Associations of dietary fat with albuminuria and kidney dysfunction^,,

Abstract

Background: Diet represents a potentially important target for intervention in nephropathy, yet data on this topic are scarce.

Objectives: The objective was to investigate associations between dietary fats and early kidney disease.

Design: We examined cross-sectional associations between dietary fats and the presence of high albuminuria (an established independent predictor of kidney function decline, cardiovascular disease, and mortality) or estimated glomerular filtration rate (eGFR) <60 mL ⋅ min⁻¹ ⋅ 1.73 m⁻² at baseline in 19,256 participants of the REGARDS (Reasons for Geographic and Racial Differences in Stroke) study, an ongoing cohort study in US adults aged ≥45 y at time of enrollment. We used logistic regression to assess associations between quintiles of total fat and subtypes of dietary fat (saturated, monounsaturated, polyunsaturated, and trans fat) and presence of high albuminuria or eGFR <60 mL ⋅ min⁻¹ ⋅ 1.73 m⁻².

Results: After multivariable adjustment, only saturated fat intake was significantly associated with high albuminuria [for quintile 5 compared with quintile 1, odds ratio (OR): 1.33; 95% CI: 1.07, 1.66; P for trend = 0.04]. No significant associations between any type of fat and eGFR <60 mL · min⁻¹ · 1.73 m⁻² were observed. ORs between the highest quintile of saturated fat and eGFR <60 mL · min⁻¹ · 1.73 m⁻² varied by race with a borderline significant interaction term (ORs: 1.24 in whites compared with 0.74 in blacks; P for interaction = 0.05) in multivariable-adjusted models, but no other associations were significantly modified by race or diabetes status.

Conclusion: Higher saturated fat intake is significantly associated with the presence of high albuminuria, but neither total nor other subtypes of dietary fat are associated with high albuminuria or eGFR <60 mL · min⁻¹ · 1.73 m⁻².

INTRODUCTION

Although a number of experimental animal models have suggested that hyperlipidemia is associated with progressive kidney failure (1), possibly through glomerular atherosclerosis leading to focal segmental glomerulosclerosis and concurrent tubulo-interstitial damage (2, 3), data remain sparse on the role of dietary fat intake on kidney disease. Specifically, the effects of different types of dietary fat on albuminuria and estimated glomerular filtration rate (eGFR), 2 widely used measures of kidney disease, have not been well explored in human studies. Dietary fat intake is a potentially modifiable behavior that could be a target for intervention in nephropathy.

In animal models, dietary fats may significantly influence kidney structure and function. Dietary administration of polyunsaturated or monounsaturated fatty acids has been reported to decrease glomerulosclerosis, glomerular enlargement, and glomeruli loss in spontaneously hypertensive rats (4). In contrast, Wistar rats fed a high-saturated animal fat diet of lard and egg yolk had decreased glomerular number and size (and greatest increase in blood pressure) when compared with rats fed a high-fat diet of canola oil, which is high in monounsaturated fat (5). In aging rats fed canola oil, fish oil, or butter, eGFR decline was greatest in those receiving butter, perhaps because of a higher percentage of saturated fat in butter compared with canola or fish oil (6).

Investigations of dietary fats and kidney function in humans, especially in population-based studies, are very limited. We investigated associations between dietary fats and high albuminuria or presence of eGFR <60 mL · min⁻¹ · 1.73 m⁻² [widely considered to be moderate or worse kidney dysfunction (7)] in the Reasons for Geographic and Racial Differences in Stroke (REGARDS) Study, a large ongoing prospective cohort study of white and black US adults. The parent study was designed to study why blacks have a higher risk of stroke than whites and why this varies by geographic location. Specifically, we hypothesized that dietary intake of total, saturated, and trans fat would be directly related to presence of high albuminuria and a higher prevalence of eGFR <60 mL · min⁻¹ · 1.73 m⁻², whereas these relations would be reversed for monounsaturated and polyunsaturated fat intake.

