Daily Intake of Recommended Servings of Fruit Improves Nutrient Intake but Shows no Major - Effect on Cardiovascular Health or Cognition in Low Fruit Consumers

Abstract

Background

Research has demonstrated that fruit consumption may benefit cardiometabolic health, cognition, and motor function; however, consumption among American adults falls well below the recommended intake of 2 cups per day.

Objectives

The hypothesis of the study was that consuming 2 cups of fruits, as recommended by the Dietary Guidelines for Americans, would improve cardiometabolic health, cognition, and motor function.

Methods

A total of 40 healthy men and women aged 30–70 y were assigned to consume 2-cup equivalents (2C) of fruits per day and to restrict fruit (RF) intake to <½ cup in a 2-period, 8-wk randomized crossover trial, with an 8-wk washout period. Anthropometrics, blood pressure, cardiometabolic biomarkers, brachial artery flow-mediated dilation, cognition, motor function, and balance measurements were assessed before and after each trial.

Results

Insulin concentrations and insulin resistance were significantly lower, and high-density lipoprotein cholesterol levels were significantly higher after the RF trial compared with the 2C trial. Processing speed scores increased during both trials, whereas 9-Hole Pegboard performance of the dominant hand improved during the 2C trial only. Energy intake, carbohydrates, fiber, vitamin K, vitamin C, thiamin, magnesium, potassium, copper, total anthocyanins, and total flavonoids intake were higher after the 2C trial compared with the RF trial. Glycemic index was lower and glycemic load was higher during the 2C trial compared with the RF trial.

Conclusions

Daily intake of 2C of fruit for 8 wk did not consistently impact cardiovascular biomarkers, body composition, cognition, or balance in habitually low fruit consumers and minimally impacted motor function; however, it improved intake of certain nutrients. Our findings suggest the need for further investigation into the optimal quantity and composition of fruit intake in diverse populations. As data accumulate, recommendations for fruit intake by the United States Department of Agriculture and the United States Department of Health and Human Services should be revisited and further studied.

This trial was registered at clinicaltrials.gov as NCT05063929.

Introduction

The 2020–2025 Dietary Guidelines for Americans recommend consuming 5 servings of fruits and vegetables per day as part of a 2000 kcal diet, including ≥2-cup equivalents (2C) from fruit specifically (2020–2025 Dietary Guidelines) [1]. Fruits and vegetables are rich in essential nutrients and other bioactive compounds that contribute to overall health. Prior observational studies suggest that diets high in fruits may help protect against the development of chronic diseases. Several studies have shown inverse relationships between consumption of fruits and vegetables and risk of cardiovascular diseases [2,3], type 2 diabetes [4], certain cancers [5,6], and all-cause mortality [7]. However, adherence to the current fruit intake recommendations remains low across all age and sex groups. In fact, fewer than 20% of individuals meet the recommendations for fruit intake (2020–2025 Dietary Guidelines) [1].

In addition to promoting heart health and reducing risks for diabetes and metabolic syndrome, epidemiological studies [8,9] and intervention trials of specific fruits [10] suggest that fruit consumption may improve cognitive function. However, low intake of fruit and other plant foods has been associated with advancing cognitive decline in older adults [11,12]. Epidemiological research [13] also links fruit and vegetable consumption to age-related declines in balance, muscle strength, and coordination, all of which are key risk factors for falls. Notably, 1 intervention study demonstrated improvements in similar outcomes even when a fruit intervention is initiated later in life [14].

Fruits are good sources of fiber, numerous vitamins, potassium, magnesium, and bioactive phytochemicals, which have been linked to cardiometabolic health and lower risk factors for chronic diseases [15,16]. Additionally, fruit consumption promotes healthy weight regulation by maintaining satiety [17,18]. Several small randomized controlled trials (RCTs) have assessed the effect of individual fruits or fruit juices on biomarkers of cardiovascular disease and vascular function, particularly brachial artery flow-mediated dilation (FMD), a key indicator of endothelial health [[19], [20], [21], [22], [23], [24]]. However, there is a need to assess the impact of a variety of fruits in a single intervention because 1) individuals rarely consume a single fruit as the sole food within this food group, 2) the nutrient mixture provided by a variety of fruits may be complementary with regard to health outcomes, and 3) focusing on a single fruit would limit the generalizability of the study to fruits overall. Despite clear guidelines from the USDA and other organizations, much of our knowledge on the consumption of a fruit-rich diet is limited to epidemiological data, highlighting the need for more comprehensive RCTs of multifruit interventions. Overall, the majority of studies linking fruit consumption to positive health outcomes are correlational, with a very limited number of clinical trials available. The few existing clinical studies, such as the FLAvonoids and Vascular function at the University of Reading Study (FLAVURS) based on United Kingdom dietary guidelines [25], and trials on low- and high-carotenoids [26] used a combination of fruits and vegetables high in flavonoids or carotenoids, respectively. Furthermore, the general public typically consumes a variety of fruits rather than only a single fruit. To our knowledge, the health effects of consuming a variety of commonly consumed fruits as observed in a typical diet have not been thoroughly studied. Therefore, the primary aim of this study was to investigate whether fruit, consumed in amounts recommended by the dietary guidelines, will improve cardiometabolic outcomes, cognition, and motor function among individuals with habitually low fruit consumption.

