Sorting Adherence Risk Factors to Shape Better Adherence Care

Abstract: Review of recent studies -- from meta-analyses to cross-sectional cohorts -- indicates more than 30 potential reasons for poor antiretroviral adherence. Many of these risk factors overlap with others, and multivariate analysis cannot always disentwine these overlaps. In addition, some studies disagree on whether certain variables impair or promote adherence. Although people taking antiretrovirals often cite "simply forgetting" as the prime reason for poor adherence, several analyses indicate that forgetting does not explain erratic pill taking for most HIV patients and that memory aids will not improve adherence. Adherence factors amenable to change include alcohol and other substance use, smoking, depression, and antiretroviral side effects. Some well-planned studies indicate that treating depression improves antiretroviral adherence. One study offers persuasive evidence that "nonspecific" side effects and seemingly minor side effects like cough, fatigue, and taste disturbances can imperil adherence. Plentiful research shows that adherence often slips measurably in the early postpartum period.

Potential reasons for poor antiretroviral adherence -- and for good adherence -- seem limited only by the imagination of adherence researchers. A review of metaanalyses, prospective and cross-sectional cohort studies, and randomized controlled trials turned up 32 reasons for good or bad adherence (Table 1). Even grouping these reasons into broad categories still leaves seven: demographic, habitual, behavioral, economic, HIV-related, non-HIV clinical, and psychological. And this breakdown relies on studies published mostly in the past 10 years that typically used statistical analysis beyond descriptive comparisons -- such as multivariate analysis or data pooling.

Table 1. Identified Adherence Promoters (+) and Barriers (-)
Female gender (-)1-4
Postpartum vs pregnancy (-)5-8
Black race (-)9,10
Older age (+/-)9,11,12
Younger age (-)13

Forgetting (-)14,15
Being away from home (-)14,15
Change in daily routine (-)14,15
Being busy (-)15
Alcohol use (-)12,16-18
Other substance use (-)12,14,19-24
Smoking (-)9,12,25
Conduct disorder/disruptive behavior in youth (-)26,27
Caregiver factors* (-)28

Employment (+/-)15,29
Financial constraints (-)21
Food insecurity (-)30,31
Housing instability (-)32,33
Social support (+)21,28
Stress/trauma/violence (-)34-36
Poor understanding of HIV (-)15
More education (+)9,25
Concerns about ART (-)21,24
Negative beliefs about necessity/utility of ART (-)21,24
Antiretroviral side effects (-)4,37,38
CD4 count (+/-)9,25,39,40
Trust/satisfaction with provider (+)21

Non-HIV clinical
Comorbidities (-)2,41
Comedications (+/-)9,13,41
Feeling sick/feeling well (+/-)14

Depressive/psychological symptoms (-)14,17,21,22,26,42-44
Antidepressant therapy (+)10
Self-efficacy (+)4,15,21
HIV stigma (-)14,21

* Caregiver not fully responsible for medications, low caregiver well-being, adolescent perceptions of poor caregiver-youth relations, caregiver perceptions of low social support.28

Certain studies disagree on whether a particular factor promotes or impairs adherence. Most studies, for example, find better adherence with older age and worse adherence with younger age. But secondary analysis of data from 326 participants in a US randomized smoking cessation trial linked older age to worse adherence in these smokers.12 In a 28-study meta-analysis, employment promoted adherence in high-income countries and lowincome countries.29 But an older analysis of US patients at 10 AIDS Clinical Trials Group (ACTG) sites found worse adherence among people employed outside the home.15 Analysis of the international SMART trial9 and the Swiss HIV Cohort Study13 confirmed better adherence in people taking comedications for non-HIV conditions, while a prospective cohort study in Spain linked more comedications to worse adherence.41 People with a higher current CD4 count had worse adherence in a cross-sectional Dutch study,40 but a lower enrollment CD4 count predicted worse adherence in a small study of US patients,25 while an 18-study meta-analysis yielded mixed results on the impact of CD4 tallies.39

Reasons for such contradictions -- and for the vast array of potential adherence variables -- reflect wide differences in study populations, methods, and antiretroviral era. Adherence is easier today than when most regimens hinged on a ritonavir-boosted protease inhibitor or efavirenz . But difficulties in pinpointing discrete adherence promoters and barriers also reflect the confounding overlap of both individual variables and whole categories of variables. As just one example, the demographic variable race may easily overlap behavioral, economic, HIV-related, non-HIV clinical, and psychological barriers (Table 1).

Multivariate analysis can help sort out individual impacts of isolated factors, but no multivariate analysis can control for the dozens of confounders that may cloud results. One resonant example is female gender. HIV-positive women throughout the world endure disadvantages that also imperil adherence (Table 1) including (to name but a few) unemployment, financial constraints, food and housing insecurity, lack of social support, trauma and violence, less education, and HIV stigma. A 29-state 5177-person analysis of adherence in Medicaid recipients 50 to 64 years old found that men were 11% more likely to be adherent than women (adjusted prevalence ratio 1.11, 95% confidence interval [CI] 1.02 to 1.21, P = 0.0127).2 But that analysis adjusted for only age and state, not for the just-listed socioeconomic variables.

Should Adherence Motivators "Forget About Forgetting"?

