Rising Plasma Beta-Carotene Is Associated With Diminishing C-Reactive Protein in Patients Consuming a Dark Green Leafy Vegetable–Rich, Low Inflammatory Foods Everyday (LIFE) Diet

Abstract

Chronic inflammation contributes to a number of chronic diseases and can be assessed with C-reactive protein (CRP). In this longitudinal retrospective chart review, we investigate whether patients intensively counseled to eat a specific diet high in dark green leafy vegetables, and thus high beta-carotene, have reductions in plasma high-sensitivity CRP (hsCRP). We term this the Low Inflammatory Foods Everyday (LIFE) diet. Forty-three patients in a community practice instructed to eat the LIFE diet met inclusion criteria. The CRP levels were measured at least twice over the course of up to a year. Adherence to the diet was objectively assessed by measurement of plasma beta-carotene, which is abundant in dark green leafy vegetables, and subjectively by serial interviews. The change in beta-carotene was inversely correlated with change in CRP (r = −0.68, P < .0001). Additionally, patients subjectively classified as adherent had higher beta-carotene (P < .0001) and lower CRP (P = .002) as compared with patients who were classified as nonadherent. These longitudinal findings suggest that adherence to the LIFE diet leads to increased beta-carotene and decreased CRP. Thus, this type of diet may reduce risk or severity of chronic diseases involving inflammation.

‘. . . adherence to the LIFE diet leads to increased beta-carotene and decreased CRP . . .’

Chronic inflammation exacerbates a number of diseases. Low-grade chronic inflammation contributes to all stages of the atherothrombotic process ¹ and can subsequently lead to increased rates of cardiovascular events such as heart attack, stroke, and sudden cardiac death. ² In addition, chronic inflammation contributes to diabetes, ³ insulin resistance, ⁴ chronic kidney disease (CKD), ⁵ Alzheimer disease pathogenesis, ⁶ age-related macular degeneration (AMD),^7,8 uveitis, ⁹ rheumatoid arthritis, ¹⁰ systemic lupus erythema (SLE), ¹¹ and many forms of cancer.^12,13

C-reactive protein (CRP) is a widely accepted measure of systemic inflammation.^14,15 CRP is a hepatically derived acute-phase reactant that has a long plasma half-life and can be measured easily with an inexpensive blood test that has been standardized across many commercial platforms.^15,16 Previous studies have shown that baseline CRP levels are predictive of future cardiovascular events in both healthy individuals and patients with known cardiac disease.^17,18 In addition, elevated CRP levels are associated with increased risk for AMD independent of established risk factors ⁸ and may play a role in disease progression. ¹⁹

While CRP reflects the general inflammatory load of the body, it also has pro-inflammatory properties and is a mediator of disease.^15,16 Thus, lowering CRP levels is a reasonable strategy to reduce chronic disease risk. The CANTOS trial showed that decreasing inflammation with an interleukin-1β (IL-1β) inhibitor that does not decrease cholesterol levels reduced the rate of cardiovascular disease (CVD), with the greatest reduction among patients with the largest decreases in IL-6 and CRP levels.^20,21 Another study showed that rosuvastatin significantly reduced the incidence of major cardiovascular events in patients with elevated CRP without hyperlipidemia, potentially because rosuvastatin reduces CRP levels. ²² As the predictive value of CRP is strongly linear across all values, with the lowest risk among those with the lowest CRP levels, ²³ CRP should ideally be reduced to the lowest level possible.

Along with anti-inflammatory medications, CRP levels can also be decreased with lifestyle changes such as exercise and diet.^24-26 Specifically, CRP reduction has been associated with dietary intake of beta-carotene, a retinol precursor and antioxidant that is found in certain fruits and vegetables, such as dark green leafy vegetables (DGLV). ²⁷ A cross-sectional study found a strong inverse relationship between plasma beta-carotene and CRP. ²⁸ Another study showed that plasma CRP was reduced in individuals who consumed diets high in beta-carotene-rich fruits and vegetables over a 4-week period. ²⁹ However, increased beta-carotene levels due to supplement consumption have shown either no benefit or increased risk of cardiovascular events.^30-33 These results suggest that beta-carotene alone is not the mediator of CRP reduction in those who consume a vegetable-rich diet; rather, it is an indicator of the levels of other anti-inflammatory phytonutrients.