SUBJECTS AND METHODS

Participants

REGARDS participants were selected from a commercial, nationwide list of >250 million individuals in the United States (Genesys Inc, Daly City, CA) (8, 9). Between February 2003 and October 2007, 30,239 individuals (42% black, and 55% female) were recruited, with ≈56% of the cohort from the “stroke belt” (North Carolina, South Carolina, Georgia, Tennessee, Alabama, Mississippi, Arkansas, and Louisiana). Individuals were identified from commercially available lists of residents and recruited by using an initial mailing followed by telephone contact. For each household selected, one resident aged ≥45 y was randomly screened for eligibility. Exclusion criteria were race other than black or white, active treatment of cancer, medical conditions preventing long-term participation, cognitive impairment as judged by the telephone interviewer, residence in or on a waiting list for a nursing home, and inability to communicate in English. Verbal informed consent, date of birth, and a medical history were collected.

Afterward, participants underwent an in-home visit (Examination Management Systems Incorporated, Irving, TX) in the morning for phlebotomy, urine collection, blood pressure measurement, anthropomorphic measures, resting electrocardiograms (ECGs), and written informed consent. Phlebotomy was performed by trained personnel using standardized procedures. Samples were collected after patients had fasted for 10–12 h. Within 2 h of collection (mean: 97 min; SD: 127 min), samples were centrifuged and serum or plasma separated and shipped overnight on ice packs to the University of Vermont, Burlington, VT. In samples obtained from study participants, overnight shipping was achieved for 95%. On arrival, samples were centrifuged at 30,000 g and 4°C and either analyzed (general chemistries) or stored at –80°C. Several questionnaires addressing medical and social conditions, including a food-frequency questionnaire (FFQ) (Block 98, NutritionQuest Berkeley, CA), were left with the participants to be completed and mailed back to the study center (10). NutritionQuest (http://www.nutritionquest.com/) is a company founded by Gladys Block in 1993 to help researchers study diet at the population level. NutritionQuest provided all FFQ and analyzed nutrient content by using proprietary algorithms for the REGARDS study. Approximately 80% of participants returned the FFQ. Total and individual dietary fat intake was calculated by NutritionQuest with use of proprietary algorithms.

For these analyses, we excluded participants who did not complete ≥85% of the FFQ (n = 7297) or had caloric intakes outside of the range we deemed plausible, where the valid range for women was 500–4500 kcal and for men was 800–5000 kcal (n = 1267). Those who had total dietary fat values higher than the 99th percentile or lower than the first percentile (n = 1085) were also excluded for potentially unreliable data. Furthermore, participants who had missing data for urinary albumin-to-creatinine ratio (ACR) or eGFR were excluded (n = 1344) for a final analysis population of 19,246 participants (Figure 1).

FIGURE 1.

Exclusionary cascade for analysis subpopulation: data from the Reasons for Geographic and Racial Differences in Stroke (REGARDS) Study 2003–2007. FFQ, food-frequency questionnaire; ACR, albumin-to-creatinine ratio; EGFR, estimated glomerular filtration rate.

Measurements of dietary fat intake

Daily dietary intake of total, saturated, monounsaturated, polyunsaturated, and trans fat (g/d) were calculated from the Block 98 FFQ. The Block 98 has been described in detail elsewhere (http://www.nutritionquest.com/products/B98_DEV.pdf). Briefly, this instrument is a 110-item FFQ to estimate average dietary intake over the past year and was validated with use of 24-h recalls in the third National Health and Nutrition Examination Survey (NHANES III) population. Both dietary fat and total caloric intake came directly from NutritionQuest by using proprietary algorithms based on the validated Block 98 FFQ.

Measurements of high albuminuria and eGFR

High albuminuria

Urinary albumin was measured in batches during enrollment by using the BN ProSpec Nephelometer from Dade Behring (Marburg, Germany). The assay range is 2.4–76.9 mg/L on initial sampling. Automatic dilutions are performed on specimens with higher concentrations. The interassay CVs were 2.2% at 109.9 mg/L and 4.3% at 12.7 mg/L. Urinary creatinine was measured in batches during enrollment on the Modular-P chemistry analyzer from Roche/Hitachi (Indianapolis, IN). The assay range was 1–650 mg/dL on initial sampling. Automatic dilutions were performed on specimens with higher concentrations. The interassay CVs were 2.6% at 66.6 mg/dL and 8.6% at 15.6 mg/dL.

