Introduction

How do people in midlife (ages 40–60) feel about their lives in developed countries today? This may seem like a simple question, but so far research has generated answers that are muddled and contradictory. On one side, some researchers claim to have identified a “midlife nadir” in well-being that applies across countries, and may even apply to other primates (Blanchflower and Oswald 2008; Weiss et al. 2012). On the other side, many researchers claim that midlife is in fact a time of life when people experience a peak in well-being, overall as well as within the specific domains of work and personal relationships (Easterlin 2006; Sutin et al. 2013). The goal of the present paper is to add further dimensions to this portrayal of midlife well-being, by posing a new set of questions and asking them in a new way.

A Midlife Nadir?

Over the past two decades, a substantial literature has accumulated that claims to find a persistent nadir in well-being at midlife. In a comprehensive review, Blanchflower and Oswald (2008) summarized a wide variety of findings purporting to show a U-shaped pattern of well-being through adult life, in data sets including the General Social Survey (GSS) in the United States from 1972 to 2006; the Eurobarometer surveys in Western European countries from 1976 to 2002; data from the World Values Survey in 80 countries from 1981 to 2004; the Latinobarometer surveys of 18 Latin American countries over a period from 1997 to 2004; and the Asian barometer surveys of 15 Asian countries during 2003–2004. Across these surveys, well-being was found to be relatively high in the twenties, then slopes downward, reaching a minimum at some point from the mid-thirties to the late sixties before sloping upwards again.

Three kinds of objections have been raised to claims of low well-being during midlife. First, nearly all the studies on which this claim is based have used a single-item measure of well-being. For example, in the GSS, the question is, “Taken all together, how would you say things are these days? Would you say that you are very happy, pretty happy, or not too happy?” It may be doubtful that well-being can be effectively summed up in a single item. As Blanchflower and Oswald (2008) conceded, a single-item measure of well-being “cannot allow subtle differentiation, as favored in some psychology journals, into what might be thought of as different types of, or sides to, human happiness or mental health” (p. 1734).

The second objection to claims of a midlife nadir in well-being concerns the effect size and the interpretation of the U-shaped pattern. For the most part, studies investigating age-related patterns of well-being involve large national samples, in the tens of thousands or even hundreds of thousands. In such large samples, small differences in well-being between age groups may produce statistically significant results, but are these differences meaningful? For example, Cheng et al. (2015) reported a U-shaped pattern in a German sample. On a 10-point scale of life satisfaction, median well-being was about 7.2 in the early twenties, sloped downward to about 6.8 by the early fifties, then sloped upward to about 7.1 by age 70. This finding was taken by the authors to be a confirmation of low midlife well-being, but it seems reasonable to question whether a change from 7.2 to 6.8 on a scale of 1–10 constitutes an important age-related change. Furthermore, as this example illustrates, studies on well-being find consistently that most people at all ages rate their well-being as relatively high, considerably above the mid-point (Diener 2012). Consequently, even if the U-shape is regarded as valid, it appears to indicate not that midlife is a time of despondency but that it may be, at most, a time when well-being is slightly less favorable than earlier or later.

The third objection concerns the use of covariates in analyses of well-being. Most studies purporting to show a decline well-being during midlife use a variety of covariates in the analyses, including education, income, marital status, and whether or not there are children at home (Blanchflower and Oswald 2008). However, given that these characteristics vary substantially over the adult lifespan, it is questionable whether controlling for these variables provides a truer picture of the age effect on well-being or a distorted and artificial comparison (Frijters and Beatton 2012; Glenn 2009). Some investigators who have presented age patterns without covariates not only fail to find the U-shaped pattern in well-being across adulthood but find the opposite, that well-being peaks in midlife (Easterlin 2006; Mroczek and Kolanz 1998).

A Midlife Peak?

Supplementing the single-item surveys that have shown a peak in well-being at midlife without covariates, broader studies have shown that midlife is a time of contentment and high life satisfaction in most domains for most people (Freund and Ritter 2009). The most comprehensive study of midlife development to date is the Midlife in the United States (MIDUS) project, which began in the 1990s and is still ongoing. Although the focus of the study was on midlife development, which they defined as ages 40–60, the age range of the sample was 25–74, to allow comparisons between adults at midlife and younger and older adults. The initial sample has also been followed longitudinally 10 years later.

