The roles of busyness and daily routine in medication management behaviors among older adults

Abstract

Busyness (the density of activities) and daily routine (patterns of organizing time) are two understudied factors that likely impact medication-taking behaviors. We examined the association between busyness and routine with medication adherence (MA) in 405 older adults with adequate cognition using multivariable models. The final model included an interaction term between daily routine and busyness. MA scores (measured by the ASK-12, higher scores mean more barriers to adherence) were higher for individuals reporting low and moderate levels of daily routine versus those with high daily routine. MA scores were higher for individuals reporting moderate and high busyness versus those reporting low busyness. The busyness/routine interaction term was significant for MA; among highly busy individuals, those with high daily routine had lower MA scores than those with low routine. A daily routine may be a modifiable factor for improving MA among older adults, particularly among those with busy lives.

Introduction

Approximately 40% of older adults aged 65 years and older take five or more prescription medications to manage multiple chronic conditions (Kantor et al., 2015). Unfortunately, close to half of individuals take their medications less than prescribed (Osterberg & Blaschke, 2005), which increases the risk of hospitalization for poorly controlled chronic conditions. There are many well-documented individual-level barriers to medication adherence including cost (Devine et al., 2018), limited health literacy (Devine et al., 2018), cognitive abilities (Stilley et al., 2010), and beliefs and concerns about medications (Gagnon et al., 2017). Yet, medication management behaviors function within and are influenced by the surrounding environmental and social contexts. Busyness, the density of activities and obligations, and routine, established patterns of organizing time, make up the structure of daily life, and are two promising, but understudied, factors that may impact medication taking behaviors.

Busyness is often cited in daily conversation. Despite the ubiquity of busyness in everyday life, few empirical studies have considered its role in health management. On one hand, greater busyness is associated with better cognition (Festini et al., 2016), which is a well-established predictor of medication management (Stilley et al., 2010). On the other hand, busyness may challenge the execution of planned behaviors, especially among older individuals managing multi-drug regimens that must be taken at multiple times throughout the day. Previous research has found that busyness, more so than age, contributed to unintentional medication nonadherence (Ng et al., 2014; Park et al., 1999).

The formation of a daily routine may help to counteract some of the deleterious effects of busyness (Fritz, 2014). Many people have established patterns of waking, eating, sleeping and organizing their time; this is often done in the form of a routine to provide a sense of coherence and predictability to one’s day. The formation of a daily routine has frequently been described by chronically ill adults as a primary facilitating mechanism to adopt and maintain medication taking behaviors (Klinedinst et al., 2021; O’Conor et al., 2017; Sanders & Van Oss, 2013), yet the extent to which having a daily routine is related to medication taking behaviors in the context of a busy lifestyle has not been investigated. Though busyness and routine are related, they are distinct concepts, and studying one without the other provides an incomplete picture of how daily rhythms and activities challenge or support self-management behaviors. We therefore examined the association between daily routine and busyness with medication adherence. We hypothesized that greater daily routine would support medication adherence among older adults who report busy lives.

Methods

Sample and Procedure

We conducted a secondary data analysis using data from the Health Literacy and Cognitive Function in Older Adults Study (LitCog). LitCog is a cognitive aging cohort study examining changes in cognition, health literacy, and self-care skills and their relation to health outcomes. Beginning in 2008, 900 adults were recruited from one academic general internal medicine practice and six federally qualified health centers in Chicago, Illinois, United States. English speaking adults who sought regular care (defined as two clinic visits within the past 2 years) from study sites were identified through practice records, and research coordinators contacted patients by telephone to screen for eligibility. Patients were eligible if they (1) were between the ages 55 and 74, (2) spoke English, and (3) had adequate cognitive capacity, as defined by ≤2 errors on the 6-item screener (Callahan et al., 2002). Enrolled participants are invited to complete follow-up interviews every three years. Cross sectional data from the fourth timepoint were used, which 405 of the original 900 participants completed. Among those in the cohort who did not participate, 79 were not yet eligible for the fourth time point interview, 89 were deceased or no longer community dwelling, 185 declined participation, and 142 were unable to be contacted. Among those due to complete the fourth time point interview, those who did not participate were more likely to be male, identified as Black, have less income, educational attainment, have limited health literacy, and more chronic conditions at baseline (p’s < .05). The study was approved by the Institutional Review Board at Northwestern University.

