• Open Access

Fruit consumption among people living in a high deprivation New Zealand neighbourhood


Correspondence to:
Dr Sara R. Jaeger, Plant & Food Research, Mt Albert, Private Bag 92169, Auckland 1142, New Zealand. Fax: +64 9 925 7001; e-mail: sara.jaeger@plantandfood.co.nz


Objective: To investigate fruit consumption in a high deprivation population in New Zealand.

Method: In 2007, 99 door-to-door interviews were conducted in a high deprivation neighbourhood in Auckland with a focus on measuring self-reported fruit consumption.

Results: On average, participants reported eating a serving of fresh fruit five to six times per week. At the time of the interview, more than a third of participants (38%) did not have any fresh fruit in the house and 60% of respondents reported that in the past month they had thrown out fruit between one to four times per week because it was considered to be past its best in terms of eating quality. Fruit juice was consumed on average one to two times per week. Self-efficacy for fruit consumption was positively associated with consumption. Relative to participants with lower levels of self-efficacy for fruit consumption, those with higher levels of self-efficacy were more likely to achieve the target of consuming two or more servings of fruit daily.

Conclusion: Strategies that aim to increase self-efficacy beliefs for fruit consumption may contribute to improving compliance with the recommended two or more servings daily. Together with strategies that give consideration to the social and cultural context and community level interventions (involving schools, churches and local community groups) they represent a holistic approach that is likely to be necessary for improving fruit consumption in high deprivation populations.

The benefits of fruit consumption for health are well-established,1–5 and public health promotions and interventions to increase fruit consumption are used in many countries, including New Zealand (NZ).6–8 Recent NZ data show that only half NZ adults consume two or more servings of fruit each day.9 Variability in intake exists across sub-populations. Older females generally have higher intakes.10,11 Low income and neighbourhood deprivation is associated with lower fruit consumption.12–15

Deprivation is a broader concept than low income, encompassing aspects such as employment, living space and support. Several of these characteristics have also been identified as correlates of barriers to fruit consumption.13,14 A review of available literature suggests that there are no previous reports exploring fruit consumption in high deprivation populations in New Zealand.

As a moderator of fruit consumption, the study included a measure of self-efficacy. Self-efficacy pertains to a person's belief of whether they are confident that they can undertake the desired behaviour when faced with specific barriers.16 These barriers comprise of more superficial factors such as when alone, when in a hurry or during weekends, as opposed broader social, cultural and financial barriers. This measure is intended to provide an indication of a person's willingness to change behaviour in the face of such challenges. For example, people with high self-efficacy beliefs for a focal behaviour will likely find it easier to make or maintain behavioural change,17,18 for example achieving a target of consuming two or more servings of fruit per day. Previous research has found that self-efficacy for fruit and vegetable consumption is both strongly and consistently associated with higher intakes both within the general population and within low income populations.12,19–21 Again, a review of available literature suggests that this construct does not appear to have been studied in connection with deprivation.


Sample and field procedures

A suitable high deprivation population was identified using the 2001 Index of Deprivation in New Zealand NZDep.22 The Otara suburb in Auckland was chosen and 10 mesh blocks (a geographical area with approximately 90 people and 40 dwellings, which in urban areas is typically the size of a street block)23 were identified as the locations from which to collect data. Across the chosen 10 mesh blocks the average deprivation score was 1,375. The 2001 NZ average was 1,000 and across the country deprivation scores ranged from 834–1,521. Neighbourhood deprivation generally does not change quickly from census to census22 and for the Otara suburb this was confirmed by the 2006 Index of Deprivation in New Zealand,24 which became available following completion of the fieldwork. The average NZ deprivation score in 2006 was 999 (range 838–1,619). Although there had been some movement in the chosen 10 mesh blocks, they all continued to be classified as high deprivation neighbourhoods and across the 10 mesh blocks the average deprivation score was 1,345.