High albuminuria [formerly known as microalbuminuria (11)] was defined as a urinary ACR of 25–354 μg/mg in women and 17–250 μg/mg in men (12, 13). These values correlate with an albumin excretion rate of 20–200 μg/min or a 24-h urine measurement of 30–300 mg/d, which is the traditional definition of microalbuminuria. Very high albuminuria, or macroalbuminuria (11), was defined as an ACR ≥ 355 μg/mg in women and ≥ 250 μg/mg in men. Participants with very high albuminuria were excluded from the albuminuria analyses.

eGFR

Creatinine was measured in recentrifuged plasma samples at time of arrival to the REGARDS Core laboratory at University of Vermont (Ortho Vitros 950IRC; Johnson & Johnson Clinical Diagnostics, Rochester, NY; CV: 1.1%). After recruitment was complete, the REGARDS laboratory changed creatinine reagents to a method traceable to creatinine determined by isotope dilution mass spectrometry as previously described (14). The isotope dilution mass spectrometry-traceable equation was used to calibrate the REGARDS creatinine values to calculate eGFR by using the CKD-EPI equation (15):

where Scr is serum creatinine, k is 0.7 for females and 0.9 for males, a is −0.329 for females and −0.411 for males, min indicates the minimum of Scr/k or 1, and max indicates the maximum of Scr/k or 1.

Measurements of clinical and laboratory covariates

Weight and height were assessed during the in-home visit by a trained medical examiner. In addition, participants were asked to bring all prescription medications to the examiner. Medications were then reviewed by the Pharmacy Department at Samford University (Birmingham, AL). All medications were linked back to a unique generic name. Race, sex, cigarette smoking status, alcohol intake, physical activity level, and cardiovascular disease were self-reported on a baseline participant questionnaire. Presence of diabetes was defined as fasting glucose ≥126 mg/dL, nonfasting glucose ≥200 mg/dL, or use of diabetic pills or insulin. Alcohol use was quantified by asking the participant the average consumption in drinks per week. Then alcohol user group was categorized based on National Institute on Alcohol Abuse and Alcoholism definition of use, where moderate use equals 0–7 drinks/wk for women and 0–14 drinks/wk for men and heavy use equals ≥7 drinks/wk for women and ≥14 drinks/wk for men. High-sensitivity CRP (hsCRP) was analyzed in batches by particle-enhanced immunonephelometry by using the BNII nephelometer (N High Sensitivity CRP; Dade Behring) with interassay CVs of 2.1–5.7%.

Statistical analysis

Chi-square tests were used for comparing proportions, t tests for means of normally distributed continuous data, and nonparametric tests by Wilcoxon's sum rank and Kruskal-Wallis for skewed continuous variables. Univariate and multivariable logistic regression models were constructed to assess adjusted associations of quintiles of dietary fats with high albuminuria or eGFR <60 mL · min⁻¹ · 1.73 m⁻² with the lowest quintile always defined as the referent category. Multivariable analyses for the outcome of high albuminuria were adjusted for eGFR, and multivariable analyses for the outcome of eGFR <60 mL · min⁻¹ · 1.73 m⁻² were also adjusted for ACR.

A P value < 0.05 was considered statistically significant. All analyses were energy adjusted for daily caloric intake. Because we anticipated that dietary fat intake may vary in the presence of diabetes (16) o r by ethnicity or race (17), we performed stratified analyses with formal tests of interaction terms in multivariable-adjusted models.

SAS for Windows, version 9.1 (SAS Institute, Cary, NC) was used for all analyses. This study was approved by the Human Research Committee Institutional Review Board at all participating institutions.

RESULTS

These 19,246 REGARDS participants were 67% white, 33% black, and 45% male with a mean (±SD) age of 64.9 ± 9.2 y and a mean body mass index (BMI; in kg/m²) of 29.0 ± 6.0. Also, 57% had hypertension, 18% had diabetes, 22% had prevalent ischemic heart disease, and 32% were on angiotensin-converting enzyme (ACE) inhibitors or angiotensin-2 receptor blockers (ARBs). eGFR was 85 ± 19 mL · min⁻¹ · 1.73 m⁻² and median ACR was 7.1 μg/mg with 10% with eGFR <60 mL · min⁻¹ · 1.73 m⁻² (n = 1983) and 16% with high albuminuria (n = 3009) (Table 1). We assessed whether participants included in these dietary fats analyses were representative of the entire cohort and found that they were more likely to be white (67% compared with 43% in excluded) and less likely to be hypertensive (57% compared with 64%) or diabetic (18% compared with 29%) or taking ACE inhibitors or ARBs (37% compared with 32%). There were no other meaningful differences between participants who were included and who were not included in these diet analyses (data not shown).

