Trends in Angiotensin-Converting Enzyme Inhibitor and Angiotensin II Receptor Blocker Use among Those with Impaired Kidney Function in the United States

Significance Statement

Although angiotensin-converting enzyme (ACE) inhibitor and angiotensin II receptor blocker (ARB) medications are mainstays of treatment in CKD, the frequency of ACE/ARB use has not been well characterized in this population. Nationally representative data from 1999 to 2014 showed that overall ACE/ARB use during this period was only 34.9% among those with CKD; use rose significantly with time (from 25.5% in 1999–2002 to 40.1% in 2011–2014) but seemed to level off after the early 2000s, a trend similarly observed in multiple CKD subgroups. Overall, regardless of era, ACE/ARB use in CKD was the exception unless concomitant illnesses, like diabetes mellitus or cardiac disease, were present. This suggests that a significant opportunity exists for improvement in the care in community-based CKD.

Visual Abstract

Abstract

Background

Although hypertension is common in CKD and evidence-based treatment of hypertension has changed considerably, contemporary and nationally representative information about use of angiotensin-converting enzyme (ACEs) inhibitors or angiotensin II receptor blockers (ARBs) in CKD is lacking.

Methods

We examined ACE/ARB trends from 1999 to 2014 among 38,885 adult National Health and Nutrition Examination Survey participants with creatinine-based eGFR<60 ml/min per 1.73 m² or urinary albumin-to-creatinine ratio ≥30 mg/g.

Results

Of 7085 participants with CKD, 34.9% used an ACE/ARB. Across four eras studied, rates of use rose significantly (rates were 25.5% in 1999–2002, 33.3% in 2003–2006, 39.0% in 2007–2010, and 40.1% in 2011–2014) but appeared to plateau after 2003. Among those with CKD, use was significantly greater among non-Hispanic white and black individuals (36.1% and 38.2%, respectively) and lower among Hispanic individuals (26.7%) and other races/ethnicities (29.3%). In age-, sex-, and race/ethnicity-adjusted models, ACE/ARB use was significantly associated with era (adjusted odds ratios [aOR], 1.41; 95% confidence interval [95% CI], 1.14 to 1.74 for 2003–2006, 1.84; 95% CI, 1.48 to 2.28 for 2007–2010, and 2.02; 95% CI, 1.61 to 2.53 for 2011–2014 versus 1999–2002); it also was significantly associated with non-Hispanic black versus non-Hispanic white race/ethnicity (aOR, 1.40; 95% CI, 1.19 to 1.66). Other multivariate associations included older age, men, elevated BMI, diabetes mellitus, treated hypertension, cardiac failure, myocardial infarction, health insurance, and receiving medical care within the prior year.

Conclusions

Rates of ACE/ARB use increased in the early 2000s among United States adults with CKD, but for unclear reasons, use appeared to plateau in the ensuing decade. Research examining barriers to care and other factors is needed.

Hypertension, which is highly prevalent in patients with CKD, is associated with cardiovascular outcomes in both the general^1,2 and CKD populations.³ Several recent changes in the therapeutic management of hypertension are relevant in patients with CKD. For example, the Systolic Blood Pressure Intervention Trial (SPRINT) demonstrated that lower BP targets reduced cardiovascular events, primarily driven by reduced risk of cardiac failure, an effect unmodified by the presence of CKD.⁴ More recently, the American Heart Association (AHA) recommended that the threshold to define hypertension should be lowered from 140/90 to 130/80 mm Hg and that individuals with CKD should initiate antihypertensive medication at this level regardless of proteinuria.⁵ The AHA guidelines differ from the recommended Eighth Report of the Joint National Committee (JNC-8) and most recent Kidney Disease Improving Global Outcomes (KDIGO) guidelines, which both predate the SPRINT. Thus, the JNC-8 maintains a uniform threshold of 140/90 mm Hg, regardless of proteinuria,⁶ whereas KDIGO recommends a threshold of 130/80 mm Hg to be treated with angiotensin-converting enzyme (ACE) inhibitor or angiotensin II receptor blocker (ARB) when albuminuria is present or 140/90 mm Hg when it is not.⁷

ACE/ARB therapy is a cornerstone of hypertension and proteinuria management in CKD. ACE/ARB medications have been shown to have apparently salutary effects on markers of kidney function and cardiovascular events in populations with diabetes mellitus and kidney disease^8,9; similar findings have been observed in patients with CKD without diabetes.^10–16 ACE/ARB medications also are first-line therapy in patients with cardiac failure with reduced ejection fraction, a population with a very high prevalence of CKD.^17–19 They are also recommended in select patients with myocardial infarction.^20,21