Measures

Busyness and Routine

Daily routine and busyness were assessed using the Martin and Park Environmental Demands (MPED) Questionnaire (Martin & Park, 2003). The MPED is a brief instrument that was developed to evaluate self-reported environmental demands in the form of busyness and routine within an individual’s daily life. The daily routine subscale includes four items that measure the frequency that an individual perceives following a regular routine in his or her behaviors every day. These items capture the frequency someone engages in daily activities at the same time, including getting up in the morning and going to bed in the evening, eating meals, and engaging in activities at home; the final item asks participants to describe how often their days follows a basic routine. The busyness subscale includes seven questions that assess the density of events in an individual’s daily life. The items capture the frequency someone feels busy or rushed in a variety of settings. Each item is rated on a 5-point Likert scale with responses ranging from never to always. Scores on the routine subscale range from 4 to 20, with higher scores indicating greater daily routine. Scores on the busyness subscale range from 7 to 35, with higher scores indicating greater busyness. Tertiles were calculated for the routine and busyness subscales; our decision to use tertiles was based on clinical interpretability as well as empirical data (O’Conor et al., 2019). Clinically, low, medium, and high levels of busyness and routines are more easily interpreted than a continuous number or simply low and high levels. Empirically, tertiles have been used with this dataset, where low routine was related to worse physical function, depression, and anxiety (O’Conor et al., 2019).

Medication Adherence

Medication adherence was measured using the 12-item Adherence Starts with Knowledge (ASK-12) questionnaire (Matza et al., 2009). The ASK-12 is a subjective assessment of general adherence behaviors and barriers to treatment adherence, and contains 3 subscales: forgetfulness, health beliefs, and behaviors (Matza et al., 2009). Total scores and subscale scores were summed (total scores ranging from 12 to 60), with higher scores indicating greater barriers to adherence. Scores for each subscale ranged from 3–15 (forgetfulness), 4–20 (health beliefs) and 5–25 (behaviors), with higher scores indicating greater barriers.

Covariates

Several covariates that have well-established associations with medication adherence were also collected. Health literacy was measured using the Test of Functional Health Literacy in adults (TOFHLA), which assesses comprehension of actual health information and is comprised of a numeracy (17 items) and a literacy section (50 items). Scores range from 0 to 100, with higher scores indicating higher health literacy. Scores are classified as limited (0–74) or adequate (75–100) health literacy (Parker et al., 1995). The Mini Mental State Exam (MMSE) is a global measure of cognitive ability and is commonly used in medical settings (Folstein et al., 1975); scores range from 0 to 30, with higher scores indicating greater cognitive ability. We also included self-reported demographic characteristics (age, biological sex, race, income), and participants self-reported their number of active prescription medications.

Analysis Plan

Differences between participant characteristics and daily routine and busyness scores (measured continuously) were analyzed using analysis of variance (ANOVA). Next, we calculated a Pearson correlation coefficient to determine the amount of overlap between routine and busyness. A series of generalized linear regression models were conducted to identify the association of daily routine and busyness with medication adherence (overall score and the three individual subscales). First, we ran models using busyness and routine tertiles separately, and then a combined model that included both routine and busyness tertiles. To ensure that assumptions of multivariable regression were met, we performed residual analyses. We found that all residuals were acceptable and held to assumptions of normality; we also found no influential outliers. Descriptive statistics for ASK-12 total and subscales can be found in Supplementary Table 1. To further understand if busyness and medication management differed based on different levels of routine, interaction terms based on routine and busyness tertiles were added to the models that included both routine and busyness. All models included age, biological sex, race, income, health literacy, number of prescription medicines and cognitive ability as covariates. We reported least squares means (LSM) and 95% confidence intervals. Statistical analyses were performed using STATA/SE software, version 15 (College Station, TX).