Research Design and Methods

Study design and participants

This study was conducted at San Diego State University from August 2021 until August 2023. A total of 40 healthy men and women aged 30–70 y who reported typically eating <2C of fruit per day were enrolled to participate. Using a randomized crossover study design (Figure 1), participants were assigned to complete 2 8-wk trials in random order. Participants were randomly allocated to trial order using a computer-generated list of random numbers. Blinding of trial assignment was not possible. For 1 trial, participants consumed 2C per day of the 6 most commonly consumed fruits in the United States (bananas, apples, oranges, grapes, strawberries, and watermelon provided on a rotating basis), whereas during the other trial, participants were requested to restrict fruit (RF) consumption to <½ cup equivalents of fruits per day and to include only bananas, apples, oranges, grapes, strawberries, and watermelon. An 8-wk washout period occurred between the trials, where participants were instructed to return to their habitual fruit intake.

FIGURE 1.

Flow diagram of enrollment and follow-up. 2C, 2 cup equivalents; RF, restricted fruit.

Participants were recruited via posted flyers around the San Diego State University campus, on social media, within neighboring communities, and through word of mouth in San Diego. Exclusion criteria included gastrointestinal disorders (e.g., Crohn’s, celiac, inflammatory bowel disease, irritable bowel syndrome, etc.), allergy or intolerance to any of the study fruits, previous stroke or myocardial infarction, clinically diagnosed diabetes, liver or renal disease, drug treatment for lipid reduction or hypertension, current chronic use of anti-inflammatory medication such as non-steroidal anti-inflammatory drugs (NSAIDS), pregnancy or lactation, severe diseases including neurological disorders, known metabolic disorders, psychiatric medication use, a mini-mental state examination (MMSE) score of <26 points, a BMI <18 kg/m² or >32 kg/m², regular use of aspirin, daily consumption of >50 g alcohol, >10 cigarettes, or >6 cups of coffee (>450 mg caffeine/d), and use of antioxidant supplements for the prior 2 mo. In addition, a food frequency questionnaire (FFQ) was administered to determine habitual daily intake of fruits, and individuals consuming 2C/d or more were excluded [27].

Participants reported to the study site while fasted ≥10 h at baseline and post intervention during each trial. Participants were required to abstain from consuming alcohol and caffeine for 12 h before testing. To ensure adequate hydration, participants were asked to consume 20 ounces of water after waking. All subjects provided informed consent. Fasting blood samples were collected in evacuated tubes for serum and plasma and centrifuged at 4°C at 1000 × g for 10 min. Aliquots were stored at −80°C until further analysis.

Study intervention

For the 2C trial, 2C of fruit per day were provided to participants on a weekly basis. Participants were provided with a 1-wk cycle menu of provided fruits that repeated for each of the 8 wk. Two portions of different fruits totaling 2C were provided each day. Throughout the course of the week, the cycle menu included an equal distribution of each fruit to be consumed. This allowed for the supply of a variety of selected fruits to be consumed daily. Participants were instructed to strictly avoid all other fruits. To maintain the freshness of the fruits provided, a prerequisite for entering the study was that participants had adequate refrigerator or cold-storage space for the fruits provided in 1-wk supplies. Similar portions of select fruits (bananas, apples, oranges, grapes, strawberries, and watermelon) were chosen based on the most commonly consumed fruits in the United States [28]. The weekly menu cycle provided to participants with their fruit portions weekly is provided in Supplemental Table 1. During the final 3 d of the 2C trial, all fruits were provided as a blended smoothie to ensure that all fruits were consumed for the 3 d before the final testing. During the RF trial, participants were asked to monitor their fruit intake and to consume no more than the ½ cup equivalents of the provided list of fruits, which included only the fruits to be studied, to ensure that the limited intake of fruit contained the same types of phytonutrients consumed during the higher fruit intervention trial. Participants were asked to maintain their current lifestyles throughout the study period including physical activity and intake of other foods, with an emphasis on maintaining their vegetable consumption. Compliance was monitored by both the printed menu forms and daily fruit intake electronic forms sent via text message or email, which served as daily communication with participants to enhance adherence. Participants’ answers to a medical history questionnaire were recorded at baseline, and the Block 2014 FFQ + Physical Activity Screener (Nutrition Quest) was administered at each visit to assess dietary intake and physical activity over the previous month. Changes to physical activity were determined by total energy expenditure.

Anthropometric measures, blood pressure, and biomarkers

Anthropometric measures included height, weight, waist circumference (WC), and body composition, which was measured using multifrequency, multisegmental bioelectrical impedance analysis (BIA, InBody). Height was taken by measuring each participant’s stature against the wall without shoes using a stadiometer and measuring to the nearest fifth centimeter. WC was measured by using a tape measure wrapped around the participant's waist 1 inch above the navel, underneath clothing. Before the body composition assessment, participants were instructed to void their bladders and remove shoes, socks, jewelry, and excess clothing. During the measurements, participants held the electrodes in each hand, while standing on the BIA for 60 s with their arms away from their body to assess fat mass (FM) and fat-free mass (FFM) in kilograms recorded to the nearest tenth. Resting blood pressure was measured using an automated blood pressure cuff (M3, Omron Healthcare, Inc.) with participants in a seated position. Concentrations of glucose, triglycerides (TG), total cholesterol (TC), and HDL cholesterol were measured using colorimetric assay kits from EKF. Insulin and C-reactive protein (CRP) were measured using ELISA kits (ALPCO), and total antioxidant capacity (TAC) and glutathione peroxidase (GPx) were measured with kits from Cayman Chemical. Insulin resistance was calculated by assessing the HOMA-IR using the equation [insulin (mU/mL)∗glucose (mmol/L)/22.5] [29]. LDL cholesterol was calculated by using the equation of Friedewald [30] as LDL cholesterol = TC minus HDL cholesterol minus TG/5.