Given the change in the therapeutic landscape of hypertension and the public health burden of CKD, it seems reasonable to suggest that trends in ACE/ARB use in the United States CKD population should be followed closely. In the United States, ACE/ARB use increased in the general adult population from 1999 to 2012.²² ACE/ARB use from 2001 to 2010 increased among those with hypertension,²³ whereas from 1988 to 2002, ACE use alone increased in those with hypertension and CKD.²⁴ In the absence of information about contemporary practice patterns in the CKD population, we set out to assess ACE/ARB use in the United States population between 1999 and 2014.

Methods

Study Population

The National Health and Nutrition Health Examination Survey (NHANES) is a series of nationally representative cross-sectional surveys conducted by the National Center for Health Statistics among noninstitutionalized adults and children in the United States sampled in a stratified, clustered probability design.²⁵ Surveys are performed continuously and have been documented in 2-year increments since 1999. The survey includes a questionnaire, physical examination, and laboratory data. Our study was limited to those ages ≥20 years old with serum creatinine and urinary albumin-to-creatinine ratio (ACR) measurements surveyed between 1999 and 2014.

Measures of Kidney Function

Kidney function was defined by both creatinine-based eGFR and urinary ACR. The Chronic Kidney Disease Epidemiology Collaboration equation was used to calculated eGFR.²⁶ Calibration of creatinine was performed according to the recommended calibration equations for particular survey years requiring it before the use of an isotope dilution mass spectrometry-traceable reference method.^27–29 Participants were considered to have impaired kidney function if they had eGFR<60 ml/min per 1.73 m² or ACR≥30 mg/g.^30,31 Sensitivity analysis using a threshold of ACR≥100 mg/g did not produce meaningfully different results.

Medication Use, Comorbidities, and Other Characteristics

Prescription medications used within the 30 days before the NHANES assessment were documented during in-home interviews, including by inspection of prescription medication bottles. All medications solely or in combination with other medications containing ACE/ARB according to their coded ingredient category were assessed to determine if the participant was using an ACE/ARB.

Age, sex, race/ethnicity, education (known to have graduated high school versus not), health insurance status (known to be uninsured versus not), income status (known to be at or below poverty level²⁵ versus not), and medical care over the prior year were all assessed by questionnaire. Because of changes in the questionnaire regarding medical care in the prior year, participants were categorized as being known to not have had care in the prior year or not. Diabetes mellitus was defined as present with either hemoglobin A_1c ≥6.5%³² or self-reported history of the disease. Reflecting the fact that all study participants were evaluated before the SPRINT findings were known,⁴ hypertension was defined as present either when the mean of up to three resting BP measurements was ≥140/90 mm Hg or by self-report and further characterized by use of any self-described antihypertensive medication. Cardiac failure and myocardial infarction were defined by self-report.

Statistical Analyses

In this 16-year study, participant statistical weights were determined by dividing the NHANES 2-year Mobile Examination Center statistical weight (“wtmec2yr”) by eight as recommended by the National Center for Health Statistics.^33,34 For our primary analysis, the chi-squared test was used to evaluate the null hypothesis of no change over time in the proportion of individuals with eGFR<60 ml/min per 1.73 m² or ACR≥30 mg/g using ACE/ARB medications. Logistic regression was used to calculate odds ratios for ACE/ARB use adjusting for age, sex, and race/ethnicity. R version 3.4.4 was used for statistical analyses (R Foundation for Statistical Computing, Vienna, Austria).

Results

Characteristics of the general United States study population (n=38,885) are shown in Table 1. For age distribution, 38.1% (SEM=0.5%) were 20–39 years of age, 45.2% (0.4%) were 40–64 years of age, and 16.7% (0.3%) were 65 years of age or older (Table 1); 51.9% (0.2%) were women, 70.0% (1.1%) were of non-Hispanic white race/ethnicity, 10.7% (0.6%) were of non-Hispanic black ethnicity, 13.3% (0.9%) were of Hispanic ethnicity, and 6.0% (0.3%) were of other ethnicities (Table 1). Additionally, 14.1% (0.3%) of the study population had eGFR<60 ml/min per 1.73 m² or ACR≥30 mg/g (Table 1). Overall, 15.1% (0.3%) of the study population used an ACE/ARB (Table 1).