Data were collected through a door-to-door survey in June 2007 (n=99). The interviews were conducted by professional interviewers from a commercial market research provider. The interviewers, who were from multi-ethnic backgrounds, had previous experiences working with this population. Prior to data collection, interviewers were briefed to ensure consistency in delivery of the questionnaire. Issues relating to cultural sensitivity and appropriate conduct by the interviewers were also discussed. In addition to a good level of spoken English, criteria required participants to be between 18 and 65 years old and enforced a 50:50 gender split. Within a mesh block, interviewers approached every second house and interviewers discussed the participant information sheet with potential participants and if communication was not an issue and the participant offered verbal consent, the interview proceeded. This was completed at a location comfortable for the participant (i.e. in the house or the front yard). Each interview lasted approximately 30–45 minutes. The interview process was contracted out to a professional marketing agency and refusals were not recorded for which reason an accurate response rate is not available. However, the 99 interviews were completed over two weekends and two weekdays by seven qualified interviewers. The interviewers reported that people not being home was the most substantial limitation they faced. Sunday is a popular church day in this community so two weekdays were added in an attempt to access people when they were home. There was some difficulty in getting males to agree to participate compared to females. In return for their time, participants received a grocery shopping voucher. A copy of the interview guideline is available upon request.


The questionnaire comprised multiple sections. Of relevance to this paper, two sections centred around fruit consumption including frequency of consumption, and perceived ability to consume the recommended number of servings of fruit per day. Intake was measured as the frequency of having eaten one serving of fruit in the past month using a 10-point scale: 1=‘never’; 2=‘1–3 times per month’, 3=‘1–2 times per week’, 4=‘3–4 times per week’, 5=‘5–6 times per week’, 6=‘1 time per day’, 7=‘2 times per day’, 8=‘3 times per day’, 9=‘4 times per day’, and 10=‘5 or more times per day’.25 Questions were accompanied by detailed examples and photographs of what constitutes a serving of both fruit and juice (e.g. a small banana, a ¼ cup dried fruit or 250 ml of 100% fruit juice) which was explained to each participant during the interview. Questions on self-efficacy in regard to the participants’ perceived ability to find alternative uses for fruit considered to be past its best were also included in these two sections, and self-efficacy expectations for fruit intake were measured using the scale by Brug et al.16 This asked participants to rate their degree of certainty (1 =‘not very certain’, 5 =‘very certain’) that they could eat at least two servings of fruit per day in a number of difficult situations (e.g. when eating alone, during winter when there is less choice).

The final sections of the questionnaire collected measures on deprivation and general demographic information. Specifically, to gain insight into deprivation at the individual level, the NZ-specific NZiDep index was administered.26 The index, which comprises eight items (e.g. feeling cold to save heating costs; putting up with hole in shoes because they could not afford a replacement) is designed to be applicable to all ethnic groups in NZ. It differentiates individuals with no deprivation characteristics (NZiDep=1), one deprivation characteristic (NZiDep=2) and multiple deprivation characteristics (NZiDep≥3). The questionnaire concluded with questions on ethnicity, household composition, income and housing cost. The two latter questions were asked specifically with reference to participants’ Economic Family Units (EFU). The concept of EFU24 acknowledges that households may have more than one independent economic grouping. For example, a household of three adults and three children may comprise two EFUs, one that is a couple with three dependent children, and one that is a single adult (e.g. elderly parent or grown-up child).

Statistical methods

A combination of descriptive analyses, correlation analyses, chi-square analysis, and analysis of variance were used. The five-item scale measuring self-efficacy for fruit consumption was internally reliable (Cronbach α=0.77) and exploratory factor analysis indicated that it represented a single underlying construct. A summary measure was input to analysis (theoretical range: 5–25).


Sample characteristics

The sample comprised Māori (23%) and Pacific people (of which 42% were Samoan, 26% were Tongan, 20% were Cook Island Māori and 12% were Niuean). The sample was evenly split on gender and ranged in ages from 18 to 65 years old with a predominance of younger people (77% aged 18–44 years old). A majority of participants (68%) reported multiple deprivation characteristics (NZiDep ≥ 3) and only 15% did not report any deprivation characteristics (NZiDep=1). More than half the participants (57%) belonged to an EFU that comprised two adults with dependent child(ren). Participant EFUs without dependent children were in the minority (20%). Normalised to weekly figures, EFU incomes varied between $143 and $1,250, and were, on average, $504 (SD=$252). Weekly EFU incomes varied with NZiDep classification (F2,77=9.0; P=0.0003) and were higher among participants who did not report any deprivation characteristics ($725) than those who reported multiple deprivation characteristics ($433). Weekly housing costs displayed a similar, albeit weaker pattern (F2,81=2.8; P=0.07). For the sake of completeness Table 1 provides a breakdown of sample characteristics by ethnic group.