Flow-mediated dilation

Brachial artery FMD was assessed noninvasively using a Philips CX-50 ultrasound machine equipped with a 12 MHz transducer. The right arm was adducted at heart level, and the brachial artery was located 3–6 cm above the antecubital crease. To ensure the location of the same arterial segment with serial measurements, anatomical landmarks were recorded. The ultrasound probe was then clamped to avoid any involuntary movement. After obtaining baseline diameters for 30 s, reactive hyperemia was produced by inflating a blood pressure cuff placed on the lower forearm for 5 min at 232 mmHg of pressure, followed by rapid deflation. The brachial artery was scanned continuously until 2 min postocclusion to obtain the peak dilatory response. Ultrasound images were recorded for offline analysis [31] by a single investigator. FMD was calculated as absolute (Δmm) and percentage change in brachial artery diameter in response to the forearm hyperemic stimulus. Commercially available software package (Vascular Analysis Tools 5.8.1; Medical Imaging Applications) was used to acquire and analyze electrocardiogram-gated brachial artery diameters.

Cognition and motor function

Before performing cognitive and motor function measurements, a small, controlled breakfast (350 kcal) of a refined food (vanilla muffin) and water (8 oz) was provided to avoid feelings of hunger during testing. At baseline and after 8 wk of treatment, a comprehensive battery of tests was used to measure cognition and motor function.

Cognitive function was evaluated using the NIH Toolbox-Cognition Battery [[32], [33], [34]], which entails 5 computerized tests that assess 4 major cognitive subdomains including processing speed [pattern comparison processing speed test (PCPST)], executive function and attention [dimensional change card sort test (DCCST) and Flanker inhibitory control and attention test (FICAT)], working memory [list sorting working memory test (LSWMT)], and episodic memory [picture sequence memory test (PSMT)].

Motor function was assessed using select measures of the NIH Toolbox Motor Battery [35] including domains of strength (Grip Strength Test) and dexterity (9-Hole Pegboard Dexterity Test). Manual dexterity was also assessed using a series of Purdue Pegboard tasks (Lafayette Instruments). Fine motor strength was assessed using a Lafayette pinch gauge (Lafayette Instruments).

Balance testing

The Btracks Balance Plate and BTracks Balance System (Balance Tracking Systems, Inc.) were used to assess postural sway in bipedal stance. This system has previously been shown to have high test–retest reliability and to be resistant to learning effects [36,37]. The plate was set on a firm, level surface, and participants were asked to remove their shoes and stand on the plate with their hands on their hips and eyes closed. For each trial, the center of pressure deviation was measured in triplicate in centimeters over the course of 20 s.

Statistical analysis

Statistical analyses were conducted using JASP version 14 (JASP Team 2022). Data were tested for normality using the Shapiro–Wilk test with normality defined as P > 0.05. Log-transformed data were used for data that violated normality. The Friedman test was used for nonparametric analysis of data that did not meet normality after log transformation. Differences in study outcomes between trials were analyzed using a 2 (time) × 2 (trial) repeated measures analysis of variance with trial and time (baseline and 8-wk postintervention) as within-subject factors. For significant interactions or main effects, paired t-tests were used for post-hoc analyses. Different responses by sex were explored. Log-transformed data were used for outcomes of weight, HDL cholesterol, TC, CRP, GPx, pinch grip using the dominant hand, and the 9-Hole Pegboard using the nondominant and dominant hand, food energy, protein, carbohydrate, fat, dietary fiber, cholesterol, vitamin A, vitamin E, vitamin K, thiamin, riboflavin, niacin, total folate, calcium, magnesium, phosphorus, potassium, copper, and total flavonoids. The Friedman test was used for outcomes of FFM, glucose, TG, insulin, HOMA-IR, FMD, grip strength using the nondominant and dominant hand, pinch grip using the nondominant hand, Purdue Pegboard using the left hand, Purdue Pegboard using both hands, Purdue Pegboard assembly task, balance, PSMT, vitamin C, cobalamin, sodium, and total anthocyanidins. Wilcoxon signed-rank test was used as follow-up when necessary. To verify the effectiveness of the crossover design, we checked for an effect of treatment sequence using a linear mixed-effects model, including randomization sequence as a fixed effect and participant nested within sequence as a repeated effect. Data are presented as means ± SD. Significance was defined as P < 0.05. Using data from Siasos et al. [38], a priori sample size estimation suggested that 18 participants per group would provide 80% power at an α level of 0.05 for changes in FMD between trials. Allowing for higher statistical power and a greater potential to detect differences in other variables, as well as to account for potential attrition, an initial sample size of 40 participants was recruited.

Results

A total of 38 participants completed the study (24 women, 14 men) and were included in the final analysis (Figure 1), with the exception of 28 for FMD analysis and 37 for the balance testing analysis. Baseline characteristics of participants are presented in Supplemental Table 2. Participants’ MMSE score ranged from 27 to 30 (29.1 ± 1), and age ranged 30–70 y (45 ± 13). Compliance during both the RF and 2C trials was measured as the mean amount of fruits consumed per day during the 8-wk intervention: RF: 0.23 ± 0.20 cups per day; 2C: 1.96 ± 0.09 cups per day. The effect of randomization sequence was not statistically significant for any outcome variable (P > 0.05), indicating no evidence of sequence- or carryover-related effects.