Some research rates "just forgetting" the primary reason for shaky antiretroviral adherence. A 125-study meta-analysis involving 17,061 adults, 856 adolescents, and 1099 children found forgetting the most frequent patient-reported adherence barrier in adults (41.4%) and adolescents (63.1%), far ahead of the second-place barrier, being away from home (30.4% in adults and 40.7% in adolescents).14 A survey of 75 adult ACTG study participants at 10 trial sites found "simply forgot" the reason for missing doses in 66%, ahead of being away from home (57%) or being busy (53%).15 And multivariate analysis in an 80-person US CHARTER cohort substudy linked worse pharmacy refill adherence to worse working memory on standard tests.45

Despite the frequency of forgetting to take antiretrovirals14,15 and the intuitively compelling tie between faulty memory and poor adherence,45 other research in the United States argues that "just forgetting" does not fundamentally explain erratic pill taking for most people with HIV. This work is important because, if correct, it means clinicians should spend less time promoting adherence memory aids and more time drilling down to underlying causes of poor adherence that need to be addressed. Here are the key findings:

A University of Connecticut group analyzed 556 people with less than 95% antiretroviral adherence according to phone-based unannounced pill count, dividing them into severely nonadherent patients (75% or fewer medications taken) and moderately nonadherent patients (more than 75% to less than 95% taken).46 As in other studies,14,15 forgetting topped the list of reasons for missing doses in severely nonadherent people (54%) and in moderately nonadherent people as well (41%). Reasons related to mental health, structural barriers, and substance use proved much less frequent. High and similar proportions of the severely and moderately nonadherent groups relied on multiple dosing-reminder strategies, such as using bedtime as a cue (67% and 65%), using mealtimes as a cue (61% and 61%), storing medicines where they can be easily seen (58% and 56%), and using pill box organizers (42% and 39%).

Multivariate logistic regression to distinguish severe nonadherence from moderate nonadherence controlled for the four sets of adherence barriers analyzed: cognitive/organizational (which includes forgetting), mental health (such as depression), structural barriers (such as running out of pills or being unable to pay for them), and substance use. Neither cognitive/organizational nor mental health barrier composites distinguished severe from moderate nonadherence. But structural barriers boosted odds of severe nonadherence almost 50% (adjusted odds ratio [aOR] 1.49, 95% CI 1.17 to 1.89, P < 0.05), and substance use raised odds of severe nonadherence by one third (aOR 1.32, 95% CI 1.02 to 1.73, P < 0.01). The researchers suggest that colleagues "forget about forgetting" as a way to improve antiretroviral adherence. Rather, they propose, efforts should "concentrate on substance use treatment and providing case management to resolve structural barriers to adherence."46

In a study of 223 US adults with adherence measured by electronic MEMS pill bottle caps, the more adherence reminders people used, the worse their adherence got.47 The largely male (83%), white (66%) study group completed the Prospective Memory for Medications Questionnaire to assess use of 28 adherence strategies. Participants reported using an average 8.7 strategies at least sometimes. The most frequent memory aids were leaving pill bottles in a prominent place (69% at least sometimes), linking dose times to something done routinely (68%), and thinking about dosing times at the beginning of the day (61%). But statistical analysis linked use of more adherence strategies to worse antiretroviral adherence (-0.15, P = 0.02). The study also found that using more reminder strategies did not reflect patient belief that they would work. Perhaps the researchers discovered a dismal feedback loop in which people use more and more reminders, yet adhere less and less, leading to less belief that strategies work, prompting use of even more reminders (Figure 1, outside arrows). Or maybe the loop runs the other way: people use more and more reminders, yet believe less and less that the strategies work, fulfill their own forecast by taking fewer antiretrovirals, and thus prompt use of even more reminders (Figure 1, inside arrows).

Possible Reminder-Adherence Feedback Loops
Possible Reminder-Adherence Feedback Loops Figure 1. Research linking more adherence reminders to worse adherence -- with no link between using more reminders and believing they work47 -- suggests one of two negative feedback loops in which people use more adherence reminders to no avail.

Analysis of 1496 adults in 11 ACTG studies that ended from 2002 through 2012 also found memory-related pitfalls the most common self-reported reasons for poor adherence at treatment week 12.48 But forgetting ranked far behind other barriers (too many pills, side effects, depression) in explaining a detectable viral load at 24 weeks. The most frequently cited adherence barriers were being away from home (21.9%), simply forgetting (19.6%), a change in daily routine (19.5%), and falling asleep (18.9%). Multivariate analysis adjusted for all 14 potential adherence barriers found one independent predictor of failure to have an undetectable viral load at 24 weeks, "felt sick" (aOR 0.53, 95% CI 0.37 to 0.76, P < 0.001). Two variables marginally predicted virologic failure, "too many pills" (aOR 0.61, 95% CI 0.37 to 1.01, P = 0.06) and "felt drug was harmful" (aOR 0.62, 95% CI 0.37 to 1.04, P = 0.07). "Simply forgot" had no impact on virologic response in bivariate analysis (OR 0.99, 95% CI 0.76 to 1.30, P = 0.95).

Dominance analysis to assess the relative importance of each adherence barrier at 12 weeks in promoting virologic detectability at 24 weeks rated "simply forgot" ninth in importance. (Dominance analysis is a regression-based approach calculating effect size.49,50) The top five barriers were (1) felt sick, (2) too many pills, (3) felt drug was harmful, (4) wanted to avoid side effects, and (5) felt depressed/overwhelmed. The 1-through-14 dominance ranking did not even remotely reflect the 1-through-14 self-reporting frequency (Table 2). The ACTG researchers propose that "interventions should focus on barriers that have been associated with poor virologic outcomes rather than focusing on the most commonly reported barriers."48

Table 2. Adherence Barriers in 11 ACTG Trials Ranked for Impact on Virologic Response vs Self-Reported Frequency
Resistance BarrierRanked Impact on Undetectable Viral Load (24 wk)Patient Self-Reported Frequency (12 wk)
Felt sick16
Too many pills213
Felt drug was harmful312
Wanted to avoid side effects48
Felt depressed/overwhelmed59
Ran out of pills614
Busy with other things75
Taking pills at specified time87
Simply forgot92
Need to hide pill taking1011
Felt good1110
Change in daily routine123
Fell asleep134
Away from home141