Here, we investigate whether patients intensively counseled to eat a novel DGLV-rich, Low Inflammatory Foods Everyday (LIFE) diet have a reduction in plasma CRP. We used plasma beta-carotene as an objective measure of adherence to the DGLV-rich LIFE diet because DGLVs are high in beta-carotene and prior studies have shown that a DGLV-rich diet increases plasma beta-carotene.^34,35 We report a longitudinal retrospective study examining CRP levels among 43 patients in a community practice who were instructed to eat the LIFE diet. The goal of this study was to determine whether adherence to the LIFE diet can diminish CRP levels, which is likely to reduce the risk of CVD and other diseases involving inflammation.

Materials and Methods

Subject Population

The first criterion for eligibility was a baseline CRP concentration ≥2.0, as Ridker et al ²¹ showed in the CANTOS trial that these patients have diminished cardiovascular risk when they lower their CRP. Second, subjects could not have baseline beta-carotene levels that were above the normal range, which would have made it difficult to use beta-carotene as an objective measure of adherence. Third, subjects had to remain with one laboratory over the course of the study period, as CRP levels may vary between labs. Fourth, subjects could not have an active infection or trauma, as these processes could alter CRP levels unrelated to dietary intake of DGLV. Fifth, subjects could not eat plant-based nutritional powders, more than an allotted amount of starchy vegetables (eg, sweet potatoes), or supplements that might increase CRP levels unrelated to dietary intake of DGLV. Patients taking beta-carotene supplements or eating more than one medium sweet potato (7.5 ounces) per week or 4 cooked carrots per week (foods that are high in beta-carotene but lacking many of the phytonutrients in DGLV) were excluded. Only 1 patient in this study was a smoker and did not change his smoking habits during the course of the study. Patients classified as adherent to the life diet did not add and/or increase their doses of statins, predinose, or other immunosuppressive drugs over the course of the study. In fact, several adherent patients reduced their number and dose of immunosuppressive drugs. Data were collected from the time of first visit extending as long as 1 year afterward.

Study Design

Seventy-two charts covering a 5-year time period were reviewed for eligibility. Forty-three patients met inclusion criteria. The 29 excluded patients had switched labs, had acute infections, used plant-based nutritional powders, used supplements containing beta-carotene, and/or consumed too much (see above paragraph) beta-carotene-containing starchy vegetables such as cooked carrots or sweet potatoes. All patients were instructed to follow the LIFE diet that was designed by one of the authors (DMD), a primary care physician. The LIFE diet is a modified version of Dr Joel Fuhrman’s high nutrient density (HND) diet, ³⁶ a plant-rich diet that rates foods based on total micronutrient content per calorie.