Table 1.  Descriptive summary of participant characteristics.
 Maori N=23Pacific Island N=76
No Deprivation Levels26.111.8
Singly Deprived21.715.8
Multiply Deprived52.272.4
Solo parent with dependent children30.421.1
Two-parent family with dependent children47.859.2
Single person13.013.2
Couple only8.76.6
Income/week$539.70 (SD $246.16)$494.38 (SD $255.01)
Household cost/week$181.50 (SD $116.01)$173.96 (SD $99.40)

Summary statistics for fruit consumption

Reported results are based on the aggregate sample (Table 2). Stated consumption of one serving of fruit (fresh, as well as dried and cut-up) in the past month varied significantly among participants, with a few participants (6%) reporting having eaten 5+ servings daily and one person reporting not having eaten fruit at all. On average, participants consumed a serving of fruit 5–6 times per week (mean = 5.3, SD= 2.2). Around 10% did not eat fruit on a weekly basis. At the time of the interview, more than a third of participants (38%) did not have any fresh fruit in the house. Reported frequency of consumption of a serving of fruit (fresh, as well as dried and cut-up) did not differ from frequency of consumption of a serving of fresh fruit (t98=0.84, P=0.40). From this, we take the inference that participants’ fruit intake came almost exclusively from fresh fruit. On average participants reported drinking a glass (250 mL) of 100% pure fruit juice 1–2 times per week (mean = 3.3, SD= 2.4)

Table 2.  Summary of frequency of fruit consumption across total sample and sub-populations.
Frequency of ConsumptionA serving of fruita (%)A serving of juiceb (%)Low self-efficacyc serving of fruit (%)High self-efficacyc serving of fruit(%)
  1. Notes:

  2. a) One serving is: i) a medium–size piece of fresh fruit (e.g., a medium–size apple or a small banana), ii) a ¼ cup dried fruit or iii) a ½ cup cut–up fruit.

  3. b) A small glass (6 oz or 180ml) of 100% fruit juice (either freshly squeezed or from concentrate) such as orange, apple, grape or grapefruit.

  4. c) Sample size segment is 34.

1–3 times per month10.
1–2 times per week12.
3–4 times per week20.
5–6 times per week11.16.12.920.6
1 time per day14.25.08.817.7
2 times per day18.26.117.729.4
3 times per day4.06.105.9
4 times per day3.
5 or more times per day6.

Compliance with recommended fruit consumption guidelines

New Zealand guidelines recommend that people eat at least two daily servings of fruit.27 To assess the extent participants achieved this target, the sample was divided into two groups defined by whether or not participants ate 2+ daily servings of fresh fruit. Just over one quarter of participants (26.3%) consumed 2+ servings of fresh fruit. The data also revealed that 60% of respondents they had thrown fruit out between 1–4 times per week in the past month because it was considered to be past its best in terms of eating quality.

Female participants were more likely to consume 2+ daily servings of fresh fruit than male participants (χ2df=1=7.2; p=0.007). Age was not a predictor of segment membership (χ2df=2=0.3; p=0.57). There was a tendency for participants who ate 2+ daily servings of fresh fruit to have dependent child(ren) in their household (χ2df=1=2.7; p=0.10). Whether or not participants belonged to the group that ate 2+ daily servings of fresh fruit was not associated with the number of deprivation characteristics participants reported (χ2df=2=0.7; p=0.72). Neither were there significant associations with participants’ EFU income or housing cost (p>0.15).

Role of self-efficacy

Across the sample, self-efficacy for consuming 2+ daily servings of fruit varied significantly, with some participants at either pole of the summary index (5 and 25, respectively). The average self-efficacy score was 15.2 (SD=5.4). A triadic split procedure identified participants who were low (35%; mean=9.4), medium (30%, mean=14.6) and high (35%; mean=21.4) in self-efficacy.

Self-efficacy for consumption of 2+ daily servings of fresh fruit was significantly associated with frequency of fruit consumption (r96=0.29; p=0.04). People classified as high in self efficacy reported eating a serving of fruit more than once daily. On the 10-point measurement scale for frequency of eating a serving of fruit (1=‘never’; 10=‘5+ times daily’), this group's mean score was 6.3 (SD= 1.8). Participants who ate 2+ daily servings of fresh fruit were, on average, higher in self-efficacy than those who did not (t29.7= -1.8; p=0.087). Consumption frequency among those classified as low in self-efficacy was, on average, 4–5 times per week. This group's average was 4.4 (SD= 2.1). In the ‘non compliant’ segment, 43% of people were classified as being low in self-efficacy. Conversely, 65% of people in the 2+ daily servings ‘compliant’ segment were classified as being high in self-efficacy (χ2df=2=15.7; p=0.0004). Table 2 offers further results for fruit consumption and self-efficacy.