Anthropometric measures, blood pressure, biomarkers, and FMD

There were no significant differences for BMI, weight (Wt), FFM, FM, WC, systolic blood pressure, and diastolic blood pressure within or between trials (Table 1). HDL cholesterol concentrations were significantly lower after the 2C trial (P < 0.05) compared with the RF trial (Table 2). Insulin concentrations and HOMA-IR decreased (P < 0.05) from baseline to 8 wk during the RF trial and were significantly lower (P < 0.05) after the RF trial compared with the 2C trial. There were no significant differences (P > 0.05) for glucose, TG, LDL-C, TC, CRP, TAC, and GPx. FMD revealed no statistically significant interaction or main effects of trial or time (P > 0.05).

TABLE 1.

Impact of fruit intake on anthropometric measurements, body composition, and blood pressure.

	RF trial		2C trial		P value
	Baseline	8 wk	Baseline	8 wk	Time × trial	Trial	Time
BMI (kg/m2)	25.1 ± 3.0	25.1 ± 3.2	25.1 ± 3.2	25.2 ± 3.2	0.141	0.458	0.387
Weight (kg)	74.7 ± 13.7	74.5 ± 14.2	74.7 ± 14.4	75.1 ± 14.5	0.061	0.538	0.651
FFM (kg)	53.2 ± 12.1	53.0 ± 11.9	53.3 ± 12.0	53.3 ± 12.0	0.787	0.327	0.502
FM (kg)	21.5 ± 7.3	21.5 ± 7.2	21.4 ± 7.4	21.8 ± 7.2	0.284	0.736	0.129
WC (cm)	83.5 ± 9.9	83.0 ± 10.7	84.3 ± 11.4	84.3 ± 10.5	0.574	0.052	0.367
SBP (mmHg)	127 ± 16	126 ± 15	124 ± 14	125 ± 17	0.905	0.146	0.969
DBP (mmHg)	77 ± 10	77 ± 9	77 ± 9	76 ± 9	0.773	0.204	0.943

Values are means ± SDs, n = 38.

Abbreviations: 2C, 2-cup equivalents; DBP, diastolic blood pressure; FFM, fat-free mass; FM, fat mass; RF, restricted fruit; SBP, systolic blood pressure; WC, waist circumference.

TABLE 2.

Impact of fruit intake on cardiovascular markers.

	RF trial		2C trial		P value
	Baseline	8 wk	Baseline	8 wk	Time × trial	Trial	Time
Glucose (mg/dL)	94.1 ± 11.9	92.7 ± 11.6	92.3 ± 11.9	92.9 ± 13.3	0.250	0.178	0.756
Triglycerides (TG) (mg/dL)	70.6 ± 44.1	69.7 ± 65.8	69.9 ± 37.7	75.3 ± 75.2	0.300	0.502	0.718
HDL cholesterol (mg/dL)	52.1 ± 14.1	53.9 ± 14.6	51.6 ± 15.0	50.5 ± 14.71	0.154	0.024	0.810
LDL cholesterol (mg/dL)	116.0 ± 43.6	118.9 ± 42.5	120.0 ± 42.1	116.5 ± 40.9	0.196	0.783	0.930
Total cholesterol (TC) (mg/dL)	182.2 ± 44.1	186.7 ± 41.9	185.5 ± 38.8	182.0 ± 40.9	0.073	0.985	0.873
Insulin (μIU/mL)	8.0 ± 6.6	5.9 ± 5.22	7.6 ± 7.0	8.0 ± 5.91	0.018	0.109	0.235
Insulin resistance (HOMA-IR)	1.9 ± 1.8	1.3 ± 1.32	1.7 ± 1.8	1.9 ± 1.81	0.019	0.111	0.123
C-reactive protein (CRP) (mg/dL)	3.0 ± 3.3	2.9 ± 3.8	2.4 ± 2.9	3.5 ± 5.0	0.908	0.125	0.679
Total antioxidant capacity (TAC) (mM)	2.42 ± 0.49	2.39 ± 0.50	2.47 ± 0.43	2.41 ± 0.40	0.584	0.519	0.293
Glutathione peroxidase (GPx) (mM)	171.6 ± 31.3	167.5 ± 26.2	172.3 ± 32.9	166.5 ± 26.8	0.757	0.926	0.194
Flow-mediated dilation (FMD) (%)	5.77 ± 4.34	5.71 ± 5.12	5.58 ± 3.39	6.58 ± 6.21	0.487	0.134	0.549

Values are means ± SDs, n = 38.

Abbreviations: 2C, 2-cup equivalents; RF, restricted fruit.

¹Indicates (P < 0.05) for between-group difference at the same time point.

²Indicates (P < 0.05) for within-group difference.

Although differences between sexes were not anticipated, exploratory analyses indicated that serum glucose and insulin concentrations and HOMA-IR significantly decreased from baseline to 8 wk during the RF trial for women subjects only (P < 0.05). CRP concentration was significantly lower during the 2C trial compared with the RF trial at 8 wk only in men (P < 0.05).