Basic recommendations of the LIFE diet include (1) Consumption of the LIFE smoothie; a 64oz batch of smoothie consists of 16 ounces (by weight) of DGLV (spinach, baby bok choy, or baby kale), 4.5 cups of blueberries, 1 banana, 2 tablespoons of cocoa powder (unsweetened), 2 tablespoons of ground flax seed, 1 cup of soy milk (plain or vanilla) or unsweetened vanilla almond milk, and 1 cup of water. (Classification of “adherent” to the LIFE diet in this study required consumption of 32 ounces per day of the LIFE smoothie, 6 out of 7 days per week.) (2) Unlimited consumption of DGLV, which include cruciferous vegetables. Examples of DGLV are spinach, kale, collard greens, bok choy, broccoli, cauliflower, cabbage, Brussels sprouts, arugula, Swiss chard, endive, asparagus, mustard greens, beet greens, mache, broccolini, broccoli rabe, radish, watercress, escarole, romaine, or green leaf. (Adherence to the LIFE diet required consumption of at least 5 ounces by weight of DGLV in salad or cooked vegetables per day.) They were also encouraged to have other nonstarchy vegetables, including onions, mushrooms, garlic, green and yellow zucchini, eggplants, peppers, or tomatoes. (3) Unlimited consumption of fruit, and in particular berries. (This requirement is satisfied by fruit included in the LIFE smoothie, but additional fruit consumption is encouraged.) (4) Consumption of beans and legumes. (Adherence required consumption of at least ½ cup of beans or legumes or 3 ounces of bean pasta once per day.) (5) Limiting consumption of whole grains and starchy vegetables, including potatoes, sweet potatoes, winter squashes (butternut, spaghetti, acorn), peas, corn, and cooked carrots. (Adherence required consumption of no more than 2 servings of whole grains or starchy vegetables that do not contain beta-carotene per day. Inclusion in the study required consumption of no more than 1 medium-cooked or uncooked sweet potato or 4 cooked carrots per week, as these contain beta-carotene and can therefore mask beta-carotene levels derived from DGLVs.) (6) Limiting consumption of refined grains such as white flour. (Adherence required consuming no more than 1 serving or 4 ounces of refined grains once per week.) (7) Limiting consumption of added sugar. (Adherence required consuming no more than 24 g/d, which is in line with the American Heart Association guidelines and the World Health Organization’s most stringent guidelines.) (8) Consuming but limiting intake of nuts and seeds. (Adherence required consumption of 1-2 ounces of raw seeds and nuts daily, but no more than 2 ounces.) (9) Limiting consumption of animal protein, including any type of meat, fish, eggs, or dairy products. Among these, fish is preferred. (Adherence required consuming no more than 4 ounces of animal protein no more than once per day, with limitation of cold cuts, bacon, butter, or cheese to no more than 1 serving per month.) (10) Limiting consumption of oils. (Adherence required consuming no more than 1 tablespoon per day, though patients were encouraged to consume 1 teaspoon or less.) (11) Limiting consumption of dates. (Adherence to the LIFE diet required consumption of no more than ½ a serving (2 medium sized dates) per day. Additionally, patients with low vitamin B₁₂ (<550 pg/mL in line with the European standards) or high methlylmalonic acid (MMA) levels (above the normal range determined by the lab) were counseled to take 500 to 1000 µg vitamin B₁₂ per day.

Adherence to the LIFE diet was objectively assessed by measurement of plasma beta-carotene, which is abundant in DGLV, at least twice over the course of the year. While food frequency questionnaires were not available in the patients’ medical records, adherence with the diet was subjectively assessed by DMD at office visits with questioning about each of the adherence points listed above. Additionally, DMD performed a physical exam and body analysis at each visit to measure bioelectrical impedance analysis (BIA) using the Tanita body analysis scale (Tokyo, Japan). Measurements included body mass index (BMI), fat mass, and fat percent. This retrospective analysis of de-identified patient data was approved by the University of Pennsylvania’s Institutional Review Board (IRB Protocol #: 831566).

Laboratory Measurements

Plasma CRP levels were measured using a high-sensitivity CRP (hsCRP) test at any one of the following labs: Quest, Labcorp, Bioreference, Stony Brook University Hospital (Stony Brook, NY) or Mather Hospital (Port Jefferson, NY). Plasma beta-carotene levels were also measured by the same laboratories. As mentioned above, in order to meet inclusion criteria, patients had to remain with one laboratory over the course of the study period. Thirteen patients (30.2%) were tested at Quest, 25 patients (58.1%) were tested at Labcorp, 3 patients (7.0%) were tested at Bioreference, 1 patient (2.3%) was tested at Stony Brook University Hospital, and 1 patient (2.3%) was tested at Mather Hospital.

Statistical Analysis

Descriptive analyses were performed using mean (standard deviation, median, range) for continuous measures, and percentage for categorical measures. Comparison of means (eg, CRP, or beta-carotene) between 2 groups were made using the 2-sample t test. The longitudinal change in continuous measures (eg, CRP, beta-carotene) were first summarized using linear slope calculated from the linear regression models. The correlation between slope for change in beta-carotene levels and slope for change in CRP levels was evaluated using the Spearman correlation coefficient (r). The comparison of longitudinal change in CRP levels (analyzed either as slope for CRP level or slope for percentage change in CRP level from baseline) between LIFE adherence groups (nonadherent, adherent) and between beta-carotene change groups (slope <0, 0-0.3, >0.3) were made using the analysis of variance (ANOVA) with and without adjustment of covariates (age, gender, race, baseline CRP level, test lab, and weight change). All the statistical analyses were made in SAS v9.4 (SAS Institute Inc, Cary, NC), and 2-sided P < .05 was considered statistically significant.