Table 1.

Prevalence of angiotensin-converting enzyme inhibitor or angiotensin II receptor blocker use in the general United States adult population by era, level of kidney function, demographic characteristics, and comorbidities (the National Health and Nutrition Health Examination Survey 1999–2014)

Characteristic	Percentage with Characteristic (SEM)	Percentage Using ACE/ARB (SEM)	P Valuea	Adjusted Odds Ratio (95% CI)b	P Valueb
All		15.1 (0.3)
NHANES era			<0.001
1999–2002	23.0 (0.7)	9.9 (0.5)		1 [Reference]
2003–2006	24.5 (1.0)	14.1 (0.6)		1.43 (1.24 to 1.66)	<0.001
2007–2010	25.6 (0.9)	17.2 (0.7)		1.86 (1.62 to 2.13)	<0.001
2011–2014	26.9 (0.9)	18.3 (0.8)		1.98 (1.71 to 2.30)	<0.001
eGFR, ml/min per 1.73 m2 and ACR, mg/gc			<0.001
eGFR≥60 and ACR<30 (no CKD)	85.9 (0.3)	11.8 (0.3)		1 [Reference]
eGFR≥60 and ACR≥30	7.4 (0.2)	25.1 (1.0)		1.91 (1.69 to 2.16)	<0.001
eGFR=30–59 and ACR<30	4.5 (0.2)	44.2 (1.4)		2.23 (1.95 to 2.54)	<0.001
eGFR=30–59 and ACR≥30	1.6 (0.1)	49.6 (2.5)		2.73 (2.20 to 3.40)	<0.001
eGFR<30 and ACR<30	0.2 (0)	54.7 (4.9)		3.09 (2.03 to 4.71)	<0.001
eGFR<30 and ACR≥30	0.4 (0)	45.3 (3.9)		2.54 (1.83 to 3.52)	<0.001
Age, yr			<0.001
20–39	38.1 (0.5)	2.0 (0.1)		1 [Reference]
40–64	45.2 (0.4)	17.3 (0.4)		10.1 (8.69 to 11.7)	<0.001
≥65	16.7 (0.3)	38.7 (0.8)		30.6 (26.4 to 35.5)	<0.001
Sex			0.21
Men	48.1 (0.2)	15.3 (0.4)		1 [Reference]
Women	51.9 (0.2)	14.8 (0.4)		0.89 (0.83 to 0.95)	<0.001
Race/ethnicity			<0.001
Non-Hispanic white	70.0 (1.1)	16.2 (0.4)		1 [Reference]
Non-Hispanic black	10.7 (0.6)	17.1 (0.5)		1.45 (1.33 to 1.58)	<0.001
Hispanic	13.3 (0.9)	8.9 (0.5)		0.79 (0.69 to 0.91)	0.001
Other	6.0 (0.3)	11.8 (0.9)		0.92 (0.77 to 1.09)	0.33
Health insurance status			<0.001
Insured	84.2 (0.3)	17.6 (0.4)		1 [Reference]
Uninsured	15.8 (0.3)	1.4 (0.2)		0.09 (0.07 to 0.12)	<0.001
Received medical care in last year			<0.001
Yes	84.3 (0.3)	17.6 (0.4)		1 [Reference]
No	15.7 (0.3)	1.3 (0.2)		0.09 (0.07 to 0.12)	<0.001
Income status			<0.001
Above poverty level	86.5 (0.4)	15.5 (0.4)		1 [Reference]
Below or equal to poverty level	13.5 (0.4)	11.9 (0.5)		1.01 (0.92 to 1.10)	0.90
High school graduation			<0.001
No	18.5 (0.5)	17.4 (0.6)		1 [Reference]
Yes	81.5 (0.5)	14.5 (0.3)		0.95 (0.86 to 1.04)	0.25
Body mass index, kg/m2			<0.001
<30	66.2 (0.4)	10.9 (0.3)		1 [Reference]
≥30	33.8 (0.4)	22.9 (0.5)		2.60 (2.41 to 2.80)	<0.001
Diabetes mellitusd			<0.001
No: SR−, A1c<6.5%	90.0 (0.2)	11.4 (0.3)		1 [Reference]
Yes: SR−, A1c≥6.5%	2.0 (0.1)	28.0 (1.8)		1.95 (1.61 to 2.37)	<0.001
Yes: SR+, A1c<6.5%	2.9 (0.1)	51.5 (1.9)		5.21 (4.42 to 6.13)	<0.001
Yes: SR+, A1c≥6.5%	5.2 (0.2)	54.0 (1.3)		6.04 (5.35 to 6.82)	<0.001
Hypertensione			<0.001
No: SR−, BP<140/90	64.5 (0.5)	1.8 (0.1)		1 [Reference]
Yes: SR−, BP≥140/90	6.1 (0.2)	3.7 (0.4)		1.29 (0.98 to 1.68)	0.07
Yes: SR+, BP<140/90, no BP medication	4.1 (0.2)	2.6 (0.5)		1.45 (0.94 to 2.22)	0.09
Yes: SR+, BP≥140/90, no BP medication	1.2 (0.1)	3.8 (1.1)		1.58 (0.87 to 2.85)	0.14
Yes: SR+, BP<140/90, ≥1 BP medication	16.0 (0.4)	58.6 (0.9)		51.0 (43.5 to 59.7)	<0.001
Yes: SR+, BP≥140/90, ≥1 BP medication	8.1 (0.2)	51.3 (1.1)		33.5 (28.2 to 39.8)	<0.001
Cardiac failure			<0.001
No	97.7 (0.1)	14.2 (0.3)		1 [Reference]
Yes	2.3 (0.1)	52.7 (1.8)		3.13 (2.67 to 3.67)	<0.001
Prior myocardial infarction			<0.001
No	96.7 (0.1)	13.9 (0.3)		1 [Reference]
Yes	3.3 (0.1)	48.4 (1.7)		2.69 (2.30 to 3.15)	<0.001