The MIDUS study found that in most respects, people in midlife held a favorable view of their lives, and were more positive than younger or older adults (Brim et al. 2004). Compared to younger adults, midlife adults had higher marital satisfaction and work satisfaction, and they were more content with their financial situation and their relationships with their children. However, they reported lower sexual satisfaction, and were more likely to report problems with their physical health. Nevertheless, they also had higher overall life satisfaction. Young and midlife adults reported more frequent days of experiencing any stressors or multiple stressors compared to older adults, and the two younger groups also rated their stressors as more disruptive and unpleasant than older adults did.

Longitudinal studies add further complexity to the literature on age and well-being. In the MIDUS study, when the original sample was followed-up 10 years later, overall life satisfaction increased from the forties to the fifties and from the fifties to the sixties, and decreased from the sixties to the seventies and the seventies to the eighties (Lachman et al. 2015). In general, life satisfaction was relatively low from age 25 to 50 and relatively high from age 50 to 80. However, all age groups rated their life satisfaction relatively high, between 7.5 and 8.5 on a ten-point scale, at both Time 1 and Time 2. In contrast, Cheng et al. (2015) claimed to find the U-shaped pattern of life satisfaction in four longitudinal data sets, with a low point in midlife. Still, this low point was mild at best, and life satisfaction was relatively high at all ages.

Varieties of Well-Being

Well-being can be understood as comprised of three aspects: evaluative, experienced, and eudaimonic (Stone and Mackie 2013). Evaluative well-being entails broad judgments of life satisfaction, including overall life satisfaction as well as satisfaction in specific domains, such as relationships, work, and health. Experienced well-being refers to temporary emotional states such as anger, sadness, or joy. Eudaimonic well-being includes perceptions of the purpose, value, and meaningfulness of one’s life.

These varieties of well-being appear to have different trajectories over the adult lifespan. Evaluative well-being has been the focus of the most researches, and, as discussed earlier, some studies appear to show a U-shaped pattern, whereas other studies show in an inverted U-shape, and still others show that evaluative well-being varies across domains (Frijters and Beatton 2012; Lachman et al. 2015; Stone et al. 2010). Experienced well-being appears to improve with age, across studies, as the frequency of positive affect increases and negative affect decreases (Lachman et al. 2015; Ryff 1995). Eudaimonic well-being has not been the focus of as much research as the other two types, and so far definite age patterns are not clear. One recent study of 29 European countries found no relationship between age and eudaimonic well-being across the adult lifespan in the most affluent European countries, and fluctuating or declining well-being in the less-affluent countries (Morgan et al. 2015).

The Present Study

Given the inconsistent, contradictory, and controversial literature on well-being in midlife, a new approach may be useful. The present study takes a new approach in two ways. First, participants were asked to make self-appraisals of various aspects of their well-being at “this time of my life.” Thus, the study assessed evaluative well-being, but in a way that asked participants to make the assessment by thinking broadly about the period of life they are in, rather than asking “how satisfied” or “how happy” they are in the present alone. This approach is not necessarily better or worse than previous ways of assessing evaluative well-being, but it is different, so it may yield results that provide a new perspective on well-being in midlife. Second, the present study asked a different set of questions about well-being than previous studies have asked. As noted, previous studies have either asked a one-item question on overall well-being, or asked about life satisfaction in specific domains such as health, work, and marital relations. The present study asked participants to assess their lives with new kinds of items, such as whether they view their current lives as a time of freedom or believe that they are at a time of life when anything is still possible. The present study also asked about various “sources of enjoyment” and “sources of stress.” Again, these items are different than the items used in previous studies, so the results may add new dimensions to the understanding of well-being at midlife.

The following research questions were proposed:

  • Will midlife adults generally view their lives positively, such as a time of freedom and possibilities, or negatively, such as a time of stress and uncertainty—or both?

  • What will be the main sources of enjoyment and stress reported by adults in midlife? Will midlife adults be more likely to report joys or stresses?

  • How will contextual variables such as gender, ethnicity, work status, marital status, and socioeconomic status be related to well-being in midlife?

Method

Participants and Procedure

The participants were 834 adults ages 40–60 (M = 51.9, SD = 7.6) residing in the United States. Although there is some variation across studies, 40–60 is commonly used as the age range for midlife (Freund and Ritter 2009: Lachman et al. 2015).

The data collection for this survey, the Clark University Poll of Parents of Emerging Adults, was conducted in 2013 by Purple Strategies, a survey research firm. Three methods were used to obtain participants: 426 via the internet, 334 via landline telephone, and 74 via cell phones. The internet sample consisted of members of a demographically diverse online panel assembled by the survey research firm. The phone participants were obtained via random-digit dialing. No participants were paid or provided with other compensations in return for their participation. The research was approved by the Institutional Review Board of Clark University.