Results

Characteristics of Study Participants

Forty-three patients met the inclusion/exclusion criteria and were included in the present study. A complete summary of the baseline characteristics of participants is shown in Table 1. The gender ratio was female (65.1%) versus male (34.9%). The mean age was 61.1 years, ranging from 28 to 88 years. The racial demographics included 32 (74.4%) white patients, 10 (23.3%) black patients, and 1 (2.3%) Hispanic patient. Twenty (46.5%) patients had diabetes, 21 (48.8%) patients had hypertension, 22 (55.8%) patients had high cholesterol, and 16 (37.2%) patients had autoimmune disease. The total number of patient visits ranged from 2 to 5 over the course of the study, with 17 (39.5%) patients having 2 visits, 18 (41.9%) patients having 3 visits, 7 (16.3%) patients having 4 visits, and 1 (2.3%) patient having 5 visits. The mean length of follow-up with complete laboratory result time-points was 176 days, ranging from 29 to 348 days. The baseline CRP ranged from 2.0 to 34.8 mg/L, with a mean of 6.7 mg/L. The baseline beta-carotene ranged from 4.0 to 194 µg/dL, with a mean of 38.9 µg/dL. The baseline weight ranged from 101 to 339 pounds, with a mean of 199 pounds. The baseline fat percent ranged from 20.9% to 54.2%, with a mean of 37.8%. Based on the criteria described above, 22 (51.2%) patients were classified as adherent to the LIFE diet while 21 (48.8%) were classified as nonadherent.

Table 1.

Characteristics of Study Participants (n = 43).

Characteristic
Age (years)
Mean (SD)	61.1 (12.9)
Median (Q1, Q3)	63 (55, 70)
Min, max	28, 88
Gender, n (%)
Male	15 (34.9)
Female	28 (65.1)
Race, n (%)
White	32 (74.4)
Black	10 (23.3)
Hispanic	1 (2.3)
Chronic disease, n (%)
Diabetes	20 (46.5)
Hypertension	21 (48.8)
Hyperlipidemia	22 (58.8)
Autoimmune	16 (37.2)
Adherence to LIFE diet, n (%)
Adherent	22 (51.2)
Nonadherent	21 (48.8)
Baseline C-reactive protein (mg/L)
Mean (SD)	6.7 (7.5)
Median (Q1, Q3)	4.0 (2.8, 6.3)
Min, max	2.0, 34.8
Baseline beta-carotene (µg/dL)
Mean (SD)	38 (34.2)
Median (Q1, Q3)	26.0 (21.0, 56.0)
Min, max	4.0, 194
Baseline weight (pounds)
Mean (SD)	199 (43.5)
Median (Q1, Q3)	195 (168, 218)
Min, max	101, 339
Baseline fat percent
Mean (SD)	37.8 (8.61)
Median (Q1, Q3)	38.5 (30.0, 44.5)
Min, Max	20.9, 54.2
Number of visits, n (%)
2	17 (39.5)
3	18 (41.9)
4	7 (16.3)
5	1 (2.3)
Length of follow-up (days)
Mean (SD)	176 (90)
Median (Q1, Q3)	160 (102, 244)
Min, Max	29, 348

