possible explanation is that different criteria for depressive symptoms have been adopted in different studies.

Significant associations have been found between polypharmacy behaviors and depression in the multivariate logistic regression model. Compared with non-polypharmacy patients, polypharmacy patients were directly associated with depression (OR = 1.63, 95% CI: 1.12–2.36). After the propensity score weighting approach was applied, patients who had polypharmacy behaviors were more likely to suffer from depression (OR = 1.58, 95%CI: 1.17–2.16), which was consistent with the results of Onder’s study^[18]. The possible immediate cause for these associations is that patients taking multiple medications have higher psychological distress, which was proposed in Assari’s study^[14], and may eventually evolve into depressive symptoms or even clinical depression. Further, although patients might hope to reduce the number of medications they are taking, their final decision on medication-taking still mainly depends on the prescription issued by physicians. In other words, patients have the problem of high external rely of control, namely that physicians are in control of their health^[26]. According to Bramston’s study^[27], older adults with a high external locus of control develop depression, which is also a process of learned hopelessness^[28] since there is nothing patients can do to change polypharmacy, implying that people’s hopelessness was frequently changing their future, resulting in them suffering from more depressive symptoms^[29]. In addition, polypharmacy behaviors have been observed to be associated with a range of negative health outcomes in patients, including falls, frailties and adverse drug events^[30], all of which can increase patients' stress and negatively impact their mental health, possibly leading to depression. However, adverse drug reactions were not found to be associated with depression in this study (P = 0.327). On the other hand, it was discovered in a previous study that poorer emotional capability was associated with increased medication use, suggesting that depression might also lead to more medication use^[20]. A reasonable explanation could be that depression increases self-medication and the prevalence of polypharmacy behaviors^[31]. There were also some researchers who believed that it was multimorbidity that fully mediated the association between polypharmacy behaviors and depression. That is, depression in patients with polypharmacy behaviors might be caused by a higher number of multimorbid conditions^[32].

In terms of the physical health status of the participants, hypertension, the number of diseases they had, and their self-rated health status were all found to be risk factors for depressive symptoms in this study. The health status was a significant predictor of depression in a prior study^[33]. Perceived health-stress has strong impacts on both depressive symptoms and clinical depression. The worse the self-rated health status of patients with chronic diseases, the higher their self-perceived pressure on health, and the more likely they are to have depressive symptoms and clinical depression^[34]. The number of diseases also showed strong associations with patients’ depressive symptoms in this study. In patients with chronic diseases, the probability of developing depressive symptoms increases by 0.27 (0.11–0.47) times for each additional disease. Polypharmacy behaviors are more common in patients with multiple chronic diseases, and they are more likely to develop symptoms of depression. Similar results were shown in relevant studies of Canoui-Poitrine’s^[25], which indicated that clinical depression was associated with multimorbidity and polypharmacy behaviors independently. In addition, we also discovered that patients with hypertension were more likely to have depressive symptoms in our study, which is consistent with the findings of the study of Xue^[35], namely that patients with uncontrolled hypertension are more likely to develop clinically significant symptoms of depression. The findings of this study support the complex relationships between hypertension and depression. The high proportion of patients who have both hypertension and depression suggests that the two disorders interact, implying that hypertension might be a risk factor for depression. In other relevant studies, this association has also been attributed to the poor lifestyle of patients with depression, somatic symptoms of depression induced by antihypertensive drugs, and increased healthcare utilization of patients with depression to enhance the chance of diagnostic recognition of hypertension^{[36, 37]}. Some other researchers believe that this association might be due to the limitations of self-assessed screening tools for depression. Patients with hypertension might be confused about symptoms of hypertension and depression, such as fatigue symptoms and sleeping disorders. They may consider these symptoms commonly suffered from hypertension as an indication of depression and classify themselves as patients with depression^[38]. In addition, we discovered that whether or not patients participate in exercise was related to depressive symptoms in our study. Patients who took exercise were less likely to suffer from depression. Exercise has been demonstrated in previous studies to assist people reduce negative emotions^[39] and improve sleep quality^[22]. Exercise was found to decrease depression scores significantly^[40], even having a stable and robust effect in depression severity among older patients with clinically significant symptoms of depression^[41]. On the one hand, physical activities have been demonstrated to upregulate monoamine neurotransmission in animal brains, which might be linked to emotional disorders of human ^[42]. On the other hand, exercise can also distract people from stressful life events and reduce negative attention biases^[43].