There was a tendency for female participants to be higher in self-efficacy (mean=16.2) for fruit consumption than male participants (mean=14.2) (t96=-1.80; p=0.074), but in this sample, self-efficacy was not associated with age (t96=-0.35; p=0.73). Participants level of self-efficacy for fruit consumption did not depend on whether or not they had dependent child(ren) in their household (t96=-0.08; p=0.94). Participants’ deprivation status was not related to self-efficacy (F2,95=0.01; p=0.99) and the association between self-efficacy and participants’ EFU income was not significant (p=0.57).


Increasing intake of fruit

New Zealand guidelines recommend that people eat at least two servings of fruit per day.27 In light of the well-documented relationship between fruit consumption and health,1–5 it is of interest to consider possible avenues and potential barriers to promoting fruit consumption in high deprivation populations.

The barriers to fruit intake are many and include cost, availability and accessibility. The latter may attain additional significance in high deprivation populations where access to motor vehicles is more limited than in the general population. For example, Dibsdall et al.28 documented how lack of access to a motor vehicle can influence food purchasing behaviour in low income populations and although such data were not collected in our sample, it is anticipated that several participants would have been in similar circumstances. Tentatively, this may also be a reason why 38% of participants did not have fresh fruit in the house at the time of the interview. Without a car, it may be difficult to purchase bulk amounts of fresh fruit. This seems particularly difficult given that on average five people lived in each household, meaning that an average of around 50 pieces of fruit may be required per week for all members of the household to meet recommended guidelines. This study was conducted in Otara, which is renowned for it local market that has fresh fruit at reasonable prices in abundance. The location of the market was easily accessible to most participants, yet the average level of intake was below national recommendations. It could therefore be possible that some residents found it difficult to transport a week's supply of fruit from the market to their homes. That said, it is important to recognise that much broader barriers than accessibility are likely at play.

The finding that ∼60% of the respondents threw fruit out 1–4 times per week because it was considered to be past its best in terms of eating quality poses further difficulties for increasing fruit consumption. It is possible that the fruit that was available for purchase at a low cost in this neighbourhood was of a poor quality. This could then indicate that fruit is not being consumed due to poor taste or perhaps that it is deteriorating in quality more rapidly than it can be consumed resulting in disposal. Storage conditions may be an additional factor contributing to more rapid deterioration of fruit quality. Preliminary results of a separate study indicate that throwing out fruit is relatively common in the general population in New Zealand.29 It is therefore possible that the difficulty experienced in handling highly perishable items may be what is contributing to large amount of produce being thrown out. As a consequence it would be understandable that people may be reluctant to spend a portion of their food budget on items that will spoil quickly and therefore be wasted. As Cheer et al.30 have suggested, when money needs to be reallocated, people will tend to discount food because it is not a fixed expense. If a portion of the food budget must be traded-off in favour of surplus bills, for example, it would not be surprising that fruit may feature on the non-priority list. In instances of financial constraint it is reported that bulkier foods such as pasta and potatoes were more frequently chosen because they were more filling. With an average of five people per household it is not surprising that priority is given to purchasing food items that will feed several people and be more filling.30 Furthermore, the fact that healthy food is often considered to be expensive food counteracts the possibility of it rating as a priority purchase.30 Suggestions to improve consumption of fruit must therefore consider that trade-offs and reallocation of money for food items is a frequent requirement for people living in high deprivation areas.

The results indicate that there may be opportunities to increase awareness of the contribution to fruit intake through fruit juice, canned fruit and dried fruit. For example, consumption of 250 mL of pure (100%) fruit juice can count as one serving per day.27 This could be a relatively cost effective option because a 3 L bottle is equivalent to approximately 12 servings. It ought to be cautioned however, that fruit juice is not equivalent in nutritional properties so should not be considered a replacement for fruit consumption.27,31 Because one serving of canned fruit is equivalent to ½ cup while a serving of dried fruit is equivalent to ¼ cup this could also be worth promoting further if accessibility was a key barrier to consumption. However, although dried fruit is more compact than fresh fruit and has a longer shelf-life, it does have a much different taste and texture. If this is not enjoyed similarly to fresh fruit then it would seem irrational to suggest that this be given priority in an already restricted food budget. Suggestions for improving fruit consumption require interventions that promote nutritional health claims in the context of that social and cultural environment. Therefore interventions should change between localities and tailored to the needs/preferences of specific groups. We suggest that although time-consuming it would be time well spent for increasing the likelihood that strategies are successful.