Cognition, motor function, and balance

LSWMT scores were significantly higher at baseline for the RF trial (P < 0.05) compared with the 2C trial, and no effects of interventions were detected (Table 3). PCPST scores after 8 wk were significantly higher (P < 0.05) than the scores at baseline during both trials. PSMT, FICAT, or DCCST scores did not show any changes between or within the trials. A main effect (P < 0.05) for faster time to complete the 9-Hole Pegboard using the dominant hand was detected, which included a statistically (P < 0.05) shorter (4.5% less) time after the 2C trial than the baseline during the 2C trial; however, no difference relative to the RF trial was detected (Table 4). A main effect of time (P < 0.05) was observed for increasing Purdue Pegboard pin insertions using the left hand, but follow-up measures revealed no difference within either trial. No significant changes were observed for grip strength, pinch grip, 9-Hole Pegboard using the nondominant hand, Purdue Pegboard using the right hand, both hands and assembly task, or the balance test within or between trials.

TABLE 3.

Impact of fruit intake on cognitive function measurements.

	RF trial		2C trial		P value
	Baseline	8 wk	Baseline	8 wk	Time × trial	Trial	Time
LSWMT	19 ± 3	19 ± 3	18 ± 31	18 ± 2	0.480	0.014	0.189
PCPST	50 ± 9	54 ± 92	50 ± 10	53 ± 102	0.837	0.755	0.001
PSMT	566.9 ± 95.2	575.4 ± 101.5	560.0 ± 108.6	580.6 ± 113.5	0.527	0.939	0.093
FICAT	8.41 ± 0.67	8.48 ± 0.76	8.43 ± 0.66	8.51 ± 0.67	0.937	0.750	0.134
DCCST	8.46 ± 1.00	8.56 ± 0.95	8.44 ± 0.87	8.38 ± 1.01	0.402	0.218	0.839

Values are means ± SDs, n = 38.

Abbreviations: 2C, 2 cup equivalents; DCCST, dimensional change card sort test; FICAT, Flanker inhibitory control and attention test; LSWMT, list sorting working memory test; PCPST, pattern comparison processing speed test; PSMT, picture sequence memory test; RF, restricted fruit.

¹Indicates (P < 0.05) for between-group difference at the same time point.

²Indicates (P < 0.05) for within-group difference.

TABLE 4.

Impact of fruit intake on motor function and balance.

	RF trial		2C trial		P value
	Baseline	8 wk	Baseline	8 wk	Time × trial	Trial	Time
GS ND (lbs)	69.5 ± 29.7	68.7 ± 27.1	68.1 ± 29.6	69.6 ± 29.8	0.257	0.797	0.470
GS D (lbs)	72.6 ± 31.4	73.6 ± 29.2	74.4 ± 30.5	75.3 ± 31.0	0.956	0.181	0.503
Pinch ND (lbs)	17 ± 6	17 ± 5	17 ± 5	17 ± 5	0.778	0.701	0.742
Pinch D (lbs)	18 ± 7	19 ± 6	18 ± 6	19 ± 6	0.888	0.916	0.134
9-Hole Peg ND (s)	21.13 ± 2.79	21.06 ± 2.61	20.92 ± 2.83	21.18 ± 3.05	0.578	0.791	0.710
9-Hole Peg D (s)	20.38 ± 2.85	19.73 ± 2.54	20.30 ± 3.00	19.37 ± 2.121	0.674	0.522	0.003
Peg RH	15.7 ± 1.8	15.6 ± 2.1	15.7 ± 2.3	16.0 ± 2.0	0.084	0.410	0.421
Peg LH	14.7 ± 1.8	14.9 ± 1.7	14.8 ± 2.1	15.2 ± 1.9	0.391	0.412	0.035
Peg both	12.3 ± 1.7	12.6 ± 1.7	12.3 ± 2.0	12.8 ±1.8	0.431	0.906	0.052
Peg ASM	8.2 ± 1.9	8.7 ± 2.2	8.6 ± 2.4	8.7 ± 2.4	0.125	0.587	0.103
Balance (cm)	27.58 ± 16.77	26.45 ± 14.08	26.04 ± 17.42	26.01± 14.67	0.457	0.231	0.917

Values are means ± SDs, n = 38.

Abbreviations: 2C, 2 cup equivalents; D, dominant; GS, grip strength; ND, nondominant; Peg ASM, Purdue Pegboard assembly; Peg Both, Purdue Pegboard both hands; Peg LH, Purdue Pegboard left hand; Peg RH, Purdue Pegboard right hand; RF, restricted fruit.

¹Indicates (P < 0.05) for within-group difference.

Dietary intake and physical activity

Participants consumed significant more energy, carbohydrate, fiber, vitamin K, vitamin C, thiamin, magnesium, potassium, copper, total anthocyanidins, and total flavonoids during the 2C trial (P < 0.05) compared with the RF trial after 8 wk (Table 5). Consumption of protein, carbohydrate, fat, fiber, vitamin E, vitamin K, vitamin C, thiamin, niacin, iron, magnesium, phosphorus, potassium, and copper significantly decreased from baseline (P < 0.05) to 8 wk during the RF trials. However, consumption of fiber, vitamin C, total anthocyanidins, and total flavonoids significantly increased from baseline (P < 0.05) to 8 wk during the 2C trial. Interestingly, glycemic index was lower (P < 0.05); however, glycemic load was significantly higher secondary to shifts in total carbohydrate consumption after 8 wk during the 2C trial compared with the RF trial at 8 wk (P < 0.05). Additionally, glycemic load was lower at 8 wk compared with baseline during the RF trial (P < 0.05). No differences (P > 0.05) in physical activity energy expenditure (kcal) were detected within groups or between trials (RF baseline: 991 ± 654; RF 8 wk: 1058 ± 1040; 2C baseline: 926 ± 776; 2C 8 wk: 829 ± 727).

TABLE 5.