TABLE 1.

Demographic and clinical characteristics of participants in the Reasons for Geographic and Racial Differences in Stroke (REGARDS) Study, 2003–2007¹

	All REGARDS (n = 19,246)	No high albuminuria (n = 15,783)	High albuminuria present (n = 3009)2	P value	eGFR >60 (n = 17,263)	eGFR ≤60 (n = 1983)	P value
Age (y)	64.9 ± 9.23	64.3 ± 9.1	67.6 ± 9.6	<0.0001	64.0 ± 8.9	72.2 ± 8.6	<0.0001
Age (%)
40–54 y	12	12	9	<0.0001	13	3	<0.0001
55–64 y	39	41	30		41	16
65–74 y	33	32	36		32	39
75–96 y	16	15	25	14	42
Race (%)
White	67	69	63	<0.0001	68	67	0.56
Black	33	31	37	32	33
Sex (%)
Male	45	42	56	<0.0001	45	46	0.40
Female	55	58	44	55	54
Hypertension4 (%)	57	53	72	<0.0001	54	80	<0.0001
Diabetes5 (%)	18	14	33	<0.0001	17	30	<0.0001
Ischemic heart disease6 (%)	22	20	29	<0.0001	20	36	<0.0001
Cigarette smoking (%)
Never	45	47	39	<0.0001	45	46	<0.0001
Current	14	12	16		14	10
Past	41	41	45	41	44
Alcohol intake7 (%)
None	59	58	63	<0.0001	58	69	<0.0001
Moderate	36	37	32		37	29
Heavy	5	5	5	5	2
Activity level (%)
None	32	31	36	<0.0001	31	44	<0.0001
1–3 times/wk	37	38	34		38	31
≥4 times/wk	31	31	30	31	25
Calories (kcal/d)	1681 ± 631	1679 ± 628	1699 ± 644	0.11	1694 ± 634	1571 ± 592	<0.0001
BMI (kg/m2)	29.0 ± 6.0	28.8 ± 5.9	29.7 ± 6.4	<0.0001	28.9 ± 6.0	29.3 ± 6.3	0.008
BMI categories (%)
Normal (18.5–24.9 kg/m2)	25	25	23	<0.0001	25	23	0.25
Overweight (25–29.9 kg/m2)	38	39	35		38	39
Obese (≥30 kg/m2)	36	35	41		26	37
Underweight (<18.5 kg/m2)	1	1	1	1	1
ACE inhibitor or ARB medication use (%)	32	30	43	<0.0001	30	53	<0.0001
eGFR (mL · minminus1 · 1.73 mminus2)	85 ± 19	86 ± 17	81 ± 22	<0.0001	89 ± 15	48 ± 11	<0.0001
Urinary ACR8 (μg/mg)	7.1 (0.5, 11,592.6)	6.0 (0.5, 25.0)	41.2 (17.0, 353.8)	<0.0001	6.8 (0.5, 7255.6)	12.9 (0.9, 11,592.6)	<0.0001

¹eGFR, estimated glomerular filtration rate; ACE, angiotensin-converting enzyme; ARB, angiotensin-2 receptor blocker; ACR, albumin-to-creatinine ratio.

²Participants with macroalbuminuria (n = 454) were excluded from the high albuminuria analyses, which is why the total participants for the no high albuminuria plus high albuminuria equals 18,792.

³Mean ± SD (all such values).

⁴Blood pressure ≥140/90 mm Hg or taking medication.

⁵Fasting glucose ≥126 mg/dL, nonfasting glucose ≥200 mg/dL, or taking diabetic pills or insulin.

⁶Self-reported myocardial infarction, coronary artery bypass grafting, angioplasty, or stenting or evidence of myocardial infarction via electrocardiogram.

