Hyperglycemia and carotenoid intake are associated with serum carotenoids in youth with type 1 diabetes

Abstract

Background

Serum carotenoids are commonly used as biomarkers of fruit and vegetable intake in the general population. Although hyperglycemia induces oxidative stress, it is unknown whether this pathway is associated with lower serum carotenoids in individuals with type 1 diabetes. Consequently, the utility of serum carotenoids as markers of fruit and vegetable (FV) intake in individuals with type 1 diabetes is unclear.

Objective

The study objectives were: 1) to investigate the relationship of glycemic control, oxidative stress, dietary carotenoid and FV intake with serum carotenoid concentrations in youth with type 1 diabetes, and 2) to investigate whether glycemic control or oxidative stress moderates the association of carotenoid and FV intake with serum carotenoids.

Design

The study was a secondary analysis of baseline data from youth with type 1 diabetes. Blood samples were drawn from youth with type 1 diabetes to assess carotenoids and markers of glycemic control (HbA1c and 1,5-anhydroglucitol (1,5-AG)), urine samples were used to assess oxidative stress (8-iso-prostaglandin F2α (8-iso-PGF_2α)), and 3-day diet records completed by families were used to determine FV and carotenoid intake.

Participants/ Setting

The study participants were youth with type 1 diabetes (n=136, age: 8–16.9 years; diabetes duration ≥ 1 year; glycated hemoglobin (HbA1c): 5.8% −11.9%) enrolled in a nutrition intervention trial from 2010 to 2013 at a tertiary diabetes center in Boston, MA.

Main outcome measures

Serum carotenoids (total carotenoids, and α-carotene, β-carotene, lycopene, β-cryptoxanthin, and lutein plus zeaxanthin).

Statistical analysis

Regression analyses estimated the association of glycemic control, oxidative stress, FV and carotenoid intake with serum carotenoids, as well as the role of glycemic control and oxidative stress in moderating diet-serum carotenoid associations.

Results

Greater fruit and vegetable (β=0.35, p <0.001) and carotenoid intake (β=0.28, p<0.01) were associated with higher total serum carotenoids, and no moderation by glycemic control or oxidative stress was observed. Greater hyperglycemia, as indicated by lower 1,5-AG (β=0.27, p<0.01), was related to lower serum carotenoids; however, HbA1c was not associated with serum carotenoids. 8-isoprostaglandin F2α was not associated with glycemic control or serum carotenoids.

Conclusions

Findings support the validity of serum carotenoids as markers of FV and carotenoid intake in youth with type 1 diabetes.

Introduction

Carotenoids are fat-soluble pigments that occur naturally in fruits and vegetables¹. The antioxidant potential of these carotenoids² is thought to play a role in the association of higher serum/plasma carotenoids with reduced risk of cancer^3,4 and cardiovascular disease^5–8, a primary complication in individuals with type 1 diabetes⁹.

Despite inter-individual variations in absorption and metabolism of carotenoids¹⁰, serum concentrations are responsive to dietary intake and are widely used as biomarker of FV intake in the general population¹¹. However, oxidative stress (for example, as induced by smoking) can deplete serum carotenoids¹², and thus attenuate the association of serum carotenoids with intake^13,14. Since hyperglycemia in individuals with type 1 diabetes leads to increased oxidative stress^15–18, this may similarly reduce serum carotenoid concentrations and weaken the association of serum carotenoids with intake. Consistently, studies have indicated greater oxidative stress^19–22, and lower serum carotenoids^20,23,24 in this population as compared to healthy controls. However, the association of glycemic control with serum carotenoids is inconsistent^25–29, with only one study in type 1 diabetes patients²⁹. Further, no study has examined the relationship of oxidative stress with serum carotenoids in type 1 diabetes patients, or if serum carotenoids can be used as markers of intake across a range of glycemic control and oxidative stress values. Additional studies are needed to determine how serum carotenoids are associated with glycemic control and oxidative stress in persons with type 1 diabetes, and whether hyperglycemia and oxidative stress impact the utility of serum carotenoids as markers of intake in this population.