Association Between Changes in Plasma Beta-Carotene Levels and CRP Levels

Over a median of 160 days (range: 29-348 days) of follow-up, the mean (SD) change (calculated from linear slope) was 0.23 (0.35) µg/L/d for beta-carotene levels, and −0.005 (0.059) mg/L/d for CRP levels. The change in beta-carotene levels was inversely correlated with change in CRP levels (Table 2, r = −0.68, P < .0001). The percent change in beta-carotene levels from baseline was also inversely correlated with change in CRP levels (Table 2, r = −0.78, P < .0001). Additionally, change in beta-carotene levels was correlated with reductions in weight (r = −0.45, P = .002) and fat percent (r = −0.38, P = .01; Table 2). Mean CRP level from the initial visit to final visit dropped from 7.01 to 1.75 mg/L among patients (n = 14) whose beta-carotene levels at last visit rose out of the normal range defined by the laboratory, while mean CRP levels increased (6.51 mg/L at baseline and 7.74 mg/L at last visit) in patients whose beta-carotene at last visit did not rise out of the normal range (n = 29; Figure 1). The mean CRP change between baseline and final visit was significantly different between patients with last visit beta-carotene levels above the normal range (−5.26 mg/L) and patients with beta-carotene levels at last visit within the normal range (1.23 mg/L; 2-sample t test P = .02). Longitudinal CRP levels decreased among patients with last visit beta-carotene values above the normal range (slope −0.04 mg/L/d) and increased among patients with beta-carotene levels at last visit within the normal range (slope 0.01 mg/L/d), and these were significantly different (2-sample t test P = .01).

Table 2.

Correlation Between Longitudinal Change (Linear Slope) of Beta-Carotene and Longitudinal Change (Linear Slope) of C-Reactive Protein (CRP) and Other Measures (n = 43).

Measures for Correlation With Change of Beta-Carotene	Spearman Correlation, r (P)
Slope of CRP	−0.68 (<.0001)
Slope of CRP percent change from baseline	−0.78 (<.0001)
Slope of weight	−0.45 (.002)
Slope of fat percent	−0.38 (.01)

Figure 1.

Boxplots comparing the mean last visit C-reactive protein (CRP) levels in patients with beta-carotene levels at last visit above the normal range (1.75 ± 0.57 mg/L, n = 14) as compared to patients with beta-carotene levels at last visit within the normal range (7.74 ± 3.87 mg/L, n = 29). Statistical analysis was performed using a 2-sample t test.

Changes in CRP levels were larger among patients with larger increases in beta-carotene levels, with CRP slope −0.03 mg/L/d for patients with a beta-carotene slope greater than 0.3 µg/L/d (n = 13), CRP slope −0.01 mg/L/d for patients with beta-carotene slope from 0 to 0.3 µg/L/day (n = 14), and CRP slope of 0.02 mg/L/d for patients with a beta-carotene slope less than 0.3 µg/L/d (n = 16). The mean CRP slopes among these 3 subgroups were significantly different (linear trend P = .02) and remained significant after adjusting for baseline CRP levels, test lab, age, gender, race, and weight change (linear trend P = .03, Table 3). The analysis for percent change in CRP levels from baseline provided similar results (Table 4).

Table 3.

Comparison of Slope in CRP Between Patients Grouped by Beta-Carotene Slope and by Adherence to the LIFE Diet.

		Univariate Analysis		Multivariate Analysis a
	n	Mean Slope of CRP (SE)	P	Mean Slope of CRP (SE)	P
Grouped by beta-carotene slope			.02 b		.03 b
<0	16	0.02 (0.01)		0.02 (0.02)
>0-0.3	14	−0.01 (0.02)		−0.01 (0.02)
>0.3	13	−0.03 (0.02)		−0.04 (0.02)
Grouped by adherence			.003		.001
Nonadherent	21	0.02 (0.01)		0.02 (0.01)
Adherent	22	−0.03 (0.01)		−0.04 (0.01)

Abbreviations: CRP, C-reactive protein; LIFE, Low Inflammatory Foods Everyday; SE, standard error.

^aAdjusted by baseline CRP level, test lab, age, gender, race, and weight change.

^bFrom test of linear trend.

Table 4.

Comparison of Percentage Change in CRP From Baseline Between Patients Grouped by Beta-Carotene Slope and by Adherence to the LIFE Diet.

		Univariate Analysis		Multivariate Analysis a
	n	Mean Slope of Percent Change in CRP From Baseline (SE)	P	Mean Slope of Percent Change in CRP From Baseline (SE)	P
Grouped by beta-carotene slope			.005 b		.01 b
<0	16	0.35 (0.20)		0.28 (0.27)
>0-0.3	14	−0.09 (0.21)		−0.18 (0.28)
>0.3	13	−0.52 (0.22)		−0.70 (0.30)
Grouped by adherence			.0006		.001
Nonadherent	21	0.37 (0.16)		0.28 (0.20)
Adherent	22	−0.47 (0.16)		−0.66 (0.20)

Abbreviations: CRP, C-reactive protein; LIFE, Low Inflammatory Foods Everyday; SE, standard error.