⁷Alcohol consumption: moderate = 0.1–7 drinks/wk for women, 0.1–14 drinks/wk for men; heavy = >7 drinks/wk for women and >14 drinks/wk for men.

⁸Values are medians; ranges in parentheses. ACR was determined on the basis of median (minimum, maximum) and P values obtained from nonparametric test on the medians.

The participants with high albuminuria were older; were more often male or black and more likely to have hypertension, diabetes, or cardiovascular disease; and were more likely to have a higher BMI and to be taking ACE inhibitor/ARB medication. Moreover, those with high albuminuria were less likely to have ever smoked, to have moderate alcohol intake, or be physically active. Those with high albuminuria had a significantly lower eGFR, but no difference in caloric intake was noted. Similarly, those with eGFR <60 mL · min⁻¹ · 1.73 m⁻² were also older with comorbidities such as hypertension, diabetes, and cardiovascular disease; however, there were no significant differences in race, sex, or BMI. Participants with reduced eGFR were less likely to be current smokers, physically active, or have moderate alcohol intake but more likely to be taking ACE inhibitor/ARB medication. Individuals with eGFR <60 mL · min⁻¹ · 1.73 m⁻² had lower caloric intake and significantly higher median ACR levels (Table 1).

In age- and energy-adjusted models of dietary fats, only higher saturated fat consumption was associated with the presence of high albuminuria; this was slightly attenuated but remained statistically significant after multivariable adjustment [odds ratio (OR): 1.33; 95% CI: 1.07–1.66] (Table 2). Results were not meaningfully different when stratified by race or diabetes status. Results for associations between dietary fat and high albuminuria were also not different when analysis was restricted to those without diabetes or ischemic heart disease (n = 6553). Addition of hsCRP to multivariable-adjusted models did not change associations of saturated fat intake with high albuminuria.

TABLE 2.

Associations (odds ratios and 95% CIs) between quintiles of dietary fat intake and presence of high albuminuria in the REGARDS (Reasons for Geographic and Racial Differences in Stroke) Study, 2003–2007¹

	Fat intake
High albuminuria	Quintile 1 (referent)	Quintile 2	Quintile 3	Quintile 4	Quintile 5	P for trend
Total dietary fat
Total daily calories (%)	31 ± 72	35 ± 6	38 ± 7	40 ± 6	42 ± 6	NA
Daily intake by quintile (g/d)	34.2 ± 6.0	50 ± 4.2	64.9 ± 4.2	82.9 ± 6.3	119.8 ± 20.0	NA
Age- and energy-adjusted	1.0	1.06(0.93, 1.21)	0.99(0.86, 1.14)	1.00(0.85, 1.17)	0.99(0.79, 1.23)	0.75
Multivariable-adjusted3	1.0	1.0(0.95, 1.24)	1.03(0.88, 1.20)	1.03(0.87, 1.23)	0.96(0.75, 1.21)	0.80
Saturated fat
Total daily calories (%)	9 ± 2	10 ± 2	11 ± 2	12 ± 2	13 ± 2	NA
Daily intake by quintile (g/d)	9.6 ± 1.8	14.2 ± 1.2	18.5 ± 1.3	23.8 ± 1.9	34.9 ± 6.3	NA
Age- and energy-adjusted	1.0	1.21(1.07, 1.38)	1.15(1.00, 1.33)	1.21(1.04, 1.42)	1.42(1.16, 1.74)	0.007
Multivariable-adjusted3	1.0	1.23(1.07, 1.42)	1.14(0.98, 1.32)	1.21(1.03, 1.44)	1.33(1.07, 1.66)	0.04
Monounsaturated fat
Total daily calories (%)	11 ± 3	13 ± 3	14 ± 3	15 ± 3	17 ± 3	NA
Daily intake by quintile (g/d)	12.5 ± 2.3	18.7 ± 1.6	24.4 ± 1.8	31.6 ± 2.4	46.3 ± 8.3	NA
Age- and energy-adjusted	1.0	0.97(0.85, 1.10)	0.91(0.79, 1.04)	0.93(0.79, 1.09)	0.87(0.70, 1.06)	0.18
Multivariable-adjusted3	1.0	1.03(0.90, 1.19)	0.98(0.84, 1.13)	1.00(0.84, 1.18)	0.91(0.72, 1.13)	0.52
Polyunsaturated fat
Total daily calories (%)	7 ± 2	9 ± 2	10 ± 2	11 ± 3	12 ± 3	NA
Daily intake by quintile (g/d)	8.0 ± 1.6	12.4 ± 1.1	16.4 ± 1.2	21.5 ± 1.8	32.5 ± 6.3	NA
Age- and energy-adjusted	1.0	1.01(0.89, 1.15)	1.00(0.87, 1.14)	0.99(0.85, 1.15)	0.86(0.72, 1.04)	0.22
Multivariable-adjusted3	1.0	1.05(0.92, 1.21)	1.03(0.89, 1.19)	1.03(0.88, 1.21)	0.85(0.70 1.04)	0.29
trans Fat
Total daily calories (%)	2 ± 1	2 ± 1	3 ± 1	3 ± 1	4 ± 1	NA
Daily intake by quintile (g/d)	2.2 ± 0.5	3.5 ± 0.4	4.9 ± 0.4	6.7 ± 0.7	10.9 ± 2.5	NA
Age- and energy-adjusted	1.0	0.92(0.81, 1.04)	0.97(0.85, 1.11)	0.98(0.85, 1.12)	1.14(0.96, 1.34)	0.17
Multivariable-adjusted3	1.0	0.93(0.81, 1.07)	0.99(0.86, 1.14)	0.99(0.85, 1.15)	1.10(0.91, 1.31)	0.33