Source: Saberi et al.48

Another study used dominance analysis49,50 to rank the importance of nine barriers to self-reported nonadherence (a 4-day interruption) in 1217 US adults (95% men, 76% white, 87% with an undetectable viral load).51 The analysis relied on an online survey to rate self-reported frequency of the nine barriers in this order:

  1. Simply forgot
  2. Day-to-day life
  3. Alcohol or illicit drug use
  4. Felt depressed/overwhelmed
  5. Ran out of pills
  6. Fell asleep
  7. Pharmacy/insurance problems
  8. Wanted to avoid side effects
  9. Felt sick

In the dominance analysis "simply forgot" fell to sixth place while "fell asleep" jumped to first:

  1. Fell asleep
  2. Felt depressed/overwhelmed
  3. Day-to-day life
  4. Wanted to avoid side effects
  5. Alcohol or illicit drug use
  6. Simply forgot
  7. Ran out of pills
  8. Felt sick
  9. Pharmacy/insurance problems

The ACTG researchers suggest that forgetting to take pills "may be ... multi-faceted and may include other barriers such as stigma, depression, drug and alcohol use, and lack of social support."48 They propose that basing adherence strategies solely on patient report frequency "may potentially lead investigators in the wrong direction and may result in ineffective interventions."

Turning Adherence Research Into Clinical Action

Most findings on factors tied to good or bad adherence bear out a priori intuition. Social support,21,28 more education,9,25 and provider satisfaction21 favor good adherence. Alcohol and other substance use,12,14,16-24 food insecurity,30,31 and concerns about ART21,24 foster poor adherence. But research on adherence correlates comes laden with nuance and sprinkled with surprise. The second paragraph of this article sampled contradictory findings related to adherence and age, employment, comedications, and CD4 count. The rest of this article spotlights some fine points clinicians can use to shape clinical adherence strategies.

Adherence Drops After Delivery

Studies of pregnant women in the United States5,8,19 and an international meta-analysis52 agree that antiretroviral adherence during pregnancy drops 10% to 25% after delivery. In the bustle of perinatal and postnatal concerns, even clinicians well aware of this adherence threat may miss the opportunity to prevent a clinically meaningful slip in pill taking. With 14 of 51 studies in the meta-analysis coming from the United States, this study found that a 75.7% pooled prevalence of better than 80% adherence before delivery dropped to 53% postpartum (P = 0.005).52 Barriers to adherence included depression; physical, economic, and emotional stress; alcohol or drug use; and antiretroviral dosing frequency.

A US Pediatric ACTG study involved 519 women, 75% of whom reported perfect adherence (no missed doses in preceding 4 day) during pregnancy.5 Six, 24, and 48 weeks after delivery, the perfect adherence rate dipped to 65%, 64%, and 66% (P < 0.01). Among 149 women in ACTG protocol A5084, 57% reported adherence during pregnancy (no missed doses in past 3 months), and that rate dropped to 45% in the first 12 postpartum weeks (P = 0.03).19 Multivariate analysis determined that women who ever used illicit drugs had almost 6-fold higher odds of nonadherence (P = 0.002), and those who missed prenatal vitamins had almost 5-fold higher odds (P = 0.001). The latter finding suggests vitamin taking during pregnancy may offer a signal of antiretroviral adherence after delivery. In a US Women and Infants Transmission Study of 309 women, the self-reported complete adherence rate fell from 61% during pregnancy to 44% postpartum.8 More health-related symptoms and alcohol use emerged as independent predictors of nonadherence before and after delivery.

Adherence may be better during pregnancy because women want to protect their infant from HIV and because frequent monitoring during pregnancy can promote adherence (Table 3). After delivery, the HIV transmission motivation declines for nonbreastfeeding women, while they confront new demands in caring for an infant and often face postpartum depression. US Department of Health and Human Services (DHHS) guidelines for pregnant women with HIV caution, however, that adherence may also falter during the first trimester because of the nausea and vomiting common during that phase of pregnancy.53 These guidelines recommend more frequent viral load monitoring during pregnancy if adherence is a concern, and they endorse a protease inhibitor over an integrase inhibitor for women at risk of stopping therapy after delivery. For the same reason, switching from twice- to once-daily dosing could make sense.

Table 3. Reasons for Better Adherence During Pregnancy Than Postpartum
  • More frequent monitoring/adherence reinforcement during pregnancy
  • Mother's desire to protect fetus/neonate from HIV
  • Refraining from drug/alcohol use during pregnancy
  • Postpartum challenges in caring for infant
  • Postpartum physiologic changes/health challenges
  • Postpartum depression

"Because the immediate postpartum period poses unique challenges to antiretroviral adherence," the DHHS experts counsel, "arrangements for new or continued supportive services should be made before hospital discharge."53

Depression Undermines Adherence

The well-appreciated link between depression and wavering adherence in people with HIV rests on data from three meta-analyses14,21,44 and countless smaller studies. Meta-analysis of 207 studies presented from 1996 through 2014 determined that depression trailed only current substance use and concerns about ART in predicting poor adherence (Figure 2), but depression had a much greater negative impact than HIV stigma, protease inhibitor therapy, dosing frequency, financial constraints, or pill burden.21

Ranked Predictors of Poor Adherence
Ranked Predictors of Poor Adherence Figure 2. Meta-analysis of 207 studies involving 103,836 people with HIV ranked depressive symptoms as the third strongest independent predictor of poor adherence, with effect size calculated as standard mean difference (SMD).21 (PI, protease inhibitor.)

A 111-study meta-analysis involving 43,366 HIV-positive people found no difference in depression prevalence by country income group.44 Overall chances of attaining at least 80% adherence was 42% lower in people with depressive symptoms (pooled OR 0.58, 95% CI 0.55 to 0.62), and that association did not differ by country income group, study design, or adherence rate.