Impact of fruit intake on nutrient consumption.

	RF trial		2C trial		P value
	Baseline	8 wk	Baseline	8 wk	Time × trial	Trial	Time
Food energy (kcals)	1595 ± 737	1352 ± 568	1586 ± 691	1550 ± 6611	0.051	0.118	0.044
Protein (g)	65.7 ± 35.4	54.3 ± 23.42	64.3 ± 30.3	60.3 ± 28.1	0.200	0.300	0.013
Carbohydrate (g)	166.7 ± 80.3	137.5 ± 63.42	167.8 ± 77.9	178.3 ± 77.61	0.001	0.004	0.118
Fat (g)	71.1 ± 35.5	60.7 ± 26.92	71.1 ± 33.1	64.8 ± 33.1	0.719	0.733	0.018
Dietary fiber (g)	16.3 ± 10.1	13.0 ± 8.62	16.9 ± 10.2	18.9 ± 11.01	0.001	0.001	0.274
Cholesterol (g)	217.4 ± 148.6	214.9 ± 133.8	224.2 ± 137.6	212.5 ± 125.1	0.717	0.478	0.839
Vitamin A (μg)	735.0 ± 447.0	655.7 ± 365.9	728.3 ± 476.6	744.0 ± 550.3	0.489	0.706	0.304
Vitamin E (mg)	9.1 ± 4.8	7.2 ± 3.82	8.9 ± 5.3	8.7 ± 5.7	0.155	0.373	0.021
Vitamin K (μg)	144.8 ± 111.7	127.1 ± 105.82	181.9 ± 250.2	182.4 ± 283.41	0.377	0.048	0.047
Vitamin C (mg)	63.6 ± 34.1	49.8 ± 31.52	74.6 ± 55.6	105.6 ± 59.31	0.001	0.001	0.478
Thiamin (mg)	1.3 ± 0.7	1.1 ± 0.72	1.2 ± 0.6	1.2 ± 0.61	0.043	0.440	0.015
Riboflavin (mg)	1.6 ± 0.7	1.5 ± 0.7	1.7 ± 0.8	1.6 ± 0.6	0.716	0.234	0.068
Niacin (mg)	19.9 ± 10.0	16.5 ± 7.92	19.0 ± 8.9	17.9 ± 8.7	0.156	0.619	0.012
Total folate (μg)	342.8 ± 184.4	301.6 ± 178.7	340.1 ± 186.8	334.1 ± 193.3	0.956	0.800	0.800
Cobalamin (μg)	4.1 ± 2.3	3.8 ± 2.3	4.0 ± 2.4	3.9 ± 2.2	0.425	0.510	0.187
Calcium (mg)	813.3 ± 365.2	740.4 ± 333.9	813.9 ± 379.0	733.5 ± 290.4	0.083	0.452	0.083
Iron (mg)	12.1 ± 6.9	10.4 ± 6.02	11.6 ± 6.4	10.9 ± 5.6	0.159	0.641	0.023
Magnesium (mg)	279.6 ± 133.9	235.6 ± 107.82	276.5 ± 130.8	268.3 ± 128.71	0.078	0.169	0.011
Phosphorus (mg)	1149.0 ± 550.1	998.0 ± 429.92	1142.3 ± 509.3	1038.5 ± 457.9	0.524	0.700	0.009
Potassium (mg)	2240.1 ± 924.8	1867.3 ± 763.82	2310.1 ± 987.3	2386.5 ± 997.21	0.004	0.003	0.059
Sodium (mg)	2807.0 ± 1415.6	2378.7 ± 1008.2	2799.4 ± 1308.7	2577.2 ± 1291.3	0.394	0.418	0.020
Copper (mg)	1.2 ± 0.6	1.1 ± 0.52	1.2 ± 0.6	1.3 ± 0.61	0.017	0.142	0.308
Total anthocyanidins (mg)	12.6 ± 9.2	9.7 ± 8.0	17.7 ± 20.2	40.5 ± 22.61	0.001	0.001	0.205
Total flavonoids (mg)	136.0 ± 172.3	113.3 ± 127.7	139.7 ± 146.2	182.6 ± 179.81	0.001	0.001	0.116
Glycemic index	50.6 ± 2.9	51.4 ± 2.8	50.6 ± 3.2	49.8 ± 3.61	0.045	0.105	0.685
Glycemic load	76.7 ± 38.1	64.3 ± 30.72	76.4 ± 36.8	80.1 ± 38.61	0.004	0.022	0.141

Values are means ± SDs, n = 38.

Abbreviations: 2C, 2 cup equivalents; RF, restricted fruit.

¹Indicates (P < 0.05) for between-group difference at the same time point.

²Indicates (P < 0.05) for within-group difference.

Discussion

This RCT demonstrated that in comparison to restriction of fruit intake, consuming fruit at the level recommended by the Dietary Guidelines for Americans (2C per day) for 8 wk showed limited effectiveness to promote improvements of cardiovascular biomarkers, body composition, cognition, or balance in low fruit consumers and only minimally impacted motor function. However, consuming 2 cups per day of the 6 most commonly consumed fruits in the United States (bananas, apples, oranges, grapes, strawberries, and watermelon) promoted intake of certain nutrients in comparison to both usual intake and in comparison to the fruit restriction trial.