The objectives of this study were to investigate associations of glycemic control, oxidative stress, dietary carotenoid and FV intake with serum carotenoid concentrations in youth with type 1 diabetes, and to examine whether glycemic control or oxidative stress moderates the relationship of carotenoid and FV intake with serum carotenoids.

Methods

Design and Setting

This is a secondary analysis of data from the baseline assessment of a randomized clinical trial of a behavioral nutrition intervention conducted from 2010–2013 at a tertiary diabetes center in the northeast United States³⁰. All youth provided assent, and written informed consent was obtained from parents at the time of enrollment. Further, youth turning 18 years of age during the study provided written informed consent. Study procedures were approved by the [blinded for review] Institutional Review Boards.

Participants

Youth were screened for eligibility based on the following criteria: 8.0–16.9 years of age, diagnosed with type 1 diabetes for at least 1 year, insulin dose of at least 0.5 units per kilogram per day, HbA1c ≥6.5% (48 mmol/mol) and ≤10.0% (86 mmol/mol) at the most recent clinic visit, use of an insulin regimen of at least 3 injections per day or insulin pump, at least one clinic visit in the past year, and able to communicate in English. Exclusion criteria included daily use of premixed insulin, transition to insulin pump therapy over the past three months, real-time continuous glucose monitoring use over the past three months, participation in another intervention study over the past six months, or presence of gastrointestinal disease such as celiac disease, multiple food allergies, use of medications that interfere significantly with glucose metabolism, or significant mental illness. Of 622 eligible invited youth, 148 (24%) consented to participate and 139 (22%) completed baseline assessments. Data from one sibling each in 3 sibling pairs were excluded, resulting in 136 youth. Only baseline assessments were included for this analysis to avoid the potential impact of intervention participation on reporting accuracy of dietary intake³¹.

Measures

Non-fasting blood samples were collected at baseline and kept at room temperature for 20–30 min. The samples were centrifuged for 15 mins at ~3000 RPM at 4°C, aliquoted and frozen at −80°C for later assay. Similarly, aliquoted urine samples were stored at −80°C for later assay.

Serum carotenoids

Serum concentration (μg/mL) of α-carotene, β-carotene, lycopene and β-cryptoxanthin, as well as lutein plus zeaxanthin (μmol/L) were separated and quantified by a high-performance liquid chromatography method that has been previously described³². Concentration of lutein/zeaxanthin was converted to μg/mL and summed with other carotenoids (μg/mL) to represent total serum carotenoids.

Glycemic control

A laboratory assay standardized to the Diabetes Control and Complications Trial³³ [reference range: 4–6% (20–42 mmol/mol)] was used to ascertain HbA1c, which reflects overall glycemic control for the past 3 months. Initial HbA1c assays were performed using a Tosoh analyzer (Tosoh Medics) followed by a Roche Cobas Integra. Analysis of results from the two methods on similar samples showed no clinically significant bias. 1,5-anhydroglucitol (1,5-AG), an indicator of recent (within the past 1–2 weeks³⁴) hyperglycemic excursions, was assessed by an enzymatic (glucokinase) assay (GlycoMark). Higher HbA1c and lower 1,5-AG indicate more frequent hyperglycemia.

Oxidative stress

8-iso-prostaglandin F2α (8-iso-PGF_2α), an indicator of oxidative stress, was determined from urine frozen at −80°C using enzyme linked immunosorbent assay (ELISA).

Dietary assessment methods

Youth FV and carotenoid intake was assessed from food records completed for three consecutive days by families immediately following the clinic visit. Research assistants provided instructions for reporting food and beverage intake, and a sample diet record was given to families. Measuring cups and spoons were provided to aid portion size estimation. Families were requested to include a detailed description of the food items, such as brand or restaurant names, and item labeling (for example, nonfat, low fat yogurt). The completed records were reviewed by research staff, and families were contacted to solicit missing information. If a family did not complete diet records (0.74% of assessments), two non-consecutive 24-hour dietary recalls were conducted using the Nutrition Data System for Research (NDSR) multiple pass method³⁵. The NDSR 2012³⁵ (Nutrition Coordinating Center, University of Minnesota) was used to obtain estimates of fruit (servings per day) and vegetable (servings per day) for each record, which was summed to represent total FV intake. Intake per day of α-carotene, β-carotene, lutein/zeaxanthin, lycopene, β-cryptoxanthin estimated by NDSR was also summed to denote the total carotenoid intake. Total carotenoid intake, thus obtained, did not include intake from supplements.