Depression and Adherence in Youth

Adherence poses a sterner challenge to HIV-positive youth than to older or younger age groups. And depression may play a big part youth's inconsistent pill taking. A 125-study meta-analysis of 17,061 adults, 856 adolescents, and 1099 children found that similar proportions of adults (15.5%) and children (15.1%) self-reported depression as an adherence barrier, rates well below the 25.7% of adolescents attributing poor adherence to depression.14

A US Adolescent Trials Network analysis of 956 minority HIV-positive 16- to 24-year-olds found that 39% of them reported taking fewer than 90% of antiretrovirals in the past 7 days.22 Path analysis determined that higher self-efficacy (belief in one's ability to do something) predicted good adherence, while psychological symptoms (measured on the Brief Symptom Inventory) predicted lower self-efficacy, more substance use, and lower adherence (P < 0.01). A longitudinal study of 294 US youngsters 6 to 17 years old found that 38% had at least one psychiatric condition at the baseline visit.26 Multivariable logistic regression determined that youngsters with depression had 4-fold higher odds of missing more than 5% of doses in the past 3 days at follow-up week 96 (aOR 4.14, 95% CI 1.11 to 15.42).

Treating Depression Improves Adherence

If depression promotes poor adherence, one would expect that treating depression improves adherence. That doesn't always happen, learned US researchers who randomized 304 HIV-positive people with major depressive disorder to antidepressant therapy or usual care.54 After 12 months, lack of improvement in pill count-based adherence with versus without therapy, the authors suggest, could reflect high baseline adherence rates in both study groups (about 86%).

But a clutch of other recent studies, including one meta-analysis55 and one systematic review,56 did find that treating depression improves adherence in people with HIV. The meta-analysis considered 29 studies published between 2001 and 2012 involving antidepressant therapy, cognitive behavioral therapy, and mixed approaches.55 Meta-analysis indicated that odds of antiretroviral adherence were 83% better (standardized OR 1.83, 95% CI 1.27 to 2.55) in people treated for depression or psychological distress than in untreated people. In 17 intervention studies, people randomized to the intervention arm versus the control arm had twice high odds of improvement in depressive symptoms (standardized OR 2.07, 95% CI 1.38 to 3.30). Longer treatment yielded greater adherence rates than shorter treatments (random effects r = 0.43, P = 0.02). The systematic review found that 7 of 9 studies produced evidence that antidepressant treatment improved antiretroviral adherence.56

In a study of 7034 antiretroviral-treated Medicaid recipients, 66% were black, 47% experienced depression during the study period, and 32% had optimal adherence, defined as at least 90% adherence by prescription refill.10 Multivariate logistic regression determined that black participants had 30% lower odds of optimal adherence (aOR 0.70, 95% CI 0.63 to 0.78), while antidepressant therapy nearly doubled chances of optimal adherence (aOR 1.92, 95% CI 1.12 to 3.29).

Retrospective analysis of 3359 HIV-positive people in the Kaiser Permanente healthcare system in 8 states used medical records to determine that 1398 (42%) had a depression diagnosis, yet only 508 of these 1398 (36%) had a prescription for a selective serotonin reuptake inhibitor (SSRI).57 Chances of at least 90% antiretroviral adherence (determined by pharmacy refills) were almost 20% lower in people with depression not treated by SSRIs than in people without depression (aOR 0.81, 95% CI 0.70 to 0.98, P = 0.03). But chances of at least 90% antiretroviral adherence did not differ significantly between people without depression and (1) people with depression and prescribed an SSRI (aOR 0.91, 95% CI 0.72 to 1.15) or (2) people with depression and greater than 80% adherence to an SSRI (aOR 1.13, 95% CI 0.86 to 1.49).

US depression researchers calculated that only 48% of HIV-positive people with major depressive disorder get recognized clinically, only 18% get treated, only 7% get treated adequately, and only 5% achieve remission through treatment.58 See the Summer 2016 issue of RITA for a detailed review of depression in people with HIV.

Alcohol, Hard Drugs and Toxicity Beliefs

Alcohol use and drug use are well-studied predictors of poor antiretroviral adherence. Systematic review of 7 prospective observational cohorts, 11 cross-sectional analyses, and 2 randomized controlled trials found links between alcohol use disorder and poor adherence in most of these studies.16 The researchers site evidence that alcohol abuse and faulty adherence have a synergistic impact on morbidity and mortality, calling the literature on this interface "staggering."

In the United States, alcohol use affects adherence in women as well as men with HIV. Analysis of 1304 HIV-positive women in the Women's Interagency HIV Study (WIHS), 82% of them black or Hispanic, recorded low alcohol drinking in 28%, moderate to heavy drinking in 8%, and binge drinking in 4%.17 More than one third of women (37%) had symptoms of depression. The researchers defined decreasing adherence as 100% self-reported adherence at one visit followed by less adherence at a subsequent visit. Multivariate logistic regression determined that all three drinking levels (compared with no drinking) independently predicted decreasing adherence in women (Figure 3).

Impact of Drinking on Falling Adherence in Women
Impact of Drinking on Falling Adherence in Women Figure 3. Analysis of 1304 antiretroviral-treated US women in the Women's Interagency HIV Study (WIHS) linked low, moderate-heavy, and binge drinking to decreasing adherence when compared with no drinking.17 (Values are adjusted odds ratios; 95% confidence intervals for low drinking 1.06 to 1.57, for moderate to heavy drinking 1.11 to 2.07, and for binge drinking 1.20 to 2.72).