¹NA, not applicable.

²Mean ± SD (all such values).

³Adjusted for age, energy intake (calories/d), race, hypertension (blood pressure ≥ 140/90 mm Hg or taking medication), BMI, physical activity (none, 1–3 times/wk, ≥4 times/wk), diabetes (fasting glucose ≥126 mg/dL, nonfasting glucose ≥200 mg/dL, or taking diabetic pills or insulin), cardiovascular disease (self-reported myocardial infarction, cornary artery bypass grafting, angioplasty, or stenting or evidence of myocardial infarction via electrocardiogram), cigarette smoking (never, past, current), alcohol intake (none, moderate, heavy), estimated glomerular filtration rate, and angiotensin-converting enzyme inhibitor or angiotensin-2 receptor blocker medication use.

For analyses of eGFR <60 mL · min⁻¹ · 1.73 m⁻², there was a borderline significant interaction term (ORs: 1.24 in whites compared with 0.73 in blacks; P for interaction = 0.05) for associations between the highest quintile of saturated fat and eGFR <60 mL · min⁻¹ · 1.73 m⁻²; we therefore chose to present stratified analyses of dietary fats with eGFR <60 mL · min⁻¹ · 1.73 m⁻² (Table 3). Although saturated and trans fat intakes were directly associated with reduced eGFR in age- and energy-adjusted models, these relations were no longer significant after multivariable adjustment. No differences in associations were seen when the eGFR <60 mL · min⁻¹ · 1.73 m⁻² analyses were stratified by diabetes status (data not shown).

TABLE 3.

Association (odds ratios and 95% CIs) between quintiles of dietary fat intake and presence of kidney dysfunction [estimated glomerular filtration rate (eGFR) <60 mL · min^minus1 · 1.73 m^minus2] in the REGARDS (Reasons for Geographic and Racial Differences in Stroke) Study, 2003–2007