Covariates

Age, sex, height and weight were abstracted from medical records obtained on the same day as the baseline assessment. Participants were requested to indicate whether they were currently using multivitamins. Body mass index (BMI) (kg/m²) was calculated from measured height and weight, and transformed into age- and sex-adjusted z-scores³⁶. Additionally, non-fasting serum LDL-C and HDL-C concentrations were directly determined using ELISA, standardized according the Center for Disease Control reference method.

Statistical analysis

All variables used in the analyses were standardized to obtain the standardized regression coefficient, and its corresponding standard error. Standardization involved rescaling of variables to obtain a mean of zero and a standard deviation of 1. To improve normality of the residuals, natural log transformation of serum carotenoids was used for all analyses.

Multiple linear regression analyses were used to estimate associations of glycemic control (HbA1c and 1,5-AG) and oxidative stress (urinary 8-iso-PGF_2α) with serum carotenoids (total carotenoids, α-carotene, β-carotene, lycopene, lutein/zeaxanthin and β-cryptoxanthin). Since dietary intake, demographic and physiological factors such as BMI^37,38 and serum cholesterol concentrations³⁸ are associated with serum carotenoids, the regression analyses adjusted for age, sex, HDL-C, LDL-C, BMI z-score, multivitamin use and corresponding carotenoids intake. The analyses were adjusted for covariates to account for potential confounding. Similar models were used to estimate relationships of FV and carotenoid intake with serum carotenoid concentration, adjusting for age, sex, HDL-C, LDL-C, BMI z-score, multivitamin use and HbA1c. Additionally, separate models were adjusted for 1,5-AG instead of HbA1c.

To explore interaction effects, markers of glycemic control and FV intake were first centered around the mean. Multiplicative interaction terms of the centered variables were used to determine whether glycemic control moderated associations of FV and total carotenoid intake with total serum carotenoids, adjusting for age, sex, HDL-C, LDL-C, BMI z-score and multivitamin use. Similarly, 8-iso-PGF_2α was centered around the mean and multiplicative interaction terms were used to test whether oxidative stress moderated associations of FV and carotenoid intake with total serum carotenoids. A p-value less than 0.05 was used for statistical significance. SPSS version 21 were used for all analyses³⁹.

Results

Figure 1 indicates the participant flow through randomization. Baseline sample characteristics are shown in Table 1. Youth were 90% non-Hispanic white, with a mean age and diabetes duration of 12.7 and 6 years, respectively. Glycated hemoglobin ranged from 5.8 to 11.9%.

Figure 1.

Participant flow chart of enrollment and baseline data collection for a randomized controlled trial of a behavioral nutrition intervention for youth with type 1 diabetes^a

^aAdapted with permission from: Nansel TR, Laffel L, Haynie D, Mehta S, Lipsky L, Volkening L, Butler D, Higgins L, Liu A. Improving dietary quality in youth with type 1 diabetes: randomized clinical trial of a family-based behavioral intervention. International Journal of Behavioral Nutrition and Physical Activity 2015; 12:58³⁰

Table 1.

Baseline characteristics, dietary intake, and serum carotenoid levels of a sample of youth with type 1 diabetes (n=136) participating in a behavioral intervention trial