A prospective study of 559 men and 84 women enrolled in an ACTG trial linked hard drug use (cocaine, amphetamines, or heroin) to poor adherence and AIDS progression or death.23 The researchers defined nonadherence as missed antiretroviral doses in the 48 hours before study visits. The ACTG team collected data from October 1997 through April 2007, and median follow-up measured 6 years. High proportions of participants reported ever using cocaine (39%), amphetamines (24%), or heroin (10%), while 8% called themselves heavy drinkers and 32% binge drinkers. Generalized estimating equation models determined that hard drug use doubled the odds of nonadherence (aOR 2.14, 95% CI 1.36 to 3.38, P < 0.001), while binge drinking upped the odds about 50% (aOR 1.53, 95% CI 1.21 to 1.92, P < 0.01). Time-updated nonadherence almost doubled the odds of a new AIDS condition or death in a Cox proportional hazards model (adjusted hazard ratio 1.84, 95% CI 1.15 to 2.94, P = 0.01).

People who drink alcohol59,60 or use drugs61 may interrupt their antiretrovirals because they incorrectly fear interactive toxicity between the alcohol or drugs and their antiretrovirals. In one study 57 men and women reported alcohol use over 45 days by text messaging, while researchers recorded antiretroviral adherence by the Wisepill device, a wireless pill container.59 Participants who believed alcohol interacts with antiretrovirals missed doses on significantly more days. Multivariate regression determined that interactive toxicity beliefs predicted daily missed doses regardless of depression, general medication concerns, or the amount of alcohol consumed. In an interview in this issue, Seth Kalichman stresses that mixing alcohol and antiretrovirals poses risks only in people with a compromised liver (see "Correcting Mistakes and Misperceptions in Managing Antiretroviral Adherence").

When Good Adherence Goes Up in Smoke

Studies disagree on whether smoking cigarettes independently predicts poor adherence in people with HIV. Sorting out the independent impact of smoking is complicated because so many people with HIV smoke and so many smokers have other addictive habits -- and socioeconomic traits -- that also promote poor adherence. Even robust multivariate analyses cannot hope to adjust for all adherence-related risks, some of which may remain unknown to the investigators.

Two recent studies could not confirm an independent link between cigarette smoking and poor antiretroviral adherence62,63 and four studies could.9,12,25,64 The two studies that could not link smoking to poor adherence have their limits. One study focused only on men who have sex with men who report heavy drinking -- itself a notable adherence risk factor.62 Almost half of the 185 study participants smoked. A significantly higher fraction of current smokers had imperfect adherence (37% versus 22%, P < 0.05). But multivariate regression singled out only lower education as an independent predictor of faulty adherence.

The second inconclusive study focused on 203 HIV-positive people enrolled in a US trial to test adherence interventions, which did not improve adherence.63 So these 87 daily smokers, 48 occasional smokers, and 68 nonsmokers had tough adherence problems. One third had a monthly household income below $500, 43% recently used drugs, and 56% drank alcohol. Proportions of antiretroviral doses taken at five intervals up to 48 weeks (measured by MEMS caps) did not differ significantly between daily smokers, occasional smokers, and nonsmokers. Longitudinal generalized estimating equation analysis found no association between smoking status and adherence over time.

Four recent studies that did pinpoint smoking as an independent predictor of flawed adherence are (1) a 5295-person analysis of the international randomized SMART trial (current smoking aOR 1.54, 95% CI 1.41 to 1.68, P < 0.0001),9 (2) a 326-person US randomized smoking-cessation trial (nicotine dependence aOR 1.13, 95% CI 1.02 to 1.25, P = 0.023),12 (3) a 64-person prospective US study using MEMS monitoring (current smoking P = 0.001),25 and (4) a 218-person Belgian study in which smoking, neurocognitive complaints, and female sex independently predicted nonadherence.64

"Mild" Side Effects Can Thwart Adherence

Clinical trials of new antiretrovirals typically find most side effects are "mild or moderate," but research shows that even side effects managed by pat-on-theback counseling can have a big impact on antiretroviral adherence. Teaming with a British research firm, Merck investigators conducted a meta-analysis of studies weighing the impact of specific and nonspecific treatment-related adverse events on adult adherence to antiretroviral therapy (Figure 4).38 Focusing on 18 antiretroviral side effects in 19 studies, the researchers found significantly lower pooled odds of adherence in people who endured "nonspecific" adverse events than in those with no adverse events (pooled OR 0.62, 95% CI 0.46 to 0.83, P = 0.001). Patients reporting certain specific but seemingly second-tier side effects proved significantly less likely to adhere to ART: cough (pooled OR 0.65, 95% CI 0.53 to 0.79, P < 0.001), fatigue (OR 0.63, 95% CI 0.43 to 0.92, P = 0.016), anxiety (OR 0.63, 95% CI 0.41 to 0.95, P = 0.028), confusion (OR 0.35, 95% CI 0.18 to 0.66, P = 0.001), taste disturbance (OR 0.49, 95% CI 0.30 to 0.77, P = 0.003), nausea (OR 0.57, 95% CI 0.43 to 0.77, P < 0.001), vomiting (OR 0.49, 95% CI 0.24 to 1.02, P = 0.056), and tingling in mouth or tongue (OR 0.67, 95% CI 0.42 to 1.05, P = 0.079).

Minor Antiretroviral Side Effects That Hurt Adherence
Minor Antiretroviral Side Effects That Hurt Adherence Figure 4. Even antiretroviral side effects typically considered "mild" or "minor" can impair adherence, according to results of a 19-study meta-analysis. (Source: Al-Dakkak I, et al.38 See text for pooled odds ratios. Clinical images from Servier PowerPoint Image Bank.)