	Fat intake
	Quintile 1 (referent)	Quintile 2	Quintile 3	Quintile 4	Quintile 5	P for trend
Whites: eGFR <60 mL · minminus1 · 1.73 mminus2
Total dietary fat
Age- and energy-adjusted	1.0	1.17 (0.96, 1.42)	1.06 (0.86, 1.32)	1.24 (0.96, 1.59)	1.47 (1.05, 2.07)	0.08
Multivariable-adjusted1	1.0	1.09 (0.88, 1.33)	0.96 (0.77, 1.21)	1.08 (0.83, 1.41)	1.05 (0.73, 1.51)	0.83
Saturated fat
Age- and energy-adjusted	1.0	1.18 (0.97, 1.43)	1.33 (1.08, 1.63)	1.25 (0.99, 1.59)	1.69 (1.25, 2.30)	0.004
Multivariable-adjusted1	1.0	1.08 (0.87, 1.33)	1.18 (0.94, 1.47)	1.11 (0.87, 1.44)	1.24 (0.89, 1.74)	0.24
Monounsaturated fat
Age- and energy-adjusted	1.0	1.28 (1.06, 1.55)	1.07 (0.87, 1.32)	1.20 (0.94,1.53)	1.32 (0.96, 1.83)	0.23
Multivariable-adjusted1	1.0	1.22 (0.99, 1.49)	1.02 (0.81, 1.27)	1.09 (0.84, 1.42)	1.08 (0.77, 1.53)	0.93
Polyunsaturated fat
Age- and energy-adjusted	1.0	0.96 (0.79, 1.16)	1.04 (0.85, 1.28)	1.10 (0.88, 1.37)	1.30 (0.98, 1.71)	0.07
Multivariable-adjusted1	1.0	0.94 (0.77, 1.15)	0.95 (0.77, 1.18)	1.02 (0.81, 1.30)	1.07 (0.80, 1.43)	0.48
trans Fat
Age- and energy-adjusted	1.0	1.10 (0.91, 1.33)	1.08 (0.88, 1.31)	1.19 (0.96, 1.48)	1.43 (1.11, 1.85)	0.01
Multivariable-adjusted1	1.0	0.99 (0.81, 1.22)	0.98 (0.79, 1.21)	1.02 (0.81, 1.28)	1.14 (0.86, 1.50)	0.47
Blacks: eGFR <60 mL · minminus1 · 1.73 mminus2
Total dietary fat
Age- and energy-adjusted	1.0	0.91 (0.71, 1.16)	0.89 (0.66, 1.20)	0.89 (0.62, 1.28)	0.74 (0.44, 1.24)	0.38
Multivariable-adjusted1	1.0	0.87 (0.66, 1.14)	0.83 (0.60, 1.14)	0.80 (0.54, 1.20)	0.63 (0.36, 1.12)	0.17
Saturated fat
Age- and energy-adjusted	1.0	0.76 (0.59, 0.97)	0.79 (0.59, 1.05)	0.94 (0.67, 1.33)	0.82 (0.50, 1.33)	0.53
Multivariable-adjusted1	1.0	0.73 (0.56, 0.96)	0.69 (0.51, 0.96)	0.85 (0.58, 1.24)	0.73 (0.43, 1.25)	0.23
Monounsaturated fat
Age- and energy-adjusted	1.0	0.90 (0.70, 1.15)	0.90 (0.67, 1.19)	0.89 (0.63, 1.25)	0.92 (0.57, 1.47)	0.56
Multivariable-adjusted1	1.0	0.91 (0.70, 1.19)	0.84 (0.61, 1.15)	0.87 (0.60, 1.27)	0.92 (0.55, 1.55)	0.50
Polyunsaturated fat
Age- and energy-adjusted	1.0	1.28 (1.00, 1.64)	1.13 (0.85, 1.50)	1.32 (0.96, 1.81)	1.30 (0.86, 1.95)	0.19
Multivariable-adjusted1	1.0	1.31 (0.99, 1.72)	1.10 (0.82, 1.48)	1.21 (0.85, 1.71)	1.19 (0.77, 1.83)	0.62
trans Fat
Age- and energy-adjusted	1.0	0.97 (0.75, 1.26)	1.10 (0.84, 1.44)	1.19 (0.89, 1.59)	1.52 (1.06, 2.17)	0.03
Multivariable-adjusted1	1.0	0.96 (0.73, 1.27)	1.01 (0.75, 1.36)	1.10 (0.80,1.52)	1.42 (0.96, 2.11)	0.15

¹Adjusted for age, energy intake (calories/d), race, hypertension (blood pressure ≥ 140/90 mm Hg or taking medication), BMI, physical activity (none, 1–3 times/wk, ≥4 times/wk), diabetes (fasting glucose ≥126 mg/dL, nonfasting glucose ≥200 mg/dL, or taking diabetic pills or insulin), cardiovascular disease (self-reported myocardial infarction, coronary artery bypass grafting, bypass, angioplasty, or stenting or evidence of myocardial infarction via electrocardiogram), cigarette smoking (never, past, current), alcohol intake (none, moderate, heavy), eGFR, and angiotensin-converting enzyme inhibitor or angiotensin-2 receptor blocker medication use.