High deprivation and other populations have, in recent times, become a target for studies aiming to explore health issues such as diabetes, high blood pressure and high cholesterol.32,33 Health issues associated with food consumption have resulted in a confounding number of health messages which at times can be fragmented and contradictory.34 Recommendations themselves often change as new research emerges. One recent example can be seen in transition from promoting general vegetable consumption to a more specific recommendation to consume more dark green and orange vegetables.35 As a result of such changing messages consumers in a variety of socio-economic situations may experience confusion with regard to the best food purchasing decisions. This confusion about what is good for you versus what is not may lead some consumers to rely on anecdotal reports for guidance as opposed to scientific evidence.34 Interventions must therefore not only carefully consider how health messages are portrayed in these communities but also take into account how these messages may be interpreted in the context of that specific social and cultural environment.

Increasing self-efficacy beliefs for fruit consumption

The significant association between self-efficacy and fruit consumption established in this study suggests that strategies aimed at developing strong self-efficacy beliefs may be worth incorporating into interventions aimed at increasing daily fruit consumption in high deprivation populations. Self-efficacy may assist in improving knowledge and attitudes with regard to changing behaviour and while several barriers act to restrict people's ability to implement changes, encouraging a positive attitude to change may give people the necessary mind-set to make small changes. The benefit of this is two-fold as not only may individuals feel more positive about change, but being surrounded by others that have a positive attitude may have a ripple effect due to support and improved knowledge among people around them.36 Previous studies that have used self-efficacy in an attempt to improve fruit consumption reported an increase of 0.85 servings per day.37 Another study resulted in an increase of 0.56 servings per day along with improved knowledge, self-efficacy and attitudes.21 These interventions involved brochures, posters, church bulletins and advertising in local newspapers and on local radio stations demonstrating that in order for an intervention to be effective, it is important to have the support of, for example, local community groups, schools and churches.38,39

An intervention only at the individual level is unlikely to be the most effective approach for residents in Otara and other high deprivation populations. Given the strength of the Māori and Pacific Island community in Otara, it would perhaps be more appropriate to approach an intervention in a sense of ‘collective efficacy’.39 This notion incorporates the strength of a group or organisation thereby embodying people with a sense of solving their own problems and improving their lives through concerted effort.18 Development at the community level is therefore thought to be more effective for such cultural communities because it “incorporates empowerment both as a means and end”.39 Further, interventions must not only be based on behavioural theories, such as self-efficacy17,18 or social marketing40 for example, but should also directly involve influential people within the community.36,41

Limitations and suggestions for future research

The results from this study can not be generalised beyond the sampled population and future research with representative high deprivation populations is needed. Further research in high deprivation populations with other ethnic groups would be a worthwhile extension on this study and would provide supplementary evidence of the relationship between self-efficacy and fruit consumption in high deprivation populations.

It would be encouraging to see more research that considers the social and cultural context within which recommendations are made. People living in high deprivation population are frequently being advised about practices they should or should not be engaging with as opposed to being provided with the tools to help them to do so. Successful interventions therefore require a thorough understanding not only of the nutritional changes required, but of how these may be attainable in the relevant social and cultural environment. Research that encompasses such a holistic perspective would be a welcome addition to our study.


Findings from the study show that self-efficacy is significantly related to consuming two or more servings of fruit per day, and that self-efficacy is not significantly related to high deprivation, indicating that this may be useful in interventions aimed at increasing fruit consumption. We therefore suggest that development of effective interventions to increase fruit consumption in high deprivation populations requires a holistic and multi-faceted approach which incorporates strategies for increasing self-efficacy. Further, such interventions require consideration of how recommendations may be best conveyed and applied in the specific social and cultural environment being targeted.


The authors would like to gratefully acknowledge the participation of the Otara residents who participated in this study. This research was funded by The New Zealand Institute for Plant & Food Research Ltd. Additional thanks to Timoti Pahi for input on the manuscript. Thanks are also extended to Lynette Ferguson at the School of Medical Sciences at the University of Auckland and Julie Park at the Department of Anthropology at the University of Auckland for their input and suggestions on an earlier draft.