Dermatologic side effects and lipodystrophy did not dim adherence in the 19-study meta-analysis.38 But a 39-article systematic review identified 16 studies (41%) that found visually noticeable antiretroviral side effects contributed to poor adherence (weight gain/loss, excess sweating, darkening skin color, body odor, hair loss, rash).37 Seven studies (18%) produced evidence that psychological adverse reactions impaired antiretroviral adherence. Six studies (15%) found that neuropsychological side effects in people taking efavirenz weakened adherence. Overall 33 of 39 articles (85%) reported that adverse antiretroviral drug reactions impair adherence, while 6 did not.

A cross-sectional study of the national Swedish HIV cohort also linked side effects to adherence.4 The analysis involved 2846 HIV-positive people who completed a 9-item health questionnaire from 2012 through 2014. Two thirds of participants were men, and heterosexual HIV transmission was more common than male-to-male transmission (49% versus 40%). Logistic regression modeling determined that chances of adherence (missing no doses in the past week) were almost twice higher in people reporting no side effects (aOR 1.79, 95% CI 1.38 to 2.31).

Low Perception of Side Effects as Adherence Barrier, but High Impact

AIDS Clinical Trials Group (ACTG) investigators studying 1496 people enrolled in 11 clinical trials found that side effects ranked low as a patient-reported barrier to adherence.48 But dominance analysis (a regression-based approach calculating effect size49,50) determined that side effects have a big impact on viral load. Study participants reported adherence barriers after the first 12 weeks of therapy, and researchers measured viral load at week 24. On a list of 14 potential self-reported barriers to resistance, wanting to avoid side effects ranked eighth in patient-reported frequency and feeling an antiretroviral was harmful ranked twelfth (Table 2). But the dominance analysis rated wanting to avoid side effects the fourth most important factor in affecting week-24 viral load, while feeling an antiretroviral was harmful came in third in that analysis (Table 2).

Keys to Counseling on Side Effects

People starting antiretrovirals cope with manageable side effects better if they know what to expect, and as a result they have a better shot at good adherence. That guidance emerged from the systematic review of 39 studies exploring how antiretroviral side effects shape adherence.37 "When people receiving ART had prior knowledge of antiretroviral adverse drug reactions, or had been informed in advance of potential adverse reactions and how to manage them," four studies showed, "they felt these reactions were expected, normal and manageable, and therefore reported less non-adherence."37 Patients had better adherence when armed with coping strategies for antiretroviral side effects -- like drinking fluids and resting to reduce dizziness or taking pills with a snack or meal to avoid nausea.