DISCUSSION

We report that higher intake of saturated fat is independently associated with the presence of high albuminuria in middle-aged and older adults and that this association did not vary by race or diabetes status. No statistically significant independent effect of saturated fat intake on presence of eGFR <60 mL · min⁻¹ · 1.73 m⁻² was observed, however. Because high albuminuria is powerfully associated with subsequent overt kidney disease (18), cardiovascular risk (19, 20), all-cause mortality (20, 21), and incident chronic diseases and death (22, 23), these data suggest that lowering dietary intake of saturated fat may be associated with improved outcomes through reducing the risk of high albuminuria.

Our results are consistent with an analysis of 3348 women (97% white) in the Nurses' Health Study, which found that higher dietary intake of animal fat was the primary nutrient associated with presence of high albuminuria (for highest compared with lowest quartile of intake—OR: 1.72; 95% CI: 1.12, 2.64) (24). In this study, animal fat was very highly correlated with saturated fat, as expected (Spearman's r = 0.84, P < 0.001). As in the Nurses' Health Study cohort, no effect modification by presence of diabetes was detected. This current investigation confirms that higher saturated fat intake is associated with high albuminuria and that the association is seen in a cohort that includes men and blacks. Furthermore, our findings are consistent with an analysis of diet patterns and albuminuria in >5000 participants of the Multi-Ethnic Study of Atherosclerosis (MESA), which reported a direct association between each additional serving per day of nondairy animal food intake with ACR (25).

We hypothesized that inflammation may be a pathologic link between saturated fat intake and presence of high albuminuria. For example, the MESA cohort reported that a diet pattern high in fats and processed meats was directly associated with markers of inflammation, including hsCRP (26). Higher hsCRP, in turn, has been directly and significantly associated with albuminuria levels in type 2 patients with diabetes (27, 28) as well as with presence of high albuminuria in a large, nationally representative US cohort (NHANES 1999–2004) after multivariable adjustment (29). Including hsCRP in the multivariable models for high albuminuria and saturated fats in REGARDS participants did not substantially change the ORs (data not shown), which does not support the hypothesis that hsCRP may lie on the causal pathway between saturated fat intake and high albuminuria. It is still possible, however, that other inflammatory biomarkers that have not been measured in this cohort may mediate the association.

A number of limitations in this investigation need to be acknowledged. The study design is cross-sectional and does not allow for inference of causality. We do not have information on change in ACR or eGFR over time in these participants. Because our study cohort was composed of whites and blacks only, our findings may not be generalizable to other ethnic groups, and the potential effect modification of race in the association of saturated fat intake with eGFR <60 mL · min⁻¹ · 1.73 m⁻² also needs to be tested in a different population. Compared with REGARDS participants who were excluded (most because of lack of adequate dietary data), the included participant sample for these analyses had a higher representation of whites as well as lower prevalence of hypertension and diabetes, which may reflect a selection bias. Black race (12), hypertension (30), and diabetes (30) are each associated with higher prevalence of high albuminuria, however, so one would expect this to bias the results toward the null, whereas a statistically significant relation was still observed. Conversely, however, if participants with these "unhealthy" characteristics would have reported lower dietary fat intake on their FFQs, this selection bias could have biased the results away from the null. Because lack of reliable data on dietary intake was the exclusion criterion for the majority of those not included in these analyses, we can only note this potential selection bias without conclusive comment on how this may have affected the results. As with any observational study, there is always the possibility of residual confounding. Notable strengths of this study, however, include the large sample size of black and white US adults, availability of detailed information for many important covariates, and the quantification of both albuminuria and eGFR in this REGARDS cohort.

In conclusion, higher dietary intake of saturated fats is associated with presence of high albuminuria but not with eGFR <60 mL · min⁻¹ · 1.73 m⁻². No significant associations between total or other types of dietary fats and measures of nephropathy were observed, however. Further investigations on how saturated fat may influence progression of albuminuria or kidney function decline are warranted.