These three methods were used to obtain a diverse sample that would reflect the population of 40–60-year-olds in the United States. Using telephone sampling alone is no longer viable for survey research. Survey sampling of cell phones via random-digit dialing is restricted by federal law in the United States, and rates of participation for those who are reached via landline have declined and are very low (Blumberg and Luke 2013). Consequently, the internet sample was necessary to reach segments of the population that would not be accessible via either landlines or cell phones.

Table 1 shows the demographic characteristics of the sample. About half of the participants were female and half were male. In terms of ethnicity, 72% identified themselves as White, 13% African American, and 10% Latino/a. Most (71%) were currently married, whereas 15% reported “no current relationship.” Participants were sampled from all regions of the country. They were from diverse social classes, as represented by their educational attainment. Slightly over half (57%) were employed full-time (40 or more hours per week), whereas 10% were unemployed but looking for work, and 16% were retired. Because the data for this paper were collected as part of a study of parents of 18–29-year-olds, no non-parents were in the sample. However, in this generation of American midlife adults, over 90% have had children (Brim et al. 2004).

Table 1 Demographic characteristics of the sample (N = 834)
Full size table

The total sample was demographically similar to the United States population of 40–60-year-olds. With regard to ethnicity, the overall United States population of adults aged 30 and older is 70% White, 13% Latino/a, 11% African American, 5% Asian American; and 1% Other (Taylor and Keeter 2010); in the present study the sample was 72% White; 10% Latino/a; 13% African American; 2% Asian American; and 4% Other. With regard to region, 18% of Americans live in the Northeast; 22% in the Midwest; 37% in the South; and 24% in the West (United States Census Bureau 2012); in the present study, the distribution was 21% Northeast; 22% Midwest; 33% South; and 24% West. With regard to SES (as measured by educational attainment), in the total United States population of persons aged 44–64, 31% have a 4-year college degree (National Center for Education Statistics 2014); in the present study, 43% of participants had obtained a 4-year college degree.

Measures

The survey covered a wide range of topics. For this paper, the focus was on items that pertained to well-being and self-assessment of their lives at this time.

Aspects of Well-Being

Participants were asked to respond to eight items regarding their perceptions of their well-being with respect to “this time of my life.” They were also asked about general life satisfaction (“Overall, I am satisfied with my life”). In addition, four items assessed negative well-being: “This time of my life is stressful,” “I often feel depressed,” “I often feel anxious,” and “I often feel that my life is not going well.” Items were answered with responses on a 4-point Likert scale. All the well-being items are shown in Table 2.

Table 2 Aspects of midlife well-being (N = 834)
Full size table

Sources of Enjoyment and Stress

Participants were asked, “Which of the following are current sources of enjoyment for you? Indicate ‘yes’ or ‘no’ for each.” There were 11 possible sources of enjoyment. Items are shown in Table 3. Participants were also asked, “Which of the following are current sources of stress for you? Indicate ‘yes’ or ‘no’ for each.” There were eight possible sources of stress. Items are shown in Table 4. Both the “enjoyment” and the “stress” items were based on the literature on midlife development (e.g., Brim et al. 2004; Lachman et al. 2015).

Table 3 Sources of enjoyment (N = 834)
Full size table
Table 4 Sources of stress (N = 834)
Full size table

Demographic Variables

The survey included items on a range of demographic characteristics, including age, gender, ethnicity, number of children, relationship status, educational attainment, and occupational status.

Results

The well-being of midlife adults in this study was highly positive. As shown in Table 2, a strong majority (82%) agreed that “Overall, I am satisfied with my life.” A majority also agreed that “this time of my life” is “fun and exciting” (71%), a time of freedom (71%), and a time when “anything is possible” (77%). Most regarded their current time of life as a time of changes (77%), as well as a time of focusing on themselves (56%) and “finding out who I really am” (55%). There was also a substantial proportion that endorsed negative aspects of their current well-being. Nearly two-thirds of the participants assessed this time of their lives as stressful (65%), half reported this time of their lives as a time of uncertainty (50%), and many agreed that they often feel anxious (39%), depressed (25%), or that “my life is not going well” (27%).