  1. Abara WE, Xu J, Adekeye OA, Rust G. The association between combination antiretroviral adherence and AIDS-defining conditions at HIV diagnosis. AIDS Care. 2016;28:1013-1017.
  2. Abara WE, Adekeye OA, Xu J, Heiman HJ, Rust G. Correlates of combination antiretroviral adherence among recently diagnosed older HIV-infected adults between 50 and 64 years. AIDS Behav. 2016;20:2674-2681.
  3. Puskas CM, Kaida A, Miller CL, et al. The adherence gap: a longitudinal examination of men's and women's antiretroviral therapy adherence in British Columbia, 2000-2014. AIDS. 2017;31:827-833.
  4. Marrone G, Mellgren A, Eriksson LE, Svedhem V. Adherence, treatment outcome and satisfaction with care using a nine-item health questionnaire in InfCareHIV. PLoS One. 2016;11:e0156916.
  5. Bardeguez AD, Lindsey JC, Shannon M, et al. Adherence to antiretrovirals among US women during and after pregnancy. J Acquir Immune Defic Syndr. 2008;48:408-417.
  6. Gertsch A, Michel O, Locatelli I, et al. Adherence to antiretroviral treatment decreases during postpartum compared to pregnancy: a longitudinal electronic monitoring study. AIDS Patient Care STDS. 2013;27:208-210
  7. Huntington S, Thorne C, Newell ML, et al. The risk of viral rebound in the year after delivery in women remaining on antiretroviral therapy. AIDS. 2015;29:2269-2278.
  8. Mellins CA, Chu C, Malee K, et al. Adherence to antiretroviral treatment among pregnant and postpartum HIV-infected women. AIDS Care. 2008;20:958-968.
  9. O'Connor JL, Gardner EM, Mannheimer SB, et al. Factors associated with adherence amongst 5295 people receiving antiretroviral therapy as part of an international trial. J Infect Dis. 2013;208:40-49.
  10. Kong MC, Nahata MC, Lacombe VA, Seiber EE, Balkrishnan R. Association between race, depression, and antiretroviral therapy adherence in a low-income population with HIV infection. J Gen Intern Med. 2012;27:1159-1164.
  11. Ghidei L, Simone MJ, Salow MJ, et al. Aging, antiretrovirals, and adherence: a meta analysis of adherence among older HIV-infected individuals. Drugs Aging. 2013;30:809-819.
  12. King RM, Vidrine DJ, Danysh HE, et al. Factors associated with nonadherence to antiretroviral therapy in HIV-positive smokers. AIDS Patient Care STDS. 2012;26:479-485.
  13. Glass TR, Battegay M, Cavassini M, et al. Longitudinal analysis of patterns and predictors of changes in self-reported adherence to antiretroviral therapy: Swiss HIV Cohort Study. J Acquir Immune Defic Syndr. 2010;54:197-203.
  14. Shubber Z, Mills EJ, Nachega JB, et al. Patient-reported barriers to adherence to antiretroviral therapy: a systematic review and meta-analysis. PLoS Med. 2016;13:e1002183.
  15. Chesney MA, Ickovics JR, Chambers DB, et al. Self-reported adherence to antiretroviral medications among participants in HIV clinical trials: the AACTG adherence instruments. Patient Care Committee & Adherence Working Group of the Outcomes Committee of the Adult AIDS Clinical Trials Group (AACTG). AIDS Care. 2000;12:255-266.
  16. Azar MM, Springer SA, Meyer JP, Altice FL. A systematic review of the impact of alcohol use disorders on HIV treatment outcomes, adherence to antiretroviral therapy and health care utilization. Drug Alcohol Depend. 2010;112:178-193.
  17. Lazo M, Gange SJ, Wilson TE, et al. Patterns and predictors of changes in adherence to highly active antiretroviral therapy: longitudinal study of men and women. Clin Infect Dis. 2007;45:1377-1385.
  18. Kalichman SC, Grebler T, Amaral CM, et al. Viral suppression and antiretroviral medication adherence among alcohol using HIV-positive adults. Int J Behav Med. 2014;21:811-820.
  19. Cohn SE, Umbleja T, Mrus J, Bardeguez AD, Andersen JW, Chesney MA. Prior illicit drug use and missed prenatal vitamins predict nonadherence to antiretroviral therapy in pregnancy: adherence analysis A5084. AIDS Patient Care STDS. 2008;22:29-40.
  20. Gross IM, Hosek S, Richards MH, Fernandez MI. Predictors and profiles of antiretroviral therapy adherence among African American adolescents and young adult males living with HIV. AIDS Patient Care STDS. 2016;30:324-338.
  21. Langebeek N, Gisolf EH, Reiss P, et al. Predictors and correlates of adherence to combination antiretroviral therapy (ART) for chronic HIV infection: a meta-analysis. BMC Med. 2014;12:142.
  22. MacDonell KK, Jacques-Tiura AJ, Naar S, Fernandez MI, ATN 086/106 Protocol Team. Predictors of self-reported adherence to antiretroviral medication in a multisite study of ethnic and racial minority HIV-positive youth. J Pediatr Psychol. 2016;41:419-428.
  23. Cohn SE, Jiang H, McCutchan JA, et al. Association of ongoing drug and alcohol use with non-adherence to antiretroviral therapy and higher risk of AIDS and death: results from ACTG 362. AIDS Care. 2011;23:775-785.
  24. Kalichman S, Kalichman MO, Cherry C. Medication beliefs and structural barriers to treatment adherence among people living with HIV infection. Psychol Health. 2016;31:383-395.
  25. Shuter J, Bernstein SL. Cigarette smoking is an independent predictor of nonadherence in HIV-infected individuals receiving highly active antiretroviral therapy. Nicotine Tob Res. 2008;10:731-736.
  26. Kacanek D, Angelidou K, Williams PL, Chernoff M, Gadow KD, Nachman S; International Maternal Pediatric Adolescent AIDS Clinical Trials Group (IMPAACT) P1055 Study Team. Psychiatric symptoms and antiretroviral nonadherence in US youth with perinatal HIV: a longitudinal study. AIDS. 2015;29:1227-1237.
  27. Walkup J, Akincigil A, Bilder S, Rosato NS, Crystal S. Psychiatric diagnosis and antiretroviral adherence among adolescent Medicaid beneficiaries diagnosed with human immunodeficiency virus/acquired immunodeficiency syndrome. J Nerv Ment Dis. 2009;197:354-361.
  28. Naar-King S, Montepiedra G, Garvie P, et al. Social ecological predictors of longitudinal HIV treatment adherence in youth with perinatally acquired HIV. J Pediatr Psychol. 2013;38:664-674.
  29. Nachega JB, Uthman OA, Peltzer K, et al. Association between antiretroviral therapy adherence and employment status: systematic review and meta-analysis. Bull World Health Organ. 2015;93:29-41.
  30. Singer AW, Weiser SD, McCoy SI. Does food insecurity undermine adherence to antiretroviral therapy? A systematic review. AIDS Behav. 2015;19:1510-1526.
  31. Kalichman SC, Grebler T. Stress and poverty predictors of treatment adherence among people with low-literacy living with HIV/AIDS. Psychosom Med. 2010;72:810-816.
  32. Harris RA, Xue X, Selwyn PA. Housing stability and medication adherence among HIV-positive individuals in antiretroviral therapy: a meta-analysis of observational studies in the United States. J Acquir Immune Defic Syndr. 2017;74:309-3317.
  33. Aidala AA, Wilson MG, Shubert V, et al. Housing status, medical care, and health outcomes among people living with HIV/AIDS: a systematic review. Am J Public Health. 2016;106(1):e1-e23.