^aAdjusted by baseline CRP level, test lab, age, gender, race and weight change.

^bFrom test of linear trend.

Changes in Plasma Beta-Carotene Levels and CRP Levels Among Adherent and Nonadherent Patients

Patients were classified as “adherent” and “nonadherent” to the LIFE diet based on criteria described above. Twenty-two (51.2%) patients were categorized as adherent to the LIFE diet while 21 (48.8%) were categorized as nonadherent. Mean beta-carotene levels from the initial visit to final visit rose from 45.8 to 100.5 µg/dL among adherent patients. In nonadherent patients, the beta-carotene levels remained stable (31.8 µg/dL at baseline and 27.1 µg/dL at final visit; Figure 2). The mean change in beta-carotene levels between baseline and final visit was significantly different between adherent patients (54.8 µg/dL) and nonadherent patients (−4.71 µg/dL; 2-sample t test P < .0001). Similarly, longitudinal beta-carotene levels increased among adherent patients (slope = 0.46) and decreased among nonadherent patients (slope = −0.01), and the difference was significant (2-sample t test P < .0001).

Figure 2.

Boxplots comparing the mean last visit beta-carotene levels in patients classified as adherent (100.50 ± 59.70 µg/dL, n = 22) as compared with patients classified as nonadherent (27.10 ± 16.90 µg/dL, n = 21). Statistical analysis was performed using a 2-sample t test.

Mean CRP levels from the initial visit to final visit dropped from 6.67 to 2.26 mg/L among adherent patients, while mean CRP levels increased from 6.65 to 9.49 mg/L in nonadherent patients (Figure 3). The mean CRP change between baseline and final visit was significantly different between adherent patients (−4.43 mg/L) as compared with nonadherent patients (2.83 mg/L; 2-sample t test P = .002). Longitudinal CRP levels increased among patients who were not adherent to the LIFE diet (slope = 0.02) and decreased among patients who were adherent (slope = −0.03). The difference was statistically significant (ANOVA P = .003, Table 3) and remained significant after adjusting for baseline CRP levels, test lab, age, gender, race, and weight change (ANOVA P = .001, table 3). The analysis for percent change in CRP levels from baseline provided similar results (Table 4).

Figure 3.

Boxplots comparing the mean C-reactive protein (CRP) levels at last visit in patients classified as adherent (2.26 ± 2.15 mg/L, n = 22) as compared with patients classified as nonadherent (9.49 ± 11.50 mg/L, n = 21). Statistical analysis was performed using a 2-sample t test.

Discussion

This retrospective longitudinal study shows that subjective adherence to the LIFE diet leads to decreased CRP and increased plasma beta-carotene. Objectively, beta-carotene was inversely associated with CRP. Both the magnitude of CRP reduction and the percent reduction of CRP from baseline were larger among patients with greater increases in beta-carotene levels.

Plasma beta-carotene levels were inversely correlated with longitudinal change in weight and fat percent. This implies that the LIFE diet may have additional health benefits other than lowering CRP levels. It is plausible that the decreased CRP levels could be attributed to the decrease in weight and fat percent as opposed to other anti-inflammatory effects of a phytonutrient rich diet. However, this is unlikely, as the inverse association between beta-carotene and CRP levels remained significant after the data were corrected for a number of possible confounding factors, including weight change.

Significantly, we excluded patients who consumed more than the diet’s recommended quantity of cooked carrots, sweet potatoes, or other starchy vegetables containing beta-carotene. While these foods raise plasma beta-carotene levels, they are not rich in other phytonutrients and do not have the same beneficial effects on inflammation. Patients consuming these foods could have confounded our findings, as they may not have consumed a sufficient quantity of phytonutrient-rich DGLVs, and thus may not have experienced decreases in CRP levels. We also excluded patients who had consumed plant-based nutritional powders for the same reason.