Results

A total of 405 participants were included in this analysis (Table 1). The mean age of participants was 71.3 years (SD 5.3; range 62–83 years). Participants were mostly female (71.8%), self-identified as White (56.7%) or Black (37.9%) and were concurrently taking 5 or more medications (42.2%). Nearly half of the participants (49.1%) had a yearly income greater than $50,000, and the majority (71.6%) had adequate health literacy and no cognitive impairment (94.1%). The routine and busyness scores were not linearly correlated (r = −0.06, p = 0.21).

Table 1.

Participant characteristics and daily routine and busyness (n=405)

		Routine		Busyness
Characteristic	n (%)	Mean (SD)	p-value	Mean (SD)	p-value
Age			0.25		<0.001
60–64	44 (10.9)	13.0 (2.5)		18.8 (5.7)
65–74	239 (59.0)	13.5 (3.1)		17.0 (4.9)
75+	122 (30.1)	13.8 (2.9)		15.6 (4.2)
Biological sex			0.47		0.004
Male	114 (28.2)	13.7 (2.9)		15.7 (4.7)
Female	291 (71.8)	13.4 (3.0)		17.2 (4.9)
Income			<0.001		0.32
<$10,000	38 (9.5)	11.1 (3.1)		16.8 (4.9)
$10,000 – $24,999	76 (19.0)	13.1 (3.1)		15.9 (5.0)
$25,000 – $49,999	90 (22.4)	13.4 (2.9)		17.0 (4.6)
>$50,000	197 (49.1)	14.2 (2.6)		17.1 (4.9)
Race			<0.001		0.31
White	229 (56.7)	14.0 (2.8)		16.5 (4.9)
Black	153 (37.9)	12.6 (3.0)		17.0 (4.9)
Other	22 (5.4)	14.5 (2.9)		18.0 (4.9)
Prescription Medications			0.07		0.78
0 – 1	77 (19.0)	14.0 (2.6)		17.0 (4.8)
2 – 4	157 (38.8)	13.7 (2.8)		16.6 (4.6)
5+	171 (42.2)	13.1 (3.2)		16.9 (5.2)
Health Literacy			<0.001		0.09
Adequate	290 (71.6)	13.8 (2.8)		17.1 (4.9)
Limited	115 (28.4)	12.7 (3.2)		16.1 (4.8)
MMSE			0.02		0.34
No Impairment	380 (94.1)	13.6 (2.9)		16.9 (4.9)
Mild Impairment	24 (5.9)	12.2 (3.8)		15.9 (5.3)

Daily routine did not vary by age (p = 0.25), biological sex (p = 0.47), or number of prescription medications (p = 0.07) (Table 1). However, there were group differences in routine based on income (p < 0.001), where individuals who make more than $50,000 per year tended to have higher levels of routine than those with lower incomes. There were also differences between racial groups (p < 0.001), where White and Other respondents tended to have higher levels of routine than Black respondents. Lastly, there were group differences in health literacy (p < 0.001) and cognitive impairment (p = 0.02), where people with limited health literacy and mild cognitive impairment reported lower levels of daily routine.

Busyness responses did not vary by income (p = 0.32), race (p = 0.31), number of prescription medications (p = 0.78), health literacy (p = 0.09), or cognitive impairment status (p = 0.34) (Table 1). There were group differences in busyness based on age (p < 0.001), where younger respondents tended to be busier, and biological sex (p = 0.004), where females tended to be busier.