Ample evidence from previous literature supports the long-term benefits of consuming a nutrient-rich diet, suggesting that our intervention may not have been of a sufficiently long duration to detect changes in key health outcomes between the 2 trials. Several observational studies demonstrate that greater adherence to the Dietary Guidelines for Americans is associated with lower risk of developing cardiovascular diseases and diabetes, as well as cause-specific and all-cause mortality [39,40]. Additionally, adherence to the dietary approaches to stop hypertension (DASH) diet, which is also rich in fruits, lowers the risk of all-cause mortality and cancers [41].

Consuming 2 cups of fruits per day did not influence glycaemia and insulinemia; however, improvement of insulin concentrations and HOMA-IR during the RF trial might be attributed to the overall decrease in consumption of sugars compared with the 2C trial. It is also possible that, in comparison to when the subjects were consuming 2 cups of fruit, during the RF trial, the participants may have substituted the fruit servings with foods that promoted adverse effects on insulin concentrations. Although participants were not daily habitual consumers of 2 cups of fruit, it is worth noting that insulin concentrations and HOMA-IR ratios did not significantly increase from baseline to 8-wk follow-up during the 2C trial.

Notably, HDL cholesterol concentrations were modestly decreased during the 2C trial. A potential explanation for this may be higher carbohydrate/sugar intake in comparison to the RF trial, despite participants consuming more dietary fiber during the 2C trial. Although not significant, the very slight tendencies for LDL cholesterol concentrations to decrease from baseline to 8 wk during the 2C trial, while increasing from baseline to 8 wk during the RF trial, suggest that a longer feeding period or perhaps higher intake of fruit may have yielded more positive results. Research on LDL cholesterol and HDL cholesterol highlights their roles in cardiovascular disease risk, suggesting higher HDL cholesterol levels are generally protective [42]. However, some studies indicate excessively high HDL cholesterol concentrations may also pose risks. A recent retrospective study showed that HDL cholesterol levels of over 80 mg/dL had a 96% higher risk of all-cause mortality and a 71% higher risk of cardiovascular mortality compared with those with HDL cholesterol levels in the range of 40–60 mg/dL [43]. Although the current study demonstrated a minor decrease in HDL cholesterol after consumption of 2 cups of fruit daily for 8 wk, mean concentrations remained within a protective range.

Although we did not observe statistically significant changes to TC levels, improvements in TC concentrations have been observed in previous studies assessing fruit consumption on cardiometabolic biomarkers. One study found that consuming 2 medium-sized apples daily for 6 mo significantly decreased TC concentrations in postmenopausal women compared with dried plums [44]. Notably, however, apples served as only 1 of the 6 provided fruits in the current study. The modest tendency for a trial × time interaction in the current study, with very modestly decreasing TC levels in the 2C trial and increasing TC in the RF trial, warrants further research with a larger sample size and perhaps longer duration in a more controlled setting (feeding trials). The potential effects can be partly attributed to the role that fiber and phytochemicals play in improving lipid responses. A cross-sectional study investigated the role of flavonoids on anthropometrics and blood biomarkers and identified an inverse relationship between the flavonoid quercetin (high levels in grapes and apples) and TC levels among Japanese women [45]. Additionally, fibers such as pectin (found in many fruits fed in this study) have been reported to reduce blood cholesterol by reducing the reabsorption of bile acids, increasing synthesis of bile acids from cholesterol, and ultimately reducing the circulating blood cholesterol [46].

Although no significant changes in weight or body composition were observed in the present study, previous studies have demonstrated that higher intake of fruits and vegetables promotes weight reduction [17]. In that study, implementing consumption of water-rich foods like fruits and vegetables was more effective at maintaining satiety, controlling hunger, and improving physiological measures than reducing fat intake alone. Although the current study implemented the consumption of more water-rich foods (fruits), participants were not instructed to make any other changes to the diet, nor were they encouraged to replace usual snacks with the provided fruit. Future studies may consider replacing 1 snack per day with 2 cups of fruit to further examine fruits’ impacts on weight regulation and body composition.

This study detected no significant changes in brachial artery FMD with nonparametric testing of the data, which were not normally distributed; however, after log-transforming the data, analyses revealed a potential improvement in FMD for the 2C trial at 8 wk. Importantly, the data remained non-normally distributed, which renders them inconclusive. Because research has demonstrated that daily consumption of other foods from the fruit group (freeze-dried blueberries) increases FMD in healthy older adults after 12 wk [47], further research on the impacts of consuming other fruits on endothelial function is warranted. Moreover, consumption of the fruits provided in the present study increased consumption of flavonoids, further highlighting the potential for these fruits to promote vascular health.

Many of the fruits provided to the participants in the present study, including grapes, strawberries, apples, bananas, and oranges, are rich in flavonoids. Some studies have demonstrated that flavonoid-rich fruits may improve cognition in adults, perhaps through increasing cerebrovascular blood flow [48]; however, the current findings showed that the daily consumption of 2 cups of fruit for 8 wk did not improve cognition compared with 8 wk of fruit restriction. Speed on attention tasks improved in healthy young adults aged 18–35 y old after the acute consumption of 230 mL of purple grape juice [49]. However, results from the current study did not show any significant improvements in attention following the 2C trial. Another study found that the consumption of a daily serving of 200 mL of anthocyanin-rich cherry juice improved verbal learning, memory, and executive function after 12 wk in adults 70 y old and older with mild-to-moderate dementia [50]. In a separate study, global cognitive function significantly improved in healthy older adults following the daily consumption of 500 mL of a high-flavanone orange juice for 8 wk [51]. Although these studies demonstrate how individual fruits containing a variety of bioactive phytochemicals may benefit cognition, results from the current study were inconsistent with these findings. Performance on the test for processing speed seemed to improve over time during both trials, indicating a possible learning effect. Overall, no clinically significant improvements in processing speed, executive function, working memory, or episodic memory were observed after consuming 2C of fruit for 8 wk. One possible reason for the lack of significant findings is that a variety of whole fruits was provided to the participants in the current study rather than a single fruit or fruit juice; therefore, the participants in the current study may have consumed a variety of phytonutrients at a lower dose instead of receiving a high dose of 1 specific phytonutrient as seen in previous studies investigating only single fruits. It is unclear if consuming additional servings of the fruits provided or a longer intervention duration would have altered the outcomes. Likewise, it is possible that the fruits provided may have yielded different results in a population with mild cognitive impairment.