There were many sources of enjoyment reported by a majority of the midlife participants (Table 3). At the top of the list, in terms of prevalence, were “relationship with my 18–29-year-old children” (88%), hobbies or leisure activities (86%), travel or holidays (82%), and watching television (80%). A majority of participants also reported drawing enjoyment from “relationship with spouse or partner” (75%), pets (63%), “relationship with parents or spouse’s/partner’s parents” (62%), and exercise or playing sports (62%). Overall, work was a source of enjoyment for 49% of participants, but among those currently employed, it was a source of enjoyment for 59%. Using social media, i.e., “Sending emails and Facebook posts to family and friends,” was reported as a source of enjoyment by 43%, and “relationships with my grandchildren” by 37%.

Sources of stress were not as prevalent as sources of enjoyment, but were nevertheless reported by a substantial proportion of participants (Table 4). A majority reported being stressed by financial issues (62%) and work issues (51%). Among those who were working 40 or more hours per week, work issues were a source of stress for 63%. “My physical health” was reported as a source of stress by 42%, and “relationships with my 18–29-year-old children” by 36%. Less common sources of stress were relationship with spouse or partner (28%), caring for elderly parents or in-laws (25%), spouse or partner’s physical health (25%), and caring for grandchildren (10%).

Simultaneous regression analyses were conducted to investigate associations between the key demographic variables (i.e., age, gender, ethnic group, educational attainment, work hours, and close relationship) and four aspects of well-being (i.e., positive well-being, negative well-being, sources of joy, sources of stress). First, scales were created for positive and negative well-being. The four items comprising positive well-being were those appraising “this time of my life” as a time of freedom, as “fun and exciting,” and as a time when “anything is possible,” as well as the item on overall life satisfaction (M = 12.10, SD = 2.61). For negative well-being, the five items were those appraising this time of life as stressful and as full of uncertainty, as well as the items assessing whether the participant “often” felt depressed, anxious, or that “my life is not going well” (M = 11.12, SD = 3.61). Cronbach’s alphas were .71 for the positive well-being scale and .78 for the negative well-being scale. Summary scales were also created for “enjoyment” (11 items) and “stress” (eight items), by adding the “yes” responses for each scale, so that higher levels of each composite variable indicated higher levels of enjoyment (M = 7.28, SD = 1.94) and stress (M = 2.78, SD = 1.73), respectively.

To facilitate regression analyses, some of the key demographic background variables were recoded. For ethnic group, the Asian American (2%) and “Other” (4%) categories were combined, because the number of participants in these categories was too small to be analyzed separately. African Americans were used as the reference group for the contrasts concerning ethnicity in the regression equations, because studies have shown that the well-being of African Americans is distinctly influenced by racism, including in the areas of stress, depression, and anxiety (Pieterse et al. 2012). Thus, there were three ethnic contrast variables included in the regression analyses: African Americans vs Hispanic, White, and Other. For work hours, there were two categories: those who worked less than 21 h per week (32%; the referent group) and those who worked at least 21 h per week (68%). For close relationship, there were also two categories: occasional dating or no current relationship (19%; the referent group), and married, cohabiting, or close partnership (81%).

Table 5 shows correlations and descriptive statistics for all the variables in the regression analysis (except for ethnicity, for which correlations would not make sense because it was represented by contrasts). Several of the correlations were notable. Positive well-being was positively correlated with sources of enjoyment (r = .36, p < .01) and negatively correlated with negative well-being (r = − .56, p < .001) and sources of stress (r = − .31, p < .01). Negative well-being was positively correlated with stress (r = .45, p < .01) and negatively correlated with enjoyment (r = − .25, p < .01). All other correlations were small (.20 or less).

Table 5 Correlations and descriptive statistics for contextual and well-being variables (N = 834)
Full size table

The results of the regression analyses are shown in Table 6. There were no consistent patterns across analyses, and none of the regression analyses explained more than a miniscule amount of the variance. This indicates that the overall patterns of positive well-being, negative well-being, enjoyment, and stress were consistent across age, gender, ethnic group, educational attainment, work hours, and having a close relationship (or not).

Table 6 Simultaneous regression analyses predicting midlife well-being from contextual background variables (N = 834)
Full size table

Discussion

The primary goal of this study was to add new dimensions to the diverse and often contradictory literature on well-being in midlife. Previous studies have depicted midlife as a time of relatively negative well-being (Blanchflower and Oswald 2008; Weiss et al. 2012), as a time of positive well-being (Easterlin 2006; Sutin et al. 2013), or as a time when well-being depends on the domain in question (Frijters and Beatton 2012; Lachman et al. 2015; Stone et al. 2010) or on socioeconomic status (Morgan et al. 2015). The findings of the present study show that midlife is generally a time of positive well-being, but that for most Americans, it is stressful as well, and a substantial proportion frequently experiences anxiety or depression. Overall, Americans at midlife appear to be happily stressed: generally contented with their lives, with many sources of enjoyment, yet negative emotional states are not uncommon.

There was little support in this study for the claim that midlife is a slough of despond to be glumly endured between the happier periods of early adulthood and late adulthood (Blanchflower and Oswald 2008; Cheng et al. 2015). Substantial majorities of the midlife adults in the present study reported high evaluative well-being, viewing their lives as fun and exciting, as a time of freedom, and as a time when anything remains possible. Over 8 in 10 agreed that overall, they were satisfied with their lives. Similarly high proportions reported experiencing enjoyment in their current lives, from sources including their relationships (with their children, spouse/partner, and parents/spouse’s parents) as well as from activities such as travel, hobbies/leisure activities, watching television, and exercise/playing sports.

However, the experience of midlife is complex. It cannot be simply concluded that it is a happy time of life. For many, there appears to be an ongoing search for eudaimonic well-being, that is, for purpose and meaning. Over half agreed that “This is a time of my life for finding out who I really am,” a strikingly high proportion, given that identity development is more often associated with adolescence and emerging adulthood (McLean and Syed 2014). Also notable is that a strong majority characterized their current life as full of changes, and half as a time of uncertainty. Previous studies have sometimes depicted midlife as time of adopting “maintenance goals” rather than striving for new achievements (Freund and Ritter 2009), but for most of the participants in the present study, their self-development appears to be continuing during their midlife years.

Even amidst their overall high evaluative well-being, midlife is stressful in many ways, according to the participants in this study. Nearly two-thirds agreed that this is a stressful time of life, and they identified many specific sources of stress. Financial issues and work issues were a source of stress for a majority, and substantial proportions named sources such as their physical health and relationships with children or with spouse/partner. Furthermore, experienced well-being was reported as problematic for many participants, as 39% agreed that they often feel anxious and 25% agreed that they often feel depressed. However, other studies have found that midlife adults often learn from stressful experiences and can find meaning—i.e., eudaimonic well-being—in confronting and overcoming adversity (Lachman et al. 2015; Lee et al. 2015).

Given the complexity and diversity of these findings, perhaps the effort to determine if midlife is “overall” positive or negative is misguided and simplistic. Although the findings of the present study reinforce previous studies that have reported that midlife is a time of high evaluative well-being, clearly positive well-being often coexists with feeling stressed and anxious for many. Most people report a variety of sources of enjoyment in their daily lives, from the mundane (watching television) to the meaningful (close personal relationships), but most also experience stress from a variety of sources—especially financial and work issues, but also, often, from the very relationships that serve as sources of enjoyment.

Limitations of the Study

The present study adds a new perspective on well-being in midlife, using a broad national sample of persons aged 40–60. However, there were several limitations to the study that should be mentioned. First, many aspects of well-being in midlife were measured with a single item. Although the present study employed far more items to measure midlife well-being than has been typical in studies of well-being, which have often used only one item of overall well-being, most studies in psychology use scales of multiple items to assess constructs such as anxiety, depression, and stress (e.g., Luthar and Ciciola 2016). It would be advisable to seek to replicate the findings of the present study using scales with multiple items. Second, all the midlife adults in the present study were parents of emerging adults ages 18–29. Studies have found that when children reach emerging adulthood, parents’ well-being often rises, both their overall well-being and their satisfaction with parenting (Arnett 2015; Bouchard 2014). Consequently, in the present study, the well-being of 40–60 year-olds may be higher than it would be for 40–60 year-olds who have children of younger ages, especially for mothers of children in early adolescence (Luthar and Ciciola 2016). Further research in this area should seek to include midlife adults with children of diverse ages.

Conclusion

In sum, the findings of this study indicate that the well-being of midlife Americans is generally high, in multiple respects, but that stress and anxiety are common. Furthermore, midlife is not a period of stagnation or stability for most people, but a time when they continue to explore and develop their identities. This search is frequently accompanied by change and uncertainty. Future research should continue to explore the diverse aspects of well-being in midlife, and should compare it to other periods of the lifespan. There should also be further exploration of what makes midlife satisfying and enjoyable and what does not. Because life expectancies are extending steadily longer worldwide, midlife is no longer the entryway to decline and debility but—potentially—the springboard to a new period of freedom and joie de vivre.