ACE, angiotensin-converting enzyme; ARB, angiotensin II receptor blocker; 95% CI, 95% confidence interval; NHANES, National Health and Nutrition Examination Survey; ACR, albumin-to-creatinine ratio; SR−, no by self-report; SR+, yes by self-report.

^aChi-squared test.

^bLogistic regression adjusted for age, sex, and race/ethnicity.

^ceGFR is by the Chronic Kidney Disease Epidemiology Collaboration equation. Normal kidney function is eGFR≥60 ml/min per 1.73 m² and ACR<30 mg/g.

^dDiabetes mellitus status is defined by self-reported history and hemoglobin A_1c.

^eHypertension status is defined by self-reported history and resting BP (in millimeters of mercury); it is further characterized by use of any self-described antihypertensive medication (BP medication) among those who are SR+.

Characteristics of participants with eGFR<60 ml/min per 1.73 m² or ACR≥30 mg/g (n=7085) are shown in Table 2. Overall, 34.9% (0.8%) of this subgroup used an ACE/ARB medication (Table 2); corresponding values for eGFR<60 ml/min per 1.73 m² (regardless of ACR) and ACR≥30 mg/g (regardless of eGFR) were 45.8% (1.2%) and 30.2% (0.9%), respectively (data not tabulated). ACE/ARB use in the subgroup with eGFR<60 ml/min per 1.73 m² or ACR≥30 mg/g increased over time from 25.5% (1.7%) in 1999–2002 to 33.3% (1.3%) in 2003–2006, 39.0% (1.5%) in 2007–2010, and 40.1% (1.6%) in 2011–2014 (P<0.001) (Table 2). In models adjusted for age, sex, and race/ethnicity, odds ratios for ACE/ARB use were 1.41 (95% confidence interval [95% CI], 1.14 to 1.74 [versus 1999–2002]; P=0.002) for 2002–2006, 1.84 (95% CI, 1.48 to 2.28; P<0.001) for 2007–2010, and 2.02 (95% CI, 1.61 to 2.53; P<0.001) for 2011–2014 (Table 2). Regarding era, similar patterns were present within subgroups defined by eGFR<60 ml/min per 1.73 m², ACR≥30 mg/g, age, sex, race/ethnicity, and the presence of diabetes mellitus or cardiac failure (Figure 1 and Supplemental Figure 1). Other adjusted associations of ACE/ARB use in the population with impaired kidney function included older age, men, non-Hispanic black race/ethnicity, body mass index ≥30 kg/m², awareness of diabetes, awareness and treatment of self-reported hypertension, prior myocardial infarction, cardiac failure, being uninsured, and having had no medical care in the prior year (Table 2).

Table 2.

Prevalence of angiotensin-converting enzyme inhibitor or angiotensin II receptor blocker use in the United States adult population with impaired kidney function by era, demographic characteristics, and comorbidities (the National Health and Nutrition Health Examination Survey 1999–2014)

Characteristic	Percentage with Characteristic (SEM)a	Percentage Using ACE/ARB (SEM)	P Valueb	Adjusted Odds Ratio (95% CI)c	P Valuec
All		34.9 (0.8)
NHANES era			<0.001
1999–2002	22.0 (1.0)	25.5 (1.7)		1 [Reference]
2003–2006	25.1 (1.3)	33.3 (1.3)		1.41 (1.14 to 1.74)	0.002
2007–2010	24.4 (1.0)	39.0 (1.5)		1.84 (1.48 to 2.28)	<0.001
2011–2014	28.5 (1.2)	40.1 (1.6)		2.02 (1.61 to 2.53)	<0.001
Age, yr			<0.001
20–39	15.9 (0.7)	7.0 (1.2)		1 [Reference]
40–64	35.0 (0.7)	31.7 (1.4)		6.06 (4.11 to 8.93)	<0.001
≥65	49.1 (0.9)	46.3 (1.1)		11.7 (8.03 to 16.9)	<0.001
Sex			<0.001
Men	42.5 (0.8)	37.8 (1.1)		1 [Reference]
Women	57.5 (0.8)	32.7 (1.0)		0.81 (0.72 to 0.92)	0.001
Race/ethnicity			<0.001
Non-Hispanic white	70.9 (1.3)	36.1 (1.0)		1 [Reference]
Non-Hispanic black	11.9 (0.8)	38.2 (1.6)		1.40 (1.19 to 1.66)	<0.001
Hispanic	11.5 (1.0)	26.7 (1.8)		1.01 (0.81 to 1.26)	0.93
Other	5.7 (0.5)	29.3 (3.6)		0.97 (0.69 to 1.37)	0.88
Health insurance status			<0.001
Insured	91.4 (0.4)	37.8 (0.8)		1 [Reference]
Uninsured	8.6 (0.4)	3.8 (0.8)		0.09 (0.05 to 0.13)	<0.001
Received medical care in last year			<0.001
Yes	91.5 (0.5)	37.8 (0.8)		1 [Reference]
No	8.5 (0.5)	3.3 (0.8)		0.08 (0.05 to 0.12)	<0.001
Income status			<0.001
Above poverty level	84.3 (0.7)	36.0 (0.9)		1 [Reference]
Below or equal to poverty level	15.7 (0.7)	28.8 (1.4)		0.94 (0.81 to 1.09)	0.43
High school graduation			0.10
No	26.8 (0.8)	36.9 (1.4)		1 [Reference]
Yes	73.2 (0.8)	34.2 (0.9)		1.02 (0.88 to 1.20)	0.78
Body mass index, kg/m2			<0.001
<30	58.7 (0.9)	29.4 (1.0)		1 [Reference]
≥30	41.3 (0.9)	42.6 (1.4)		2.06 (1.76 to 2.43)	<0.001
Diabetes mellitusd			<0.001
No: SR−, A1c<6.5%	72.5 (0.8)	27.3 (0.8)		1 [Reference]
Yes: SR−, A1c≥6.5%	4.2 (0.3)	32.7 (3.1)		1.26 (0.96 to 1.65)	0.10
Yes: SR+, A1c<6.5%	7.2 (0.4)	62.3 (2.4)		3.61 (2.84 to 4.58)	<0.001
Yes: SR+, A1c≥6.5%	16.1 (0.7)	57.9 (2.0)		3.42 (2.79 to 4.18)	<0.001
Hypertensione			<0.001
No: SR−, BP<140/90	33.5 (0.9)	8.6 (1.0)		1 [Reference]
Yes: SR−, BP≥140/90	9.4 (0.4)	9.0 (1.2)		0.76 (0.52 to 1.10)	0.15
Yes: SR+, BP<140/90, no BP medication	2.3 (0.3)	13.1 (3.4)		1.72 (0.94 to 3.14)	0.08
Yes: SR+, BP≥140/90, no BP medication	1.6 (0.2)	15.6 (5.1)		1.48 (0.69 to 3.21)	0.32
Yes: SR+, BP<140/90, ≥1 BP medication	29.5 (0.8)	62.3 (1.5)		13.2 (9.85 to 17.8)	<0.001
Yes: SR+, BP≥140/90, ≥1 BP medication	23.7 (0.6)	56.0 (1.7)		9.94 (7.42 to 13.3)	<0.001
Cardiac failure			<0.001
No	91.7 (0.4)	32.9 (0.8)		1 [Reference]
Yes	8.3 (0.4)	57.3 (2.5)		1.93 (1.57 to 2.38)	<0.001
Prior myocardial infarction			<0.001
No	90.5 (0.4)	32.8 (0.8)		1 [Reference]
Yes	9.5 (0.4)	55.2 (2.3)		1.73 (1.41 to 2.13)	<0.001

ACE, angiotensin-converting enzyme; ARB, angiotensin II receptor blocker; 95% CI, 95% confidence interval; NHANES, National Health and Nutrition Examination Survey; SR−, no by self-report; SR+, yes by self-report.

^aImpaired kidney function defined by eGFR<60 ml/min per 1.73 m² by the Chronic Kidney Disease Epidemiology Collaboration equation or albumin-to-creatinine ratio ≥30 mg/g.

^bChi-squared test.

^cLogistic regression adjusted for age, sex, and race/ethnicity.

^dDiabetes mellitus status is defined by self-reported history and hemoglobin A_1c.

^eHypertension status is defined by self-reported history and resting BP (in millimeters of mercury); it is further characterized by use of any self-described antihypertensive medication (BP medication) among those who are SR+.

Figure 1.

Temporal trends in angiotensin II converting enzyme (ACE) and angiotensin II receptor blocker (ARB) use in CKD-relevant subgroups. Adjusted odds ratio (OR) with 95% confidence interval (95% CI) ACE/ARB use in the United States adult population with impaired kidney function (eGFR <60 mL/min per 1.73 m², by the Chronic Kidney Disease Epidemiology Collaboration equation, or albumin-to-creatinine ratio [ACR] ≥30 mg/g) by era compared with 1999–2002 in subgroups defined by levels of kidney function, age, sex, race/ethnicity, presence of diabetes mellitus (self-report or hemoglobin A_1c ≥6.5%), and presence of cardiac failure. ORs adjusted for age, sex, and race/ethnicity.

No statistically significant change in the distribution of kidney function, defined by both eGFR and ACR, within the United States population comparing 1999–2006 with 2007–2014 was seen (Supplemental Table 1). Further demographic and comorbidity data by NHANES era are shown in Supplemental Table 2.

Discussion

In this study of Unites States adults with CKD, we found that, although use of ACE/ARB increased between 1999 and 2014, it appeared to plateau after 2003, with less than one half of the CKD population using an ACE/ARB. Overall, regardless of era, ACE/ARB use was the exception unless concomitant illnesses, like diabetes mellitus or cardiac disease, were present. It is unclear why this plateau has occurred, and it was similarly observed in multiple CKD subgroups. Among those with reduced eGFR, similar rates of ACE/ARB use were seen regardless of albuminuria. The proportion of participants with preserved eGFR and albuminuria receiving ACE/ARB therapy was particularly low.

The design of this study did not allow us to evaluate the relative contributions of lack of awareness of CKD by individuals and health care systems to its findings. This may include failure to recognize an “abnormal” creatinine measurement if not accompanied by an eGFR. An increased likelihood of identifying CKD may have contributed to increased use of ACE/ARBs whether due to guideline-based assessment of kidney function in diabetes³⁵ and hypertension⁵ or simply due to access to care. Use of any medication self-described as for treating hypertension was associated with ACE/ARB use. Of note, few participants reported a history of hypertension, and they denied use of any antihypertensive medication; however, a subset of them was using an ACE/ARB. It seems reasonable to suggest that they either used these medications for another indication and/or lacked knowledge about their role in reducing BP.

It seems unrealistic to expect that all individuals with CKD and an appropriate indication would use ACE/ARB medications because of reasons like intolerance of typical effects, such as hypotension, hyperkalemia, eGFR decline, and allergies and anaphylactoid reactions beyond issues of access to care. This being said, the greater use of ACE/ARB in the CKD subgroups with diabetes mellitus or cardiac failure suggests that medication intolerance is an inadequate explanation for nonuse of ACE/ARB medications. Such complications are not assessible in this study design.

Although our analyses focused primarily on associations between era and ACE/ARB use in the CKD population overall, several disparities were present. In models that adjusted for age and sex, the likelihood of ACE/ARB use was greater with non-Hispanic blacks than with non-Hispanic whites. We also found a lower likelihood of ACE/ARB use with women. Concerningly, younger adults with CKD were much less likely to use an ACE/ARB than older adults, with 1 in every 14 individuals age 20–39 years old receiving them compared with 1 in 3 for age 40–64 years old and 1 in 2 for age ≥65 years old. Unsurprisingly, ACE/ARB use was markedly lower in those with CKD who were without health insurance or who did not receive medical care in the preceding year. Yet, even among the large majorities of those with CKD who have health insurance and who received care within the prior year, ACE/ARB use was below one in two. Education and income, as other socioeconomic markers, were not predictive of ACE/ARB use in the CKD population after adjusting for demographic differences.

Our study has similarities to a study in patients with hypertension, including those with CKD, from the late 1980s to early 2000s that also examined ACE use.²⁴ In that study, hypertension was defined as BP≥140/90 mm Hg or use of a self-reported antihypertensive medication, a strategy that should lead to larger estimates for ACE use in the population with hypertension, because all participants using ACE would have hypertension by definition. Participants were considered to have CKD with eGFR<60 ml/min per 1.73 m² or ACR≥200 mg/g. Among that population, 32% were seen to be on ACE therapy in 1999–2002. Our results differed by including those without evidence of hypertension in the population with impaired kidney function, including those with ACR between 30 and 200 mg/g, including ARB medications, and examining more contemporary populations. Other studies from this era have shown that ACE/ARB use have risen in the general United States population,²² the population with diabetes mellitus,³⁶ and the population with hypertension.²³

Strengths of this study are the generalizability to the entire United States noninstitutionalized adult population, calibration of creatinine values, standardized BP measurements, and stringent medication survey methods.

Limitations of this study include the availability of single measurements only of creatinine and albuminuria rather than longitudinal measurements. Although participants with personally atypical “impaired” kidney function may be included and those with personally atypical “normal” kidney function may be excluded, it seems reasonable to suggest that differential misclassification by the NHANES period is unlikely, allowing for relative comparisons across eras. Additionally, ACE/ARB medications often decrease eGFR levels, leading to a situation where the study outcome could partially define the study population; this being said, such an effect would lead to an overestimation of ACE/ARB use among those with impaired kidney function measured cross-sectionally. This effect is countered by the reduction in albuminuria expected from ACE/ARB use, potentially biasing against ACE/ARB use in the population with measured impaired kidney function with unknown overall effect on the estimated ACE/ARB use over time. A prior study in the population with diabetic kidney disease showed decreasing albuminuria and decreasing eGFR over time, although such changes may be due to increasing ACE/ARB and other medication use.³⁷ Although the threshold eGFR and urinary ACR levels used to define impaired kidney function are frequently used in other studies and conform to existing guidelines, they are necessarily arbitrary.^30,31 This low threshold for ACR was chosen to reduce risk for misclassification among those with more significant degrees of albuminuria.

It is unclear why ACE/ARB use among those with CKD and among CKD-relevant subgroups has plateaued over the last decade. This study suggests that comprehensive research examining additional barriers to care, identification of CKD, and guideline adherence are needed.

Disclosures

None.

Supplemental Material

This article contains the following supplemental material online at http://jasn.asnjournals.org/lookup/suppl/doi:10.1681/ASN.2018100971/-/DCSupplemental.

Supplemental Figure 1. Temporal trends in angiotensin II converting enzyme (ACE) and angiotensin II receptor blocker (ARB) use in CKD-relevant subgroups. Adjusted odds ratio (OR) with 95% confidence interval (95% CI) and P-value (hover over point-estimate) ACE/ARB use in the United States adult population with impaired kidney function (eGFR <60 mL/min per 1.73 m², by the Chronic Kidney Disease Epidemiology Collaboration equation, or albumin-to-creatinine ratio [ACR] ≥30 mg/g) by era compared with 1999-2002 in subgroups defined by levels of kidney function, age, sex, race/ethnicity, presence of diabetes mellitus (self-report or hemoglobin A_1c ≥6.5%), and presence of cardiac failure. ORs adjusted for age, sex, and race/ethnicity.

Supplemental Table 1. Percentage of levels of kidney function, demographic characteristics, and comorbidities in the general United States adult population in 2007–2014 (versus 1999–2006), NHANES.

Supplemental Table 2. Prevalence of demographic characteristics and comorbidities in the United States adult population with impaired kidney function by era, NHANES 1999–2014.