  34. Mugavero MJ, Raper JL, Reif S, et al. Overload: impact of incident stressful events on antiretroviral medication adherence and virologic failure in a longitudinal, multisite human immunodeficiency virus cohort study. Psychosom Med. 2009;71:920-926.
  35. O'Donnell JK, Gaynes BN, Cole SR, et al. Stressful and traumatic life events as disruptors to antiretroviral therapy adherence. AIDS Care. 2017;28:1-8.
  36. Quinn K, Voisin DR, Bouris A, Schneider J. Psychological distress, drug use, sexual risks and medication adherence among young HIV-positive black men who have sex with men: exposure to community violence matters. AIDS Care. 2016;28:866-872.
  37. Li H, Marley G, Ma W, et al. The role of ARV associated adverse drug reactions in influencing adherence among HIV-infected individuals: a systematic review and qualitative meta-synthesis. AIDS Behav. 2017;21:341-351.
  38. Al-Dakkak I, Patel S, McCann E, Gadkari A, Prajapati G, Maiese EM. The impact of specific HIV treatment-related adverse events on adherence to antiretroviral therapy: a systematic review and meta-analysis. AIDS Care. 2013;25:400-414.
  39. Bock P, James A, Nikuze A, et al. Baseline CD4 count and adherence to antiretroviral therapy: a systematic review and meta-analysis. J Acquir Immune Defic Syndr. 2016;73:514-521.
  40. Pasternak AO, de Bruin M, Bakker M, Berkhout B, Prins JM. High current CD4+ T cell count predicts suboptimal adherence to antiretroviral therapy. PLoS One. 2015;10:e0140791.
  41. Cantudo-Cuenca MR, Jiménez-Galán R, Almeida-Gonzalez CV, Morillo-Verdugo R. Concurrent use of comedications reduces adherence to antiretroviral therapy among HIV-infected patients. J Manag Care Spec Pharm. 2014;20:844-850.
  42. Tedaldi EM, van den Berg-Wolf M, Richardson J, et al. Sadness in the SUN: using computerized screening to analyze correlates of depression and adherence in HIV-infected adults in the United States. AIDS Patient Care STDS. 2012;26:718-729.
  43. Tyer-Viola LA, Corless IB, Webel A, Reid P, Sullivan KM, Nichols P; International Nursing Network for HIV/AIDS Research. Predictors of medication adherence among HIV-positive women in North America. J Obstet Gynecol Neonatal Nurs. 2014;43:168-178.
  44. Uthman OA, Magidson JF, Safren SA, Nachega JB. Depression and adherence to antiretroviral therapy in low-, middle- and high-income countries: a systematic review and meta-analysis. Curr HIV/AIDS Rep. 2014;11:291-307.
  45. Andrade AS, Deutsch R, Celano S, et al. Relationships among neurocognitive status, medication adherence measured by pharmacy refill records, and virologic suppression in HIV-infected persons. J Acquir Immune Defic Syndr. 2013;62:282-292.
  46. Kalichman SC, Kalichman MO, Cherry C. Forget about forgetting: structural barriers and severe non-adherence to antiretroviral therapy. AIDS Care. 2017;29:418-422.
  47. Blackstone K, Woods SP, Weber E, Grant I, Moore DJ; HNRP Group. Memory-based strategies for antiretroviral medication management: an evaluation of clinical predictors, adherence behavior awareness, and effectiveness. AIDS Behav. 2013;17:74-85.
  48. Saberi P, Neilands TB, Vittinghoff E, Johnson MO, Chesney M, Cohn SE. Barriers to antiretroviral therapy adherence and plasma HIV RNA suppression among AIDS Clinical Trials Group study participants. AIDS Patient Care STDS. 2015;29:111-116.
  49. Azen R, Traxel NM. Using dominance analysis to determine predictor importance in logistic regression. J Edu Behav Statistics. 2009;34:319-347.
  50. Budescu DV. Dominance analysis: A new approach to the problem of relative importance of predictors in multiple regression. Psychol Bull. 1993;114:542-551.
  51. Sauceda JA, Neilands TB, Johnson MO, Saberi P. An update on the barriers to adherence and a definition of self-report non-adherence given advancements in antiretroviral therapy (ART). AIDS Behav. 2017 Mar 28. doi: 10.1007/s10461-017-1759-9.
  52. Nachega JB, Uthman OA, Anderson J, et al. Adherence to antiretroviral therapy during and after pregnancy in low-income, middle-income, and high-income countries: a systematic review and meta-analysis. AIDS. 2012;26:2039-2052.
  53. US Department of Health and Human Services. Recommendations for use of antiretroviral drugs in pregnant HIV-1-infected women for maternal health and interventions to reduce perinatal HIV transmission in the United States. October 2016.
  54. Pence BW, Gaynes BN, Adams JL, et al. The effect of antidepressant treatment on HIV and depression outcomes: results from a randomized trial. AIDS. 2015;29:1975-1986.
  55. Sin NL, DiMatteo MR. Depression treatment enhances adherence to antiretroviral therapy: a meta-analysis. Ann Behav Med. 2014;47:259-269.
  56. Springer SA, Dushaj A, Azar MM. The impact of DSM-IV mental disorders on adherence to combination antiretroviral therapy among adult persons living with HIV/AIDS: a systematic review. AIDS Behav. 2012;16:2119-2143.
  57. Horberg MA, Silverberg MJ, Hurley LB, et al. Effects of depression and selective serotonin reuptake inhibitor use on adherence to highly active antiretroviral therapy and on clinical outcomes in HIV-infected patients. J Acquir Immune Defic Syndr. 2008;47:384-390.
  58. Pence BW, O'Donnell JK, Gaynes BN. Falling through the cracks: the gaps between depression prevalence, diagnosis, treatment, and response in HIV care. AIDS. 2012;26:656-658.
  59. Pellowski JA, Kalichman SC, Kalichman MO, Cherry C. Alcohol-antiretroviral therapy interactive toxicity beliefs and daily medication adherence and alcohol use among people living with HIV. AIDS Care. 2016;28:963-970.
  60. Kalichman SC, Amaral CM, White D, et al. Alcohol and adherence to antiretroviral medications: interactive toxicity beliefs among people living with HIV. J Assoc Nurses AIDS Care. 2012;23:511-520.
  61. Kalichman SC, Kalichman MO, Cherry C, et al. Intentional medication nonadherence because of interactive toxicity beliefs among HIV-positive active drug users. J Acquir Immune Defic Syndr. 2015;70:503-509.
  62. Cioe PA, Gamarel KE, Pantalone DW, Monti PM, Mayer KH, Kahler CW. Cigarette smoking and antiretroviral therapy (ART) adherence in a sample of heavy drinking HIV-infected men who have sex with men (MSM). AIDS Behav. 2017;21:1956-1963.
  63. Moreno JL, Catley D, Lee HS, Goggin K. The relationship between ART adherence and smoking status among HIV+ individuals. AIDS Behav. 2015;19:619-625.
  64. Degroote S, Vogelaers D, Vermeir P, et al. Determinants of adherence in a cohort of Belgian HIV patients: a pilot study. Acta Clin Belg. 0 02014;69:111-115.

[Note from This article was originally published by The Centfor for AIDS Information and Advocay in September, 2017. We have cross-posted it with their permission.]