It is important to note that reduction of CRP levels in patients adherent to the LIFE diet could simply reflect a possible increase in the number and/or dose of medications that reduce CRP levels. However, this is unlikely, as patients classified as adherent did not add and/or increase their doses of statins, prednisone, or other immunosuppressive drugs over the course of the study. In fact, several adherent patients reduced their number and dose of immunosuppressive drugs and still experienced decreases in CRP levels. Similarly, CRP levels could have been confounded by subject smoking habits (ie, initiating, changing in quantity, or cessation), as individuals who smoke have been shown to have elevated CRP levels ³⁷ that decrease with cessation over time. ³⁸ However, only 1 patient in the present study was a smoker. Because of his diet, he was classified as nonadherent. He had smoked for over 30 years and had no change in his smoking regimen during the course of the study. Thus, his increase in plasma CRP level by the end of the study should not be attributed to his smoking habits as opposed to nonadherence to the LIFE diet.

Our findings are consistent with prior studies. Several cross-sectional studies have shown an inverse association between beta-carotene and CRP concentrations.^28,39 In addition, high intake of fruits and vegetables is associated with lower plasma CRP levels. ⁴⁰ Most recently, a prospective randomized controlled trial enrolling healthy, nonsmoking men showed that a 4-week diet consisting of 8 servings/day of fruits and vegetables led to a significant reduction in plasma CRP concentration. ²⁹

The present study provides some important additional information relative to this recent prospective study, which is the only other longitudinal study to our knowledge. First, Watzl and colleagues ²⁹ demonstrate a strong longitudinal inverse relationship between beta-carotene and CRP over a 4-week period. Here, we examine data collected over the course of up to a year. Second, Watzl et al ²⁹ enrolled subjects who were healthy males. Our study population included both men and women (with a higher percentage of women), many of whom had common chronic diseases such as diabetes and hypertension. Third, Watzl et al ²⁹ instructed subjects to consume a relatively flexible diet, in which subjects only had to consume a set number of total servings of fruits and vegetables. They did not specify how many servings should come from each group or limit consumption of other potentially unhealthy foods. Additionally, vegetable servings included carrots and starchy vegetables that contain beta-carotene, which may not have the same effect on CRP or general health as beta-carotene-rich DGLV. In contrast, subjects in the present study were on a highly specific LIFE diet that emphasizes consumption of food with high total micronutrient content per calorie (such as DGLV) and limits consumption of potentially pro-inflammatory foods (such as starchy vegetables, processed sugar, white flour, animal products, and oils).

Interestingly, while we showed the same inverse relationship as Watzl et al, ²⁹ we found a larger magnitude negative slope correlating longitudinal changes in beta-carotene and CRP levels. We corrected for the time denominator of our slope in order to compare the 2 slopes, as our slope was calculated over a 10-day time period while their slope was calculated over a 4-week time period (r = −2.04 vs r = −0.51). This difference in slope magnitude could be attributed to the more specific LIFE diet, and may suggest that it is more effective in lowering CRP levels and ultimately disease risk factors than a more flexible diet. However, this could also be attributed to a possible difference in the mean baseline levels of CRP between studies. Our subjects had a mean baseline CRP level of 6.67mg/L, which is considered substantially above the normal range (≥3 mg/L). We cannot compare the impact of the mean baseline CRP levels on our findings because this value was not reported by Watzl et al. ²⁹ However, it is plausible that subjects with normal baseline CRP levels may not experience the same magnitude of decrease regardless of dietary intervention. To clarify whether specific diets may differently affect the magnitude of CRP level reduction, a prospective RCT could be designed to compare the magnitude of change of CRP levels among patients who have similar baseline characteristics (including CRP levels) randomized to different diets.