Finally, in terms of other medication information, patients taking Chinese medicine to treat chronic diseases were less likely to be depressed (OR = 0.66, 95% CI: 0.48–0.91). At present, there are few international studies on this aspect. The reason might be that the majority of the patients we surveyed are the elderly, and the most of Chinese elderly have a stronger trust in traditional Chinese medicine. The elderly believe that the traditional Chinese medicine has fewer side effects, therefore they pay less attention to the side effects of drugs^[44].

Limitations

Our findings should be interpreted cautiously since there were several limitations to this study. Firstly, various definitions of polypharmacy behaviors existed in the literature, we only considered the number of drugs used, namely more than or equal to five drugs as polypharmacy behaviors, so it is difficult to make a distinction between necessary prescribing and polypharmacy behaviors. Secondly, the sample was from eight administrative regions in Wuhan City, so our findings might not be generalizable to other areas. Thirdly, we used no diagnostic assessment of depression, the CES-D 10 scale used in this study could only screen for the presence of depressive symptoms. A complete diagnostic assessment of clinical depression would be better in future researches. Finally, polypharmacy behaviors were measured based on the self-reported number of medications in this study, other sources should be considered to verify the responses.

In our study, a significant relation was observed between polypharmacy behaviors and depressive symptoms. Furthermore, multimorbidity, having hypertension, no exercise, and having a poor self-rated health status were also found to be significantly associated with depressive symptoms. The results have valuable implications for promoting the health of patients with chronic diseases in clinical practice, indicating that the psychological needs of patients with chronic diseases who receive a polypharmacy should not be discounted. At the same time, interventions should also be taken to address inappropriate polypharmacy behaviors in patients with chronic diseases who have poor mental health. Given the overlap between polypharmacy behaviors and depression, such measures might contribute to improving the health and wellbeing of patients with chronic diseases.

Twitter

Da Feng @fengda

Author Contributions

Jia Wang, Jian Zou and Yuxin Liu wrote the main manuscript text, and Chaoyi Chen, Zehao Zheng, Yan Cao, Rui Huang and Da Feng critically reviewed the manuscript and propose amendments. All authors reviewed the manuscript. Yangfan Sun revised the manuscript text. Zhanchun Feng Guided this research.

Declaration of Competing Interest

None

Acknowledgments

The authors thank all the staff and participants of this study for their important contributions.

Funding

This work was supported by the National Natural Science Foundation of China [Grant numbers 71804052], National Natural Science Foundation of China [Grant numbers 72274071], and National Key Research and Development Program [Grant number: 2020YFC2006500].

Availability of data and materials

The datasets generated and/or analysed during the current study are not publicly available due institutional research ethics agreement but may be available from the corresponding author on reasonable request.

Ethics statement

The study was conducted in accordance with the Declaration of Helsinki. Written Informed consent was obtained from all patients. This study was approved by the Medical Ethics Committee of Tongji Medical College of Huazhong University of Science and Technology, and the approval number is 2020(S223).

Consent for publication

Not applicable

Oppong Asante K, Andoh-Arthur J. Prevalence and determinants of depressive symptoms among university students in Ghana. J AFFECT DISORDERS. 2015;171:161–6.
Chen H, Xiong P, Chen L, et al. Childhood neighborhood quality, friendship, and risk of depressive symptoms in adults: The China health and retirement longitudinal study. J AFFECT DISORDERS. 2020;276:732–7.
Whiteford HA, Degenhardt L, Rehm J, et al. Global burden of disease attributable to mental and substance use disorders: findings from the Global Burden of Disease Study 2010. Lancet. 2013;382:1575–86.
Ijaz S, Davies P, Williams CJ, et al. Psychological therapies for treatment-resistant depression in adults. COCHRANE DATABASE Syst REVIEWS. 2018;5:CD010558.
Garlow SJ, Rosenberg J, Moore JD, et al. Depression, desperation, and suicidal ideation in college students: results from the American Foundation for Suicide Prevention College Screening Project at Emory University. Depress Anxiety. 2008;25:482–8.
Pitpitan EV, Kalichman SC, Eaton LA, et al. Gender-based violence and HIV sexual risk behavior: alcohol use and mental health problems as mediators among women in drinking venues, Cape Town. Soc Sci Med. 2012;75:1417–25.
Nduna M, Jewkes RK, Dunkle KL, et al. Associations between depressive symptoms, sexual behaviour and relationship characteristics: a prospective cohort study of young women and men in the Eastern Cape, South Africa. J Int AIDS Soc. 2010;13:44.
De Las CC, Sanz EJ. Polypharmacy in psychiatric practice in the Canary Islands. BMC Psychiatry. 2004;4:18.
Haider SI, Johnell K, Thorslund M, et al. Analysis of the association between polypharmacy and socioeconomic position among elderly aged > or = 77 years in Sweden. CLIN THER. 2008;30:419–27.
Liu Y, Wang R, Huang R, et al. Influencing factors and their relationships of risk perception and decision-making behaviour of polypharmacy in patients with chronic diseases: a qualitative descriptive study. BMJ OPEN. 2021;11:e43557.
Rollason V, Vogt N. Reduction of polypharmacy in the elderly: a systematic review of the role of the pharmacist. Drugs Aging. 2003;20:817–32.
Pamoukdjian F, Aparicio T, Zelek L, et al. Impaired mobility, depressed mood, cognitive impairment and polypharmacy are independently associated with disability in older cancer outpatients: The prospective Physical Frailty in Elderly Cancer patients (PF-EC) cohort study. J GERIATR ONCOL. 2017;8:190–5.
Milton JC, Hill-Smith I, Jackson SH. Prescribing for older people. BMJ. 2008;336:606–9.
Assari S, Bazargan M. Polypharmacy and Psychological Distress May Be Associated in African American Adults. Pharmacy. 2019;7:14.
Schaakxs R, Comijs HC, van der Mast RC, et al. Risk Factors for Depression: Differential Across Age? Am J Geriatric Psychiatry. 2017;25:966–77.
Hölzel L, Härter M, Reese C, et al. Risk factors for chronic depression — A systematic review. J AFFECT DISORDERS. 2011;129:1–13.
Bahat G, Tufan F, Bahat Z, et al. Assessments of functional status, comorbidities, polypharmacy, nutritional status and sarcopenia in Turkish community-dwelling male elderly. Aging Male. 2013;16:67–72.
Onder G, Liperoti R, Fialova D, et al. Polypharmacy in Nursing Home in Europe: Results From the SHELTER Study. Journals Gerontol Ser A: Biol Sci Med Sci. 2012;67A:698–704.
Volaklis KA, Thorand B, Peters A, et al. Physical activity, muscular strength, and polypharmacy among older multimorbid persons: Results from the KORA-Age study. SCAND J MED SCI SPOR. 2018;28:604–12.
Khezrian M, McNeil CJ, Myint PK, et al. The association between polypharmacy and late life deficits in cognitive, physical and emotional capability: a cohort study. INT J CLIN PHARM-NET. 2019;41:251–7.
Andresen EM, Malmgren JA, Carter WB, et al. Screening for Depression in Well Older Adults: Evaluation of a Short Form of the CES-D. AM J PREV MED. 1994;10:77–84.
Chen H, Mui AC. Factorial validity of the Center for Epidemiologic Studies Depression Scale short form in older population in China. INT PSYCHOGERIATR. 2014;26:49–57.
Ma C. The prevalence of depressive symptoms and associated factors in countryside-dwelling older Chinese patients with hypertension. J CLIN NURS. 2018;27:2933–41.
Zhang J, Sun W, Kong Y, et al. Reliability and validity of the Center for Epidemiological Studies Depression Scale in 2 special adult samples from rural China. COMPR PSYCHIAT. 2012;53:1243–51.
Canoui-Poitrine F, Reinald N, Laurent M, et al. Geriatric assessment findings independently associated with clinical depression in 1092 older patients with cancer: the ELCAPA Cohort Study. Psychooncology. 2016;25:104–11.
Liu CB, Leung DS, Chi I. Social functioning, polypharmacy and depression in older Chinese primary care patients. AGING MENT HEALTH. 2011;15:732–41.
Bramston P, Tomasevic V. Health locus of control, depression and quality of life in people who are elderly. AUSTRALAS J AGEING. 2001;20:192–5.
Abramson LY, Metalsky GI, Alloy LB. Hopelessness depression: A theory-based subtype of depression. PSYCHOL REV. 1989;96:358–72.
Joiner TE, Wingate LR, Otamendi A. An interpersonal addendum to the hopelessness theory of depression: Hopelessness as a stress and depression generator. J SOC CLIN PSYCHOL. 2005;24:649–64.
Nightingale G, Skonecki E, Boparai MK. The Impact of Polypharmacy on Patient Outcomes in Older Adults With Cancer. CANCER J. 2017;23:211–8.
Guaraldo L, Cano FG, Damasceno GS, et al. Inappropriate medication use among the elderly: a systematic review of administrative databases. BMC GERIATR. 2011;11:79.
Bazargan M, Smith J, Saqib M et al. Associations between Polypharmacy, Self-Rated Health, and Depression in African American Older Adults; Mediators and Moderators. Int J Environ Res Public Health 2019; 16(9).
Jee YJ, Lee YB. Factors Influencing Depression among Elderly Patients in Geriatric Hospitals. J Phys Ther Sci. 2013;25:1445–9.
Tsai AC, Chi SH, Wang JY. The association of perceived stress with depressive symptoms in older Taiwanese-Result of a longitudinal national cohort study. PREV MED. 2013;57:646–51.
Xue J, Chen S, Bogner HR, et al. The prevalence of depressive symptoms among older patients with hypertension in rural China. Int J Geriatr Psychiatry. 2017;32:1411–7.
Jackson CA, Pathirana T, Gardiner PA. Depression, anxiety and risk of hypertension in mid-aged women: a prospective longitudinal study. J HYPERTENS. 2016;34:1959–66.
Michal M, Wiltink J, Lackner K, et al. Association of hypertension with depression in the community: results from the Gutenberg Health Study. J HYPERTENS. 2013;31:893–9.
Li Z, Li Y, Chen L, et al. Prevalence of Depression in Patients With Hypertension: A Systematic Review and Meta-Analysis. Med (Baltim). 2015;94:e1317.
Jin P. Efficacy of Tai Chi, brisk walking, meditation, and reading in reducing mental and emotional stress. J PSYCHOSOM RES. 1992;36:361–70.
Shahidi M, Mojtahed A, Modabbernia A, et al. Laughter yoga versus group exercise program in elderly depressed women: a randomized controlled trial. Int J Geriatr Psychiatry. 2011;26:322–7.
Bridle C, Spanjers K, Patel S, et al. Effect of exercise on depression severity in older people: systematic review and meta-analysis of randomised controlled trials. Br J Psychiatry. 2012;201:180–5.
Dishman RK. Brain monoamines, exercise, and behavioral stress: animal models. Med Sci Sports Exerc. 1997;29:63–74.
Hallgren M, Nguyen T, Owen N, et al. Cross-sectional and prospective relationships of passive and mentally active sedentary behaviours and physical activity with depression. Br J Psychiatry. 2020;217:413–9.
Wang J, Matthews JT. Chronic disease self-management: views among older adults of Chinese descent. Geriatr Nurs. 2010;31:86–94.

No competing interests reported.

Associations between Polypharmacy and Depressive Symptoms in Patients with Chronic Diseases: A Cross-Sectional and Prospective Study in Wuhan, Hubei Province

Status:

Version 1

Abstract

1. Background

2. Methods

2.1 Study participants and procedure

2.2 Measure

2.2.1 Demographic characteristics

2.2.2 Health status

2.2.3 Medication information

2.2.4 Dependent variable

2.3 Patient and public involvement

2.4 Statistical analysis

3. Results

3.1 Interviewee information

3.2 Propensity score weighting (PSW) results

3.3 Regression analysis results

4. Discussion

5. Conclusion

Declarations

References

Additional Declarations

Status:

Version 1