In the present study, the consumption of 2 cups of fruit per day for 8 wk only improved 9-Hole Pegboard in the dominant hand and did not significantly impact other motor function or balance measurements. Clinical studies have linked changes in walking speed and gait variability to declines in cognitive ability and life expectancy among older adults during normal aging [[52], [53], [54], [55], [56]]; therefore, nutritional interventions may offer an opportunity to impact motor function through changes in cognitive performance. Conversely, another study found that the daily consumption of 24 grams of freeze-dried strawberries for 90 d improved spatial learning and memory but did not benefit gait or balance in healthy older adults [57]. It is unknown whether or not motor function would have improved in the current study if there had been a significant improvement in cognition. Perhaps more sensitive measures were needed to better assess motor function. Additionally, gait and walking speed were not measured in the current study.

One limitation of this study is the lack of overall participant control, as the only factor controlled was fruit consumption. Although participants were instructed to maintain their regular diet and physical activity for the duration of the study, it is possible that some individuals altered their diets or exercise habits, which could have impacted the results of the study. The lack of formal assessment of overall dietary pattern changes, including healthy eating index (HEI)-scores or food component substitutions, particularly during the RF phase, may have influenced outcomes and should be considered when interpreting the findings. Future studies may consider providing participants with guidelines for overall dietary intake during the study duration or utilizing a more controlled feeding paradigm. A second limitation was that the results indicated possible learning effects for the measure of processing speed. This reveals that a longer washout period may have been necessary to prevent any learning effects. Future studies may consider increasing the length of the trials and the length of the washout period. A third limitation was that the baseline values for the LSWMT score were significantly different between the trials. This makes it difficult to accurately interpret the results for this measure when comparing the 2 trials. However, we utilized paired-comparisons t-tests to determine if the change over time between trials was different and detected no difference (P > 0.05). We acknowledge that we used this post-hoc analyses without adjustment for multiple comparisons, which may increase the risk of type I error. As such, these findings should be interpreted with caution. Results of this study should be interpreted with the knowledge that participants were a relatively homogenous group of healthy, middle-aged volunteers despite habitually low fruit consumption, that the intervention was only 8 wk in duration, and that we only compared a fruit restriction to a single dose (2C per day) of the 6 most commonly consumed fruits in the United States.

Although the study has limitations, it also had many strengths including the randomized crossover design of the study, which minimized the variability of confounding factors. This was also the first RCT to administer a multifruit intervention designed to assess the adequacy of Dietary Guidelines for Americans for fruit. It was important to include a variety of fruits rather than a single fruit to represent a general pattern of fruit consumption. Another strength of the study is the use of whole fruits rather than processed fruits. Although other studies have used freeze-dried, powdered fruit juice, or dehydrated forms of fruit for better control tactics, these results cannot be applied to the ways in which consumers typically eat fruit. Additionally, powdered forms of fruit interfere with the matrix of the food, which may impact the metabolism of nutrients [15]. Lastly, the participants were mostly middle-aged adults, which is a population that has not been extensively studied in regard to the effects of fruit consumption on cognition and motor function. Future studies should continue to include this population in this area of research to discover an effective approach to help prevent or delay the onset of cognitive decline and to retain motor function and balance capabilities.

In conclusion, consuming 2 cups of fruit for 8 wk showed limited effectiveness to impact cardiovascular biomarkers, body composition, cognition, motor function, or balance. The lack of major beneficial findings alludes to more specific questions about fruit consumption that should inspire future research to investigate similar outcomes using a multifruit intervention. The adequacy of the Dietary Guidelines for Americans for consuming 2C of fruit daily has not been validated in the current study. Future research should continue to investigate the effects of whole fruit on vascular function, cognition, and motor function under more tightly controlled conditions and with more sensitive measures to determine if more fruit can serve as a more effective approach to improve the health of Americans. With further exploration, the USDA/United States Department of Health and Human Services will be able to better establish a more informed recommended daily intake for fruit.

Author contributions

The authors’ responsibilities were as follows – SH, MK, MYH, CL: designed the research; SZ, JM, AHS, CM, MK, SH: collected the data; SZ, NJ, JM, MYH, MK, SH: analyzed data; SZ, JM, NJ, MK, SH: interpreted the findings; SZ, JM, MK, MYH, CL, SH: wrote and edited the paper; SH: primary responsibility for final content; and all authors: read and approved the final manuscript.

Data availability

Data described in the manuscript will be made available on request and pending approval.

Funding

This work is supported by the Agricultural and Food Research Initiative grant no. 2021-67017-34024 accession no. 1025332 from the USDA National Institute of Food and Agriculture.

Conflict of interest

The authors report no conflicts of interest.

Appendix A. Supplementary data

The following is the Supplementary data to this article: