Chronic constipation affects about 14% of the population and has a substantial socioeconomic burden [1]. The symptoms of constipation include infrequent bowel movements, excessive straining, hard feces, and unsatisfactory defecation experience, with or without abdominal pain [1,2].

The 2 main constipation subtypes are irritable bowel syndrome with constipation (IBS-C) and functional constipation (FC). The key difference between the 2 is the presence of abdominal pain for patients with IBS-C [3]. In addition to constipation symptoms, these conditions are present with abdominal discomfort, bloating, reflux, indigestion, nausea, and vomiting [4]. They are frequently associated with psychological comorbidity, including mood disorders and anxiety, impacting quality of life [2,5,6]. Additionally, many individuals report extra-intestinal symptoms such as somatic pain, fatigue, and fibromyalgia [7-10].

Maintaining healthy dietary habits is essential for managing constipation. International guidelines [11-13] recommend drinking plenty of fluids and incorporating high-fiber foods (eg, fruits, vegetables, whole-grain bread, and cereals) into the diet as a first-line treatment for constipation. Moreover, many individuals prefer to self-manage their constipation symptoms with dietary management as opposed to pharmacological drugs [14].

Green kiwifruit (Actinidia deliciosa “Hayward”) is known for its laxative effects [15]. Several interventional studies of fresh fruit have reported improvements in bowel movement frequency, consistency, and ease of defecation [16-22]. Three studies involving participants with constipation have reported a significant improvement in complete spontaneous bowel movements [17,20,21]. Two recent studies demonstrated improved abdominal pain in people with constipation [20,21]. Additionally, some short-term studies have shown that consuming 2 servings of green kiwifruit can improve sleep quality, lipid profile, blood pressure, and platelet aggregation [23-25].

Green kiwifruit is rich in dietary fiber (2%‐3%), which is primarily derived from plant cell walls. The fiber has unique hydration properties, such as water retention, swelling, and viscosity [15,26-28]. A recent magnetic resonance imaging (MRI) study by Wilkinson-Smith et al [29], 2019, found that eating 2 servings of green kiwifruit twice a day for 3 days increased the water content of both the small bowel and the chyme in the ascending colon and raised colonic volume. These results support earlier in vitro research, which suggested that the polysaccharides in kiwifruit cell walls have a significant capacity for swelling and water retention. This indicates that kiwifruit may exert both osmotic and bulking effects in the colon [28,30].

Additionally, in vitro and animal studies have shown that the proteolytic enzyme actinidin from green kiwifruit aids gastric and small intestinal protein digestion [31,32], gastric emptying [33,34], and stimulates gut motility [35]. Only 1 human study has been published on the impact of green kiwifruit on the gut microbiome [36]. The study found that healthy individuals who consumed freeze-dried green kiwifruit (equivalent to 2 kiwifruit servings) for 4 days experienced a favorable shift in the abundance of lactic acid bacteria, such as lactobacilli and bifidobacteria, in fecal matter [36]. Moreover, there was a trend towards a decreased abundance of Clostridium and Bacteroides genera [36].

There is promising evidence from human, animal, and in vitro studies regarding the laxative effect of green kiwifruit. However, data on the effects of sustained consumption of 2 green kiwifruit daily on abdominal pain and the possible mechanisms of action in individuals with constipation are limited. This underscores the need for a comprehensive study to evaluate both the effectiveness and the underlying mechanisms involved.

This study was based on the hypothesis that consuming 2 green kiwifruits daily over 4 weeks would improve abdominal pain.

Secondary hypotheses were that constipation and other associated symptoms improved. Additionally, the symptom improvement is due to improved physiological functions such as changes in colonic volume and water content, gut motility, gas fermentation profile, and biological functions such as the gut microbiome composition and the host-microbial metabolite profile.

The primary aim of this study was to assess the impact of consuming 2 green kiwifruits daily over 4 weeks on abdominal pain in participants with either FC or IBS-C.

The secondary aims were to determine the impact of consuming green kiwifruit daily over four weeks on (1) constipation and its associated symptoms using validated questionnaires; (2) bowel habits using a daily bowel movement diary; (3) mental health and general well-being parameters using validated questionnaires; (4) dietary fiber intake using validated questionnaires; (5) colonic volume and water content of colonic chyme using a validated MRI protocol; (6) whole gut transit time using blue dye; (7) a relative abundance of individual taxa, predictive function (gene abundances), and diversity indices of the fecal gut microbiota using shotgun metagenomic sequencing, fecal and plasma metabolites using liquid chromatography-mass spectrometry (LCMS); and (8) regional and whole gut transit time and gas fermentation profiles using Smartpill (Medtronic Inc) and Atmo gas-sensing capsules (Atmo Biosciences; Atmo capsule) in a subset of participants.

This study was a 4-week, 2-arm, parallel, open-label, negative-controlled, randomized design investigating the daily impact of 2 Zespri green kiwifruit compared to calorie-matched maltodextrin on abdominal pain in 60 adults with IBS-C or FC. Thirty of these participants were invited to participate in the wireless motility devices (WMDs) substudy.

This study’s duration was 9 weeks: 21 days (3 weeks) lead-in phase, 28 days (4 weeks) intervention phase, and 14 days (2 weeks) follow-up phase (Figure 1). The lead-in phase started with participant enrollment and finished a week after the baseline visit. The week between the baseline visit and the intervention phase was designed to give sufficient time for the WMDs to be eliminated from the body [37]. The lead-in phase was followed by the intervention phase, where participants were either randomized to kiwifruit or maltodextrin. This was followed by a 2-week follow-up phase. This study was conducted according to the CONSORT (Consolidated Standards of Reporting Trials) and the SPIRIT (Standard Protocol Items: Recommendations for Interventional Trials) guidelines and adhered to the superiority study framework [38,39].

This study was approved by the Health and Disability Ethics Committee, New Zealand (NZ/21/NTB/96) May 14, 2021 and registered with the Australian New Zealand Clinical Trial Registry (ACTRN12621000621819) on May 24, 2021. The University of Otago Māori Health Advancement review process endorsed this study. Before entering this study, eligible participants provided informed consent, and additional informed consent was obtained for the WMDs substudy.

The researchers involved in data collection were experienced in maintaining the security and confidentiality of health information. This study did not report any identifying information about participants, such as names, birth dates, or images. Identifying details were collected during enrollment and accessed through REDCap (Research Electronic Data Capture; Vanderbilt University) only by the research team.

Participants received supermarket vouchers worth NZD $300 (US $185) as a thank you for completing the study. Additionally, the participants selected for the substudy received NZD $100 (US $62) to compensate for their time and any inconvenience.

Participant recruitment started in May 2021 and ended in May 2022 in Christchurch, New Zealand. Due to slow recruitment at this study’s start, limited availability of MRI assessments, and a COVID-19 pandemic lockdown (August 2021-September 2021), study recruitment occurred over 2 kiwifruit growing seasons. Full results are expected to be published in peer-reviewed journals by the April 2026.

Eligibility parameters were collected in 2 steps: participants completed a prescreening online questionnaire and an in-person screening visit. The prescreening questionnaire included questions based on inclusion and exclusion criteria, including questions related to the Rome IV diagnostic criteria (Textbox 2). Based on these replies, participants were diagnosed with either FC or IBS-C. Participants who met all study criteria were invited for a personal screening visit to provide consent and a blood sample to exclude any significant abnormalities.

Upon enrolling in this study, each participant was sent an individualized REDCap survey link hosted by the University of Otago, New Zealand, to complete 2 online enrollment questionnaires. One of the questionnaires was the Modified Hunter New England Health Survey [40], which includes the validated 12-item Short-Form Health Survey Version 2 (SF-12v2) and selected question domains from the New South Wales Population Health Survey.

The SF-12v2 questionnaire was used to evaluate the general health status of the participants, measuring eight domains of functioning and well-being, including physical functioning, role limitations due to physical problems, bodily pain, general health perception, energy and vitality, social functioning, role limitation due to emotional problems, and mental health well-being of the participants. The SF-12v2 questionnaire was scored by using published algorithms [41] to produce the physical component summary-12 and mental component summary-12. Participants also completed the Economic Living Standard Index short form. This validated questionnaire assesses the standard of living and socioeconomic background, ranging from “very good” to “severe hardship” [42].

Participants were randomly assigned to consume Zespri 2 green kiwifruit (Actinidia deliciosa cultivar “Hayward,” ~150 g per serving, ~90 kcal) per day or the calorie-matched negative control (Maltodextrin, 25 g per dose, ~90 kcal) for 4 weeks. The participants were instructed to consume 2 kiwifruit without skin or maltodextrin (available as powder) in 1 daily dose by incorporating it into their morning meal or breakfast. The intervention provided in this study was in addition to their usual diet.

Participants were provided with a minimum of a 14-day supply of kiwifruit on the day of the baseline visit and had their supply replenished for the subsequent 14 days. A few additional days’ supply was provided to the participants to ensure good-quality fruit was available, given that fresh fruit is a perishable product. Providing a 14-day supply was considered reasonable, as the shelf life of kiwifruit is typically around 16‐20 days at room temperature.

The kiwifruit was supplied by Zespri International Limited for the duration of this study.

The negative control was commercially sourced digestible maltodextrin (Davis Food Ingredients). This low-sweet polymer of D-dextrin was selected as a placebo as it can easily match the calorie content of the kiwifruit. Maltodextrin is also completely digested into glucose and actively absorbed in the small intestine [43], which does not increase small- or large-intestinal water content [29]. Participants receiving maltodextrin were asked not to consume kiwifruit during this study’s period.

Following enrollment, participants were randomly allocated to the kiwifruit or maltodextrin group using randomized permuted blocks (block size four) drawn from sealed opaque envelopes. Due to the nature of the intervention, that is, consumption of whole fruit, blinding the participant and the administering researcher was not possible. The researchers involved in the analysis were blinded to the group allocation.

Figure 2summarizes the timing of measures. Web-based questionnaires were administered to participants at different stages of this study, including screening, enrollment, baseline, postintervention, and follow-up assessments. Individualized REDCap survey links were sent to participants to complete questionnaires. A paper copy of the questionnaires was available on request.

Participants underwent an MRI assessment, ingested blue food dye, and provided blood and fecal samples at baseline and postintervention. In addition, participants completed a three-day food diary at baseline and postintervention. They also completed a follow-up visit 2 weeks postintervention, providing blood and fecal samples. Participants recorded daily bowel movement diaries using a REDCap-based app from enrollment until the follow-up period. A subset of participants also ingested WMDs at baseline and postintervention.

This study’s participants were given a cooler bag that contained a fecal collection kit and an ice pack. They were instructed to collect 2 fecal samples in the provided 60 mL containers (2 containers were provided) ≤24 hours before clinic visits at baseline, postintervention, and follow-up. One sample was stored at 4 °C in their home refrigerator and the other in their freezer (−18 °C) and transported within 24 hours in the provided cooler bag.

Extra fecal collection kits were offered to participants with less frequent bowel movements (occurring once in 3 days or less). They were encouraged to collect fecal samples 48 to 72 hours before and up to 24 hours after the clinic visit. Participants were also instructed to label the container with the collected time and date and record it in the bowel movement diary app. The collected samples were kept chilled until aliquoted in the laboratory. The home-frozen fecal sample was stored at −80 °C until the meta-transcriptomic analysis. The chilled fecal sample was aliquoted (1 g each) for analyses of microbial composition, microbial gene abundances, metabolome, bile acid, organic acid, and immune marker analyses. All aliquots were snap-frozen in liquid nitrogen before storage at −80 °C until analysis.

The participants also provided a blood sample (described below) at the in-person screening visit, baseline, postintervention, and follow-up. At the screening visit, blood samples were collected for complete blood count in a 4 mL ethylenediaminetetraacetic acid (EDTA) vacutainer and an 8 mL heparin tube for standard biochemistry analyses (liver function, renal function, electrolytes, C-reactive protein, and blood glucose) at Canterbury Health Laboratory (Christchurch, New Zealand). These analyses were repeated at follow-up visits before participants were discharged from this study.

At baseline and postintervention, blood samples were collected in one 4 mL EDTA tube, three 9 mL EDTA tubes, and one 6 mL heparin tube. The 4 mL EDTA and 6 mL heparin tubes were centrifuged for 5 minutes at 2000 x g with full acceleration and no deceleration. The plasma was aliquoted (4×500 µL) into labeled Eppendorf tubes and stored at −80 °C until metabolome, targeted metabolite, and immune profile analyses were performed. Peripheral blood mononuclear cells (PBMCs) were isolated from approximately 30 mL of whole blood (3×9 mL EDTA vacutainers) using the SepMate technique [44]. Briefly, the blood was split into two 15 mL centrifuge tubes and diluted in a 1:2 v/v ratio by adding 15 mL of Dulbecco’s phosphate-buffered saline containing 2% heat-inactivated fetal bovine serum to each tube. Next, the diluted blood was transferred into two SepMate tubes containing 15 mL of Lymphoprep. The SepMate tubes were centrifuged at 1200 x g for 10 minutes. After centrifugation, the top layer was poured into a new tube. This tube was centrifuged at 700 x g for 8 minutes, and the supernatant was removed. The pelleted PBMC was washed with 10 mL of Dulbecco’s phosphate-buffered saline containing 2% heat-inactivated fetal bovine serum and centrifuged again at 300 x g for 8 minutes, aspirating the supernatant while leaving the pelleted PBMC intact. Finally, the PBMC pellet was resuspended in 3 mL of a freezing mixture of 90% heat-inactivated fetal bovine serum and 10% (v/v) dimethyl sulfoxide and aliquoted into three cryovials. The cryovials were placed into a CoolCell LX freezer container and stored at −80 °C overnight. The following day, all the frozen cryovials were transferred to liquid nitrogen storage at −80 °C.

The self-administered Gastrointestinal Symptom Rating Scale (GSRS, AstraZeneca) [45], a 15-item questionnaire, assessed common gastrointestinal symptoms, including abdominal pain, constipation, diarrhea, reflux, and indigestion. The abdominal pain domain of GSRS was used to determine the primary outcome of abdominal pain and discomfort relief. The GSRS is a 7-point Likert-type scale, where 1 represents the absence of symptoms, and 7 represents severe discomfort. The scale can be used as a whole, or the mean value for items under each symptom domain can be assessed to give a subscale [45]. The mean change in symptom scores between baseline and postintervention measurements was compared to evaluate the effectiveness of the intervention.

During the main study, participants were invited to ingest a Smartpill and an Atmo capsule. The dual WMD ingestion protocol was similar to a previously described protocol [78]. Participants consumed a standardized cereal bar (Smart bar, Medtronic Inc) followed by the WMD with water. They fasted for 6 hours postingestion and wore transponders across their body (or close to their body at night) until WMD excretion. The participants were asked to enter additional bowel movement data on the Smartpill and Atmo capsule transponders for the duration of the substudy. Participants with gut transit times exceeding the capsule’s or transponder’s battery life (~5 d) were requested to visualize their feces to confirm the excretion of the capsule. If the excretion of one or both capsules could not be determined from the transponder data or visual confirmation, an abdominal radiological examination was conducted.

Smartpill measures anatomical landmarks and regional transit time via intraluminal temperature, pH, and pressure changes as it passes through the gastrointestinal tract [79]. The novel Atmo capsule measures anatomical landmarks and colonic fermentation by measuring oxygen, hydrogen, carbon dioxide, volatile organic compounds, capsule orientation, environmental electromagnetic properties, and temperature changes [78]. The analyses included regional transit time and whole-gut transit time measured by SmartPill and Atmo capsule, and colonic gas (hydrogen and carbon dioxide) profiles.

Smartpill transponder data were downloaded and analyzed using the proprietary software MotiliGI (version 2.5, Medtronic Inc). Atmo capsule transponder data was collected in real time on the smartphone app and transmitted to the Atmo cloud (hosted by Atmo Bioscience) for review and analysis.

All participants were informed about potential side effects. They were encouraged to record any symptoms and contact their health care provider if needed.

In this study, compliance was not formally measured. Participants received small batches of the intervention throughout this study’s period and were encouraged to continue consuming the interventional product. The food diaries collected during the postintervention visit provided insights into the consumption of kiwifruit and maltodextrin.

This study was powered to detect an effect size of 1 within the abdominal pain domain of the GSRS, 5% significance (α), and 80% power (β)=16 participants regarding a difference between the groups based on GSRS data obtained from studies with gold kiwifruit [80-82]. In addition to the primary outcome, this study was intentionally powered to ensure the detection of differences between groups in colonic volume as measured by MRI as a secondary outcome. Based on a previous study with 300 g of kiwifruit, a parallel design with 5% significance (α) and 90% power (β), the minimum sample size was determined as 22 participants per group [29]. To account for the lower daily kiwifruit intake (150 g) over 4 weeks and an estimated dropout rate of 10%, a total of 30 participants needed to be recruited for each group.

This study was conducted as a superiority trial. The Guidelines of the Committee for Proprietary Medicinal Products (now termed Committee for Medicinal Products for Human Use) require intention to treat analysis [83]. All data from randomized participants who received at least one intervention dose and had at least one postintervention measurement were analyzed. All statistical analyses were performed using SPSS software (version 29.0; IBM Corp) or R software (version 4.3.1; R Core Team 2021) by blinded researchers under the guidance of an independent biostatistician.

Baseline characteristics were presented using descriptive statistics. The mean and SD, or median and range, were used to describe continuous variables. Frequencies and percentages were used to describe categorical variables. The group effects (ie, the mean change from baseline between groups) were assessed using (parametric) t tests and (nonparametric) Mann-Whitney or Kruskal-Wallis tests for symmetrically and asymmetrically distributed data. The mean, SD, and CIs were used to describe continuous variables. Categorical variables were assessed using chi-square tests (or Fisher exact tests for small samples). A 2-tailed statistical significance P<.05 showed statistical significance. A P value of <.1 would be considered a trend for most variables.

To analyze the fecal microbiome measures, both univariate and multivariate statistical methods were used to evaluate variations in alpha and beta diversity and differences in the relative abundance of taxa and gene abundance differences between groups. Differences in alpha diversity were assessed using the nonparametric Wilcoxon signed rank test with continuity correction, comparing baseline and postintervention data, as well as between groups postintervention. The Benjamini-Hochberg method was applied to control for the FDR associated with multiple hypothesis testing for variables in microbiome analyses. Adjusted P values were calculated to mitigate the inflated type I error rates arising from numerous comparisons. A P<.05 and FDR of P<.1 were considered significant, with a probable biologically significant difference (trend) at unadjusted P≤.1 indicating a trend.

Boxplots were created using ggplot2 to compare groups for each α diversity (Chao1 and Shannon indices) across time points. The statistical significance of microbial community composition was assessed using the permutational multivariate ANOVA and the Analysis of Dissimilarity (number of permutations: 999), which uses permutation testing and F tests to evaluate differences in beta diversity. Differential in microbial gene expression was analyzed using quasi-likelihood F tests, controlling for dispersion uncertainty. Genes with an absolute log-fold change greater than 1.1 were deemed differentially expressed.

As mentioned above, for plasma or fecal metabolome measures, the comparisons were between each group and baseline and between groups postintervention. The partial least squares projection-discriminant analysis (PLS-DA) model was performed to identify metabolites using SIMCA, version 17 (Umetrics). The PLS-DA models were subject to 100-fold permutations to evaluate performances and validated using the predictive ability of the model (Q2) and the ANOVA testing of cross-validated predictive residuals [84]. The most discriminating features were selected for building PLS-DA models. These models were again subjected to prediction accuracy and overfitting assessment by the corresponding tests (cross-validated predictive residuals) and comparison of the Q2 afterward. Features were selected based on their variable importance in the projection score. Univariate statistical analysis and heatmaps were performed using Metaboanalyst 5.0 [85]. The metabolite and lipid relative intensity differences between groups were tested using the t test or the Wilcoxon rank-sum test. The multiple testing corrections were controlled using FDR correction [86].

Several studies have demonstrated that the consumption of green kiwifruit improves laxation. However, there are few robust randomized controlled trials assessing the efficacy and mechanism of green kiwifruit ingestion for people with constipation. To our knowledge, this is the first study to evaluate the effectiveness of consuming 2 green kiwifruit daily for 4 weeks on abdominal pain as a primary end point and on colonic volume and chyme water content in individuals with constipation. In addition to physiome outcomes, the fecal microbiota composition and gene abundance, plasma and fecal metabolites, and plasma immune markers were also analyzed to understand the underlying mechanism.

Improvement in pain was chosen as a primary outcome to assess whether kiwifruit offers benefits beyond laxation. Specifically, the goal was to evaluate kiwifruit’s suitability as a natural dietary intervention for individuals who experience both pain and constipation. Several dietary fiber interventions, including wheat bran, psyllium, and inulin, are known to alleviate constipation but are not preferred by patients because they may exacerbate pain [20,87].

In addition, this study included a range of patient-reported outcomes to evaluate nongastrointestinal symptoms and mental well-being, aiming to determine whether kiwifruit positively impacts health beyond its nutritional composition. A recent study showed that habitual kiwifruit consumption improved the quality of life in people with constipation, supporting that kiwifruit provides benefits that extend beyond mere nutrition [21].

This study used an MRI protocol to measure colonic volume and water content after habitual kiwifruit consumption over 4 weeks. This approach replicates a more natural setting where kiwifruit is integrated into the regular diet of individuals with constipation. A previous MRI study used serial imaging to examine the effects of a high dose of kiwifruit in healthy adults, providing valuable insights into the potential mechanism [29]. Additionally, this study also takes into account the secondary outcomes of colonic volume and water content when calculating the sample size, aiming to optimize its clinical utility and reduce the chances of false positive results [88].

This study was an open-label design. Both participants and researchers were aware of the group assignments for kiwifruit and maltodextrin. Open-label studies are known to have potential biases in reporting and analysis, so several measures were implemented to minimize these risks [89]. Various questionnaires were used to reduce these risks by asking the same questions in different forms and styles. Additionally, only two researchers were informed about the group allocations, and multiple endpoints were assessed, including both subjective and objective outcome measures. Participants completed the subjective end points independently, without the presence of a researcher. Additionally, a blinded outcome analysis was conducted, and the trial’s protocol and statistical plan were finalized before this study commenced [90].

The inclusion criteria could be expanded to include individuals aged 65 years and older. This is a significant group worldwide [91,92]. Older adults experience changes in their gastrointestinal function, which alter taste, motility, hormone secretion, and anorexia [93], resulting in their exclusion from many studies. However, it is a growing population with a known increased prevalence of bowel irregularity. Future research could assess the effectiveness and the change in gastrointestinal function and fecal microbiota parameters following high-fiber food interventions. Such dietary interventions could be particularly beneficial for those experiencing bowel irregularities due to polypharmacy, defined as taking more than 5 medications [94].

This study was the first to investigate the effectiveness of habitual consumption of kiwifruit (2 servings daily for 4 weeks) on abdominal pain, colonic volume, and colonic chyme water content. This study also assessed patient-reported outcomes and objective regional physiological, biochemical, and microbial changes that may aid digestion, absorption, and fermentation in order to understand the mechanism of action.