Demographic characteristics	Descriptive statistics
Age, years, Mean(SD)	12.7(2.6)
Body mass index, kg/m2, Mean(SD)	21.3(4.2)
HbAlca, %, Mean(SD)	8.1(1.0)
Duration of diabetes, years, Mean(SD)	6.0(3.1)
Youth race/ ethnicity, n (%)
Non-Hispanic white	123 (90.4)
Non-Hispanic black	5 (3.7)
Hispanic	7 (5.2)
American Indian/Alaska Native	1 (0.7)
Dietary intake, Mean(SD)
Fruit and vegetable, servings/day	1.69(1.02)
Total carotenoid, μg/day	7247(5891)
α-carotene, μg/day	377(673)
β-carotene, μg/day	1688(2108)
Lycopene, μg/day	4141(4512)
Lutein/zeaxanthin, μg/day	910(833)
β-cryptoxanthin, μg/day	131(272)
Serum carotenoidb, Meanc(SD)
Total carotenoid, μg/ml	1.44(0.68)
α-carotene, μg/ml	0.10(0.10)
β-carotene, μg/ml	0.34(0.29)
Lycopene, μg/ml	0.64(0.31)
Lutein/zeaxanthin, μg/ml	0.15(0.06)
β-cryptoxanthin, μg/ml	0.21(0.26)

^aHbA1c, Glycated hemoglobin

^bTo convert from conventional (μg/ml) to SI units (μmol/ml), multiply carotenoid concentrations by the following conversion factors: α-carotene, β-carotene and lycopene = 0.001863; β-cryptoxanthin = 0.001809; Lutein/zeaxanthin = 0.001758

^cGeometric means for total carotenoid, α-carotene, β-carotene, lycopene, lutein/zeaxanthin and β-cryptoxanthin are 1.30, 0.06, 0.27, 0.55, 0.14 and 0.16 μg/ml, respectively

Table 2 shows estimates of the associations of glycemic control and oxidative stress with serum carotenoid concentrations. In model 1, one standard deviation increase in HbA1c was associated with 0.19 standard deviation decrease in β-carotene; however, it was not significantly related to β-carotene or other serum carotenoids in adjusted models. In the adjusted model, one standard deviation increase in 1,5-AG was associated with 0.22, 0.18, 0.23 and 0.27 standard deviation increases in total serum carotenoid, α-carotene, β-carotene, and β-cryptoxanthin concentrations, respectively; however, there was no significant relationship with lycopene and lutein/zeaxanthin. Higher urinary 8-iso-PGF_2α was associated only with reduced serum α-carotene in the unadjusted model, but this association was not significant in the adjusted model. Additionally, neither HbA1c (β = −0.08, standard error (SE) = 0.09, p = 0.40) nor 1,5-AG (β= 0.001, SE = 0.10, p=0.99) was significantly associated with 8-iso-PGF_2α (data not shown in table).

Table 2.

Association of glycemic control with serum carotenoid concentrations at baseline in youth with type 1 diabetes (n=136) enrolled in a behavioral intervention

	Serum concentrations, μg/ml
	Total carotenoids	β- carotene	α-carotene	Lycopene	Lutein/zeaxanthin	β-cryptoxanthin
	β(SE)
HbA1c(%)
Model 1a	−0.05(0.09)	−0.19(0.09)*	−0.17(0.09)	0.12(0.09)	−0.004(0.09)	−0.07(0.09)
Model 2b	−0.01(0.08)	−0.09(0.08)	−0.08(0.09)	0.10(0.09)	0.04(0.08)	−0.004(0.09)
1,5-AG(μg/ml)
Model la	0.27(0.08)**	0.30(0.08)**	0.26(0.08)**	0.04(0.09)	0.18(0.09)*	0.30(0.08)***
Model 2b	0.22(0.09)*	0.23(0.09)**	0.18(0.09)*	0.06(0.10)	0.14(0.08)	0.27(0.09)**
Urinary 8-iso-PGF2α (ng/ml)
Model la	−0.09(0.09)	−0.13(0.09)	−0.20(0.09)*	0.08(0.09)	−0.09(0.09)	−0.09(0.09)
Model 2b	−0.02(0.08)	−0.07(0.08)	−0.14(0.08)	0.11(0.09)	−0.07(0.08)	−0.04(0.09)

HbA1c, glycated hemoglobin; 1,5-AG, 1,5-anhydroglucitol; 8-iso-PGF2α, 8-iso-prostaglandin F2α

^aModel 1 is the unadjusted model

^bModel 2 adjusted for age, sex, low density lipoprotein cholesterol, high density lipoprotein cholesterol, bmi z-score, multivitamin use and intake of the respective carotenoid

^***p<0.001;

^**p<0.01;

^*p<0.05

Greater fruit and vegetable intake was associated with higher total serum carotenoids, and with serum α-carotene, β-carotene, lutein/zeaxanthin and β-cryptoxanthin (Table 3). Except for lycopene, the relationship of serum carotenoids with their intake was statistically significant (Table 4). The magnitude of association between intake and serum carotenoid ranged from 0.13 for lycopene to 0.28 for total carotenoids. Total carotenoid intake was also significantly associated with serum β-carotene and α-carotene concentrations.