In multivariable analyses (Table 2), we observed overall ASK-12 scores were worse for individuals reporting low (LSM 22.4, 95% CI: 21.2, 23.6) and moderate (LSM 22.4; 95% CI: 21.1, 23.8) levels of daily routine versus those with high daily routine (LSM 20.5; 95% CI: 19.5, 22.2). ASK-12 scores were also worse for individuals reporting moderate (LSM 21.9; 95% CI: 20.7, 23.2) and high (LSM 23.6; 95% CI 22.2, 24.9) busyness versus those reporting low busyness (LSM 20.5; 95% CI: 19.3, 21.8). There were no changes to significant associations when busyness and routine were added into the model together. We observed a significant interaction term between busyness and routine for the ASK-12 (p=0.04); among highly busy individuals, those with high daily routine reported better medication adherence than those with low daily routine (Figure 1). See Supplementary Table 2 for p-values for all models that explored interactions among high and low busyness and routine.

Table 2.

Results of Multivariable Models

Outcome	Model 1 Routine	Effect Size§	Model 2 Busyness	Effect Size§	Model 3 Routine + Busyness	Effect Size§
Medication Adherence	LSM (95% CI)		LSM (95% CI)		LSM (95% CI)
Total Score
Routine
High	20.5 (19.2, 21.9)				20.9 (19.5, 22.2)
Moderate	22.4 (21.1, 23.8) †	0.022			22.6 (21.3, 24.0)*	0.019
Low	22.4 (21.2, 23.6) †				22.4 (21.3, 23.6)*
Busyness
Low			20.5 (19.3, 21.8)		20.5 (19.3, 21.7)
Moderate			21.9 (20.7, 23.2)*	0.049	22.0 (20.7, 23.2)*	0.045
High			23.6 (22.2, 24.9) ‡		23.4 (22.1, 24.8) ‡
Forgetfulness
Routine
High	6.2 (5.6, 6.8)				6.4 (5.7, 7.0)
Moderate	6.6 (6.0, 7.3)	0.007			6.7 (6.1, 7.4)	0.005
Low	6.7 (6.2, 7.3)				6.8 (6.2, 7.3)
Busyness
Low			5.9 (5.3, 6.5)		5.9 (5.3, 6.5)
Moderate			6.6 (6.0, 7.2)*	0.051	6.6 (6.0, 7.2)*	0.048
High			7.4 (6.7, 8.0) ‡		7.3 (6.7, 8.0) ‡
Behavior
Routine
High	7.1 (6.5, 7.8)				7.3 (6.6, 8.0)
Moderate	8.2 (7.5, 8.9) †	0.024			8.3 (7.6, 9.0) †	0.23
Low	7.6 (7.1, 8.2)				7.6 (7.1, 8.2)
Busyness
Low			7.1 (6.5, 7.7)		7.2 (6.6, 7.8)
Moderate			7.6 (7.0, 8.2)	0.035	7.6 (7.0, 8.2)	0.34
High			8.4 (7.7, 9.0) ‡		8.4 (7.7, 9.1) ‡
Beliefs
Routine
High	7.2 (6.6, 7.8)				7.2 (6.6, 7.8)
Moderate	7.6 (7.0, 8.2)	0.020			7.6 (7.0, 8.2)	0.019
Low	8.0 (7.5, 8.5) †				8.0 (7.5, 8.5) †
Busyness
Low			7.5 (7.0, 8.0)		7.4 (6.9, 8.0)
Moderate			7.7 (7.2, 8.3)	0.004	7.7 (7.2, 8.2)	0.004
High			7.9 (7.3, 8.4)		7.7 (7.1, 8.3)

Note. Ranges for the ASK-12 total scale and subscales are as follows: Total = 12 – 60, Forgetfulness = 3 – 15, Behavior = 5 – 25, Health Beliefs = 4 – 20.

^*= p < 0.001

^†= p < 0.01

^‡= p < 0.05

^§= effect size calculated is partial eta squared (.01 = small effect, .06 = medium effect, .14 and higher = large effect)

Figure 1.

Medication adherence scores by busyness and routine status

Note. Moderate Busy group was removed for ease of interpretation. See Supplementary Figure 1 for all groups.

With regards to the forgetfulness subscale, individuals reported greater levels of forgetfulness among those with moderate (LSM 6.6; 95% CI 6.0, 7.2) and high (LSM 7.4; 95% CI 6.7, 8.0) busyness compared to those reporting low busyness (LSM 5.9; 95% CI 5.3, 6.5). On the behavior subscale, individuals reported poorer medication-taking behaviors among those with a moderate level of routine (LSM 8.2; 95% CI 7.5, 8.9) compared to those with high levels of daily routine (LSM 7.1; 95% CI 6.5, 7.8). Additionally, those with high levels of busyness reported poorer medication taking behaviors (LSM 8.5; 95% CI 7.7, 9.0) compared to those with low levels of busyness (LSM 7.1; 95% CI 6.5, 7.7). There were no changes to significant associations when busyness and routine were added into the model together.

Discussion

Among a sample of community dwelling older adults, we found that both the busyness of an individual’s life and their level of daily routine were significant predictors of medication adherence. Individuals reporting high levels of daily routine reported better medication adherence, and those reporting less busy lives also reported better medication adherence. Furthermore, as hypothesized, of older adults who reported greater busyness, those with a highly structured daily routine have better medication management than older adults with low levels of routine. This relationship did not hold for older adults who were not busy, suggesting that stability of routine is of particular importance to busy older adults.

With the onset of the coronavirus 2019 (COVID-19) pandemic in March 2020, the importance of daily routine to functioning came into full focus as daily routines were upended by the COVID-19 pandemic. As many businesses closed, employers began sending staff to work remotely, schools moved toward remote learning, residential facilities closed to visitors, and public restrictions were put in place, we all experienced dramatic changes to our daily rhythms. During this time, many mental health experts recommended keeping up daily routines to the extent possible during lockdown (Hou, Lai, et al., 2020). This idea is based on evidence that the stability of daily patterns could be a protective factor against intense disruption to daily life, and thus a protective factor for mental health (Hou, Liu, et al., 2020). In addition to promoting mental well-being, a daily routine also has the ability to promote continuous engagement in self-management behaviors, potentially even during times of ongoing stress.

Previous research examining busyness and medication management have shown that the relationship between busyness and medication non-adherence was supported only for unintentional types of adherence (e.g., forgetfulness, carelessness), but not for intentional nonadherence (e.g., beliefs and concerns about medications). This relationship held in our study, where busyness was significantly associated with higher scores on the forgetfulness and behaviors subscales of the ASK-12. It is important to note the inconsistencies found in the results, particularly in our subscale results. Medication- taking behaviors are complex and influenced by many factors; future work should look deeper at the different facets of medication taking behaviors. Previous research has also found that individuals leading busier lives had greater cognition (Festini et al., 2016). While this may be the case, in the context of carrying out complex self-management behaviors, such as multidrug medication regimens, busyness appears to be a detriment. Additional research is necessary to further understand the impact of busyness on both cognitive health and health management behaviors.

In general, older adults in our sample with low and moderate levels of daily routine had worse medication management than older adults with high routine. These findings are consistent with previous work about the role of daily routine in medication taking behaviors (Sanders & Van Oss, 2013). The novelty of the present study lies in the inclusion of the interaction terms between routine and busyness. To our knowledge, this is the first study to report both busyness and routine in the context of medication taking behaviors. We found that for busy older adults, having a high level of routine was a protective factor for medication taking behaviors. This finding has implications for developing interventions to support medication management in busy older adults; daily routines are a critical factor in interventions to diminish the impact of busyness on missed medications. Establishing medication taking routines may be even more useful for people facing limited health literacy and impaired cognition, as routines increase the efficiency of performance by diminishing demand on cognitive processes, like memory and decision making (Radomski & Davis, 2002).

In the context of health behavior change, much more has been reported about the relationship between habit strength and medication use. Habit refers to automatic, cue-contingent actions that are developed through learned stimulus–response associations and are triggered by situations or environments (Gardner, 2015). For example, someone may build a habit of using their inhaler every night by attaching it to their toothbrush. In this case, the stimulus is seeing the inhaler and the response is using it. This habit was created by integrating inhaler use into their existing nightly oral hygiene routine. This example illustrates how habits and routines are inextricably linked; health habits are often small actions that must be integrated into the existing routines of daily life to become automatic. Interventions to use habits as the cue-contingent building blocks of routine may have utility for helping older adults to better manage medications, particularly when the medication regimens are complex and influenced by the surrounding environment (Wood & Neal, 2016).

Based on this and previous research, targeting interventions to the development of skills to plan, organize, and maintain a daily routine may be beneficial. Two such intervention approaches that have potential for optimizing daily routines are action planning and behavioral activation. In action planning, patients learn the process of implementation intention, where they specify how, where, and when they complete a desired behavior (Gollwitzer & Sheeran, 2006). This dialogue, paired with learning tools, could be used to facilitate the development and maintenance of a health-promoting daily routine (Park et al., 2007). A second approach with promise for improving daily routine is Behavioral Activation, specifically, activity scheduling and monitoring (Brick et al., 2020). In approaches using activity monitoring and scheduling, health-promoting daily activities are scheduled, and the patient monitors enjoyment and importance of that activity. Though originally developed to treat clinical depression by building positive reinforcement contingencies, activity scheduling has the added benefit of assisting the patient to develop and maintain daily routines (Lejuez et al., 2011). Both interventions are brief and can be carried out in a clinic by allied health professionals (e.g., occupational therapists) or lay health workers. Interventional work should be cognizant that the development and maintenance of a daily routine may be difficult for individuals exposed to high levels of stress or instability. Related work, on the role of life chaos, has also found greater life chaos to be associated with reduced medication adherence among individuals managing HIV (Wong et al., 2007), and further underscores the need for tailoring the development of a daily routine for at-risk populations.

Limitations

Our findings should be recognized in the context of several limitations. First, our sample consisted of older adults with adequate cognitive capacity, and may not be generalizable to younger populations, or those with cognitive deficits. Though we found significant differences among scores on the ASK-12 total and subscales based on busyness and routine, there is no research indicating what constitutes a clinically meaningful difference on this scale. Researchers may wish to further validate this scale to aid in interpretation of existing and future studies. Additionally, we relied on an existing measure of daily routine and busyness but were unable to assess specific medication related routines, only the stability of general daily patterns. Although all of our associations were in the expected direction, it is not possible for us to know if the older adults in our sample had specific medication-related routines versus a more general daily routine. Future researchers may be interested in teasing apart these influences to understand this relationship more deeply. Lastly, our medication adherence data was collected via self-report as we did not have access to measures of actual medication adherence. However, previous researchers have found significant correlations between the ASK-12 and pharmacy refill data (Takemura et al., 2017).

Conclusion

Nearly half of older adults are not taking medications according to recommendations. Until now, the impact of busyness and routine on medication taking behaviors was unknown. We found that for busy older adults, having a high level of routine may be a protective factor against the detrimental effects of busyness on medication management. Interventions to promote establishing and maintaining routines have potential for improving medication taking behaviors among busy older adults.

What this paper adds:

• Though the terms busyness and routine are ubiquitous in daily life, they are relatively unstudied in the context of completing daily health management activities.

• We demonstrated that the level of daily routine and busyness impacts older adults’ ability to adhere to medication management routines.

Application of study findings:

• This research demonstrates that for busy older adults, those with a stable daily routine have scores on a medication management assessment that indicate better adherence.

• Helping busy older adults establish medication management routines has potential to improve medication adherence.