It is important to note that there are key differences between patients who take beta-carotene as supplements and those with elevated plasma beta-carotene levels due to dietary intake (as in this study) or other causes. High dose beta-carotene supplementation has been linked with adverse outcomes. Two randomized, masked, primary prevention trials found that high-dose beta-carotene supplementation increased the risk of lung cancer and total mortality in smokers and asbestos workers,^31,41 and a meta-analysis showed increased risk of stomach cancer risk in the same population. ⁴¹ Additionally, beta-carotene supplementation may reduce the efficacy of cancer therapies in smokers, leading to increased recurrence and mortality. ⁴² However, these negative outcomes have not been shown in patients with elevated plasma beta-carotene levels due to dietary intake in whole foods. ⁴³ In fact, high dietary intake of beta-carotene has been associated with reduced risk of cervical cancer ⁴⁴ and lung cancer.^43,45 Thus, patients with elevated beta-carotene plasma levels from adherence to our LIFE diet should not be at risk for these poor outcomes. Furthermore, patients on the LIFE diet are consuming foods that contain numerous phytonutrients other than beta-carotene, such as other carotenoids, bioflavonoids, polyphenols, and glucosinolates. It is likely that these phytonutrients together are responsible for disease protection, which should not be attributed to beta-carotene alone. Similarly, it is likely that beta-carotene itself does not directly decrease CRP, but instead serves as a biomarker for diets rich in DGLV, which contain high levels of phytonutrients that together may reduce CRP levels. Furthermore, in addition to high intake of DGLV, it is likely that other aspects of the LIFE diet, including high intake of beans and lentils, and low intake of refined products and animal products, contributes to the CRP-lowering effect of the diet.^46-49

Unlike its metabolic product, retinol, excessive ingestion of beta-carotene generally does not lead to toxicity.^50,51 This is because metabolism of beta-carotene from plant sources to retinol is slow and highly regulated. ⁵¹ Thus, there is little concern for toxicity in patients with elevated plasma levels of beta-carotene due to adherence to the LIFE diet. However, prolonged intake of high doses of beta-carotene (at least 20 mg/d) via supplementation or consumption can be associated with carotenodermia, a benign yellow discoloration of the skin that is reversible with decreased consumption.^51,52 Elevated plasma beta-carotene does not necessarily result in carotenodermia, but makes it more likely. One study showed that carotenodermia was only observed after plasma total carotenoid levels exceeded 400 µg/dL. ⁵² This is much larger than the mean (100.5 µg/dL) and maximum (267 µg/dL) plasma beta-carotene levels found in patients adherent to the LIFE diet at the conclusion of the study. Furthermore, while a few adherent patients had mild discoloration in their palms to a yellowish tint, adherent patients in our study did not have carotenodermia.

High-dose beta-carotene supplementation has been shown to potentiate the hepatotoxicity of ethanol. ⁵³ In baboons, this toxic interaction occurred with a beta-carotene supplement dose of 30 mg/d for 33 days, with an administered alcohol dose that was equivalent to that of the average alcoholic (50% of dietary energy).^50,54 Hepatotoxicity was found with beta-carotene plasma levels that were generally higher than those seen in patients on the LIFE diet. Specifically, this beta-carotene supplementation raised plasma beta-carotene levels in many subjects close to or above 268.44 µg/dL. ⁵⁴ Still, caution should be exercised in recommending the LIFE diet to patients who drink heavily.

There are several important strengths of this study. This was a longitudinal study and thus allowed us to examine the relationship between the change in beta-carotene and the change in CRP levels over time. The study population was varied and diverse, with a mix of gender, race, baseline chronic disease status, and adherence to the LIFE diet. Subjects received intensive dietary counseling and frequent follow-up by one physician, along with regular blood tests. While there was no formal food-frequency questionnaire in this study, intensive interviews about adherence to the LIFE diet, as well as an objective measurement, plasma beta-carotene, were utilized. A weakness of the study is that it is small and retrospective, but the positive results justify a larger prospective study.

Overall, this study suggests that the highly specific LIFE diet, which emphasizes consumption of beta-carotene-rich DGLV, is associated with increased plasma beta-carotene levels and decreased systemic hsCRP levels over time. Furthermore, this specific diet may lower CRP levels more effectively than flexible diets based solely on the quantity of fruit and vegetable consumption, though a randomized prospective study is needed. Because lower CRP levels have been associated with reduced risk of diseases promoted by chronic inflammation, such as CVD and AMD, these findings suggest that the LIFE diet would reduce disease risk and severity, improving quality of life.