Table 3.

Association of fruit and vegetable with serum carotenoid concentrations at baseline in youth with type 1 diabetes (n=136) enrolled in a behavioral intervention

	Serum concentrations, μg/ml
Intake per daya	Total carotenoids	β-carotene	α-carotene	Lycopene	Lutein/zeaxanthin	β-cryptoxanthin
	β(SE)
Fruits and Vegetables (servings)
Model lb	0.35(0.08)***	0.40(0.08)***	0.28(0.08)**	0.15(0.09)	0.21(0.08)**	0.27(0.09)**
Model 2c	0.29(0.08)**	0.37(0.08)***	0.25(0.09)**	0.12(0.10)	0.17(0.08)*	0.20(0.09)*
Fruits (servings)
Model lb	0.24(0.08)**	0.28(0.08)**	0.22(0.08)*	0.09(0.09)	0.19(0.08)*	0.23(0.09)**
Model 2c	0.20(0.08)*	0.23(0.08)**	0.18(0.08)*	0.08(0.09)	0.16(0.08)	0.18(0.09)*
Vegetables (servings)
Model lb	0.27(0.08)**	0.32(0.08)***	0.20(0.08)*	0.13(0.09)	0.12(0.08)	0.17(0.09)*
Model 2c	0.21(0.08)*	0.29(0.08)***	0.17(0.08)*	0.10(0.09)	0.08(0.08)	0.11(0.09)

^aIntake per day was assessed using 3-day diet records

^bModel 1 adjusted for age, sex, low density lipoprotein cholesterol, high density lipoprotein cholesterol, bmi z-score, glycated hemoglobin and multivitamin use

^cModel 2 adjusted for age, sex, low density lipoprotein cholesterol, high density lipoprotein cholesterol, bmi z-score, 1,5- anhydroglucitol and multivitamin use

^***p<0.001;

^**p<0.01;

^*p<0.05

Table 4.

Association of carotenoid intake with serum carotenoid concentrations at baseline in youth with type 1 diabetes (n=136) enrolled in a behavioral intervention

	Serum concentrations, μg/ml
Intake per daya	Total carotenoids	β-carotene	α-carotene	Lycopene	Lutein/zeaxanthin	β-cryptoxanthin
	β(SE)
Total carotenoids (μg)
Model lb	0.27(0.08)**	0.35(0.08)***	0.36(0.08)***	0.14(0.09)	0.05(0.08)	0.13(0.09)
Model 2c	0.22(0.08)**	0.31(0.08)***	0.33(0.08)***	0.12(0.09)	0.000(0.08)	0.06(0.09)
β-carotene (μg)
Model lb	0.16(0.09)	0.28(0.08)**	0.24(0.09)**	0.05(0.10)	−0.03(0.09)	0.08(0.09)
Model 2c	0.09(0.09)	0.24(0.08)**	0.21(0.09)*	−0.001(0.10)	−0.09(0.08)	−0.003(0.09)
α-carotene (μg)
Model lb	0.14(0.09)	0.21(0.08)*	0.22(0.09)*	0.01(0.09)	−0.07(0.09)	0.13(0.09)
Model 2c	0.06(0.09)	0.18(0.08)*	0.19(0.09)*	−0.04(0.10)	−0.13(0.08)	0.05(0.09)
Lycopene (μg)
Model lb	0.21(0.08)*	0.25(0.08)**	0.28(0.08)**	0.12(0.09)	0.04(0.08)	0.10(0.09)
Model 2c	0.19(0.08)*	0.21(0.08)**	0.25(0.08)**	0.12(0.09)	0.02(0.08)	0.06(0.09)
Lutein/ zeaxanthin (μg)
Model lb	0.22(0.08)**	0.23(0.08)**	0.20(0.08)*	0.16(0.09)	0.22(0.08)**	0.03(0.09)
Model 2c	0.17(0.08)*	0.19(0.08)*	0.17(0.08)*	0.14(0.09)	0.19(0.08)*	−0.03(0.09)
β-cryptoxanthin (μg)
Model lb	0.18(0.08)*	0.19(0.08)*	0.15(0.08)	0.05(0.09)	0.01(0.08)	0.27(0.09)**
Model 2c	0.14(0.08)	0.16(0.08)*	0.13(0.08)	0.02(0.09)	−0.03(0.08)	0.22(0.08)**

^aIntake per day was assessed using 3-day diet records

^bModel 1 adjusted for age, sex, low density lipoprotein cholesterol, high density lipoprotein cholesterol, bmi z-score, glycated hemoglobin and multivitamin use

^cModel 2 adjusted for age, sex, low density lipoprotein cholesterol, high density lipoprotein cholesterol, bmi z-score, 1,5- anhydroglucitol and multivitamin use

^***p<0.001;

^**p<0.01;

^*p<0.05

Glycemic control markers (HbA1c and 1,5-AG) and oxidative stress did not significantly moderate the relationship of FV or total carotenoid intake with serum carotenoids, as indicated by their respective interaction term estimates. Standardized regression coefficients for interaction of HbA1c with FV and total carotenoid intake were 0.03 (p=0.74) and 0.05 (p=0.56), respectively. Standardized regression coefficients for interaction of 1,5-AG with FV and total carotenoid intake were −0.09 (p=0.31) and −0.09 (p=0.31), respectively. Similarly, interaction of oxidative stress (8-iso-PGF_2α) with FV and carotenoid intake were −0.14 (p=0.09) and −0.09 (p=0.28), respectively.

Discussion

Despite a plausible pathway by which hyperglycemia could attenuate diet-serum carotenoid relationships, greater FV and carotenoid intake were related to higher serum carotenoids in this study of youth with type 1 diabetes, with the magnitude of association comparable to that observed in the general population^40–42. This supports the utility of serum carotenoids as valid markers of FV and carotenoid intake in youth with type 1 diabetes. The association of serum carotenoids with 1,5-AG, but not HbA1c, suggests the relative importance of recent hyperglycemic excursions versus longer-term overall glycemic control. In contrast to the hypothesis, oxidative stress marker was not associated with glycemic control or serum carotenoids, indicating either the marker is less sensitive or another pathway links hyperglycemia and serum carotenoids.

1,5-anhydroglucitol is a more sensitive indicator of short-term hyperglycemic peaks compared to HbA1c or fasting/ postload glucose^43,44. To the best of our knowledge, this is the first study to have explored the relationship of hyperglycemia, as reflected by 1,5-AG, with serum carotenoids. The association of 1,5-AG with serum carotenoids in this study is somewhat comparable to previous associations of plasma glucose with carotenoids in persons with normal or impaired glucose tolerance^25–27 and type 2 diabetes^25,26. Further, greater hyperglycemia, as indicated by lower 1,5-AG, was associated with lower provitamin A carotenoids (α-carotene, β-carotene and β-cryptoxanthin) in the current study, but not with non-provitamin A carotenoids (lycopene and lutein/zeaxanthin). In contrast to non-provitamin A carotenoids, provitamin A carotenoids can be metabolized into retinal and retinol⁴⁵. Thus, the current study implies reduced vitamin A synthesis from carotenoids in children with greater hyperglycemia. However, HbA1c was not associated with serum carotenoids, similar to a previous study in type 1 diabetes patients²⁹. Differences in findings between HbA1c and 1,5-AG may be attributed to the greater similarity in the time frame reflected by carotenoids⁴⁶ and 1,5-AG^47,48 (each reflecting the previous 1–2 weeks), or alternatively may suggest glycemic excursions as a stronger correlate of serum carotenoids than overall glycemic control.

Oxidative stress was not associated with serum carotenoids in this sample. Since 8-iso-PGF_2α was the only marker of oxidative stress obtained, it cannot be determined whether this null finding indicates lack of a relationship, or whether the marker is insufficiently sensitive. The latter possibility is supported by findings in other populations of association of lower plasma carotenoids with greater 8-hydroxy-2’-deoxyguanosine⁴⁹ and 2,3-dinor-5,6dihudro-15-F_2t-isoprostane (15-F_2t-IsoP-M)⁵⁰, but not with 8-iso-PGF_2α^49,50. Alternatively, the pathway linking hyperglycemia to serum carotenoids may be via mechanisms other than oxidative stress. For example, hyperglycemia is known to suppress biliary secretion⁵¹, and type 1 diabetes patients have altered bile acid profile compared to healthy controls^52,53. Since bile is essential for carotenoid absorption⁵⁴, poor glycemic control could adversely affect carotenoid absorption and serum concentrations.

Greater fruit and vegetable intake was related to higher total serum carotenoids, and serum α-carotene, β-carotene, lutein/zeaxanthin and β-cryptoxanthin; but not lycopene. The association of FV intake with total and individual serum carotenoids was comparable^40–42,55 or slightly greater^{41,42,55–58} than correlation coefficients previously reported in the general population. With the exception of lycopene, associations of greater carotenoid intake with higher serum concentrations were found for all carotenoids, comparable in magnitude to some previous research in the general population^59–62, but lower than that reported by other studies^49,63–65. The correlations were greater for α-carotene and β-carotene, and smaller for lycopene and β-cryptoxanthin compared to previous research using 3-day diet records to measure carotenoid intake in youth with type 1 diabetes²⁹. These relationships were independent of biological variables that may affect serum carotenoids, including age, sex, low density/high density lipoprotein concentrations and BMI. Further, neither glycemic control nor oxidative stress moderated the association of FV and carotenoid intake with serum carotenoids, suggesting that the utility of serum carotenoids as markers of FV and carotenoid intake is not adversely impacted by the disease process in youth with type 1 diabetes.

Results of this study are subject to certain limitations. The cross-sectional study design precludes inferences of causality. Although the analyses were adjusted for biological variables that affect serum carotenoids, food sources of carotenoids were not considered. Given that food records are susceptible to reporting error⁶⁶ and blood draw for serum carotenoid assessment preceded completion of dietary records, diet-serum carotenoid coefficients may have been potentially attenuated. Since the presence or absence of a statistical interaction is not sufficient evidence for biological processes, further mechanistic research is warranted. Further, youth with poor glycemic control (most recent HbA1c prior to recruitment >10.0%) were excluded from participation. Given that poorer glycemic control is associated with greater oxidative stress^15–17, the study may have excluded patients most affected by oxidative stress, which may have increased our type 2 error. Thus, research including those with poorer glycemic control and using more sensitive markers of oxidative stress, such as 15-F_2t-IsoP-M⁵⁰, would be informative. Finally, the predominantly white sample of the current study may limit generalizability of findings since some studies have shown differences in serum carotenoids by race/ethnicity^67,68.

Conclusions

In this study of 136 youth with type 1 diabetes, greater FV and carotenoid intake were associated with higher serum carotenoid concentrations, with magnitude comparable to that observed in other samples. As such, findings support the utility of serum carotenoids as biomarkers of carotenoid and FV intake in youth with type 1 diabetes. The association of greater 1,5-AG with lower pro-vitamin A carotenoids implies compromised concentrations of these carotenoids in type 1 diabetes youth with increased frequency of hyperglycemic excursions. Given the relationship of greater hyperglycemia with lower serum carotenoids, future studies could evaluate the use of serum carotenoids as markers of intake in individuals with type 2 diabetes.

Research question:

Are hyperglycemia and oxidative stress associated with serum carotenoids in youth with type 1 diabetes? Is the relationship of fruit and vegetable intake with serum carotenoids moderated by hyperglycemia or oxidative stress?

Key findings:

Greater fruit and vegetable intake was associated with higher serum carotenoids, supporting the utility of serum carotenoids as valid markers of fruit and vegetable intake in youth with type 1 diabetes. Additionally, greater hyperglycemia was associated with lower serum carotenoids.