Medical Sociology

lable at ScienceDirect

Social Science & Medicine 183 (2017) 28e36

Contents lists avai

Social Science & Medicine

journal homepage: www.elsevier.com/locate/socscimed

Sadness or depression: Making sense of low mood and the medicalization of everyday life

Christian Br€oer a, *, Broos Besseling b

a University of Amsterdam, Department of Sociology, The Netherlands b Free University, The Netherlands

a r t i c l e i n f o

Article history: Received 3 June 2016 Received in revised form 11 April 2017 Accepted 13 April 2017 Available online 15 April 2017

Keywords: The Netherlands Medicalization Depression Everyday life Observational methods

* Corresponding author. University of Amsterdam Nieuwe Achtergracht 166, PO Box 15508, 1001 NA Am

E-mail address: [email protected] (C. Br€oer).

http://dx.doi.org/10.1016/j.socscimed.2017.04.025 0277-9536/© 2017 Elsevier Ltd. All rights reserved.

a b s t r a c t

This research focusses on low mood as a generic category in everyday social interactions, outside the clinical realm and among non-patients. We examine if and how a clinical depression label and treatment are employed when low mood occurs in everyday life, which enables us to analyze the extent and content of medicalization and brings to the fore the interactional mechanisms and cultural concerns that potentially drive medicalization. The analysis is based on 316 observations of everyday life in the Netherlands. We observed and recorded interactions in which low mood was spontaneously expressed.

Our paper shows that the clinical depression label resonates widely even if low mood is not fully medicalized. People de-medicalize low mood, and low mood can be un-medicalized. Our analysis thus suggests that dominance is not achieved, which nuances Horwitz and Wakefield's claim that the clinical category of depression has come to encompass all forms of low mood. Moreover, uncertainties about the meaning of low mood and about the depression label remain pragmatic concerns of everyday life.

The cultural norm of happiness and active citizenship are very prominent in everyday life across medicalized and un-medicalized interactions. These norms thus seem to be a necessary but insufficient condition for medicalization. While pragmatic concerns do not seem to trigger medicalization either, one specific type of concern is consistently related to medicalization: relational conflicts.

In sum, the cultural construction of low mood is not dominated by a single medical approach; how- ever, it mirrors the diversity and uncertainties within and around the medical field.

© 2017 Elsevier Ltd. All rights reserved.

1. Low mood in everyday life

This article analyzes everyday interactions to determine how low mood is communicatively addressed and which pragmatic concerns are relevant to it. In this study, low mood designates a generic and etic category that includes emic expressions such as prolonged depression or mildly feeling down.

This article scrutinizes the “medicalization of life” (Conrad, 2007; Crawford, 1980; Szasz, 2007) hypothesis that the expansion of medical diagnosis and treatment transforms common un- derstandings of low mood. In particular, it addresses Horwitz and Wakefield's (2007) claim that people have “lost” the ability to experience “normal sadness” and approach low mood largely as clinical depression. Horwitz and Wakefield convincingly show how

, Department of Sociology, sterdam, The Netherlands.

the diagnosis and treatment of depression have come to include increasing instances of low mood. Decontextualized diagnostic criteria were gradually applied to outpatient populations and thereby imported into the community, according to Horwitz and Wakefield. However, the uptake of medical registers in commu- nities is beyond the scope of these authors' research. Whether “normal sadness” is indeed considered to be “clinical depression” must be examined.

Horwitz and Wakefield urge social scientists to distinguish be- tween normal and pathological sadness, particularly with the goal of criticizing the overexpansion of diagnostic categories. In this research, a different approach is followed: we include all expres- sions of low mood and attempt to see if a medicalized “idiom of distress” (Nichter, 2010) is at work and, if so, in which situational contexts (Bell, 2016).

Medicalization might be limited (Williams and Calnan, 1996), and de-medicalization is studied (Torres, 2014). However, medi- calization research has focused largely on patients, pills and pro- fessionals and neglected whether diagnosis and treatment are

C. Br€oer, B. Besseling / Social Science & Medicine 183 (2017) 28e36 29

relevant among non-patients, outside the clinical realm, at home, in public or during work. Most medicalization research has focused on situations in which medicalization has already taken place. This is a crucial omission because, for medicalization to occur, it needs to settle in everyday life.

2. Low mood and depression

Cross-culturally and over time, humans have addressed low mood and the category of depression in various ways (Kirmayer, 2001; Kleinman, 1985; McPherson and Armstrong, 2009). More- over, the speed by which depression diagnosis has risen, along with the marketing of anti-depressants and the lack of bio-genetic ex- planations of depression, further emphasize the interpretive flexi- bility of diagnosis and treatment. Flexibility, however, seems less pronounced in light of phenomenological studies that show regu- larities in illness careers (Kangas, 2001; Westerbeek and Mutsaers, 2008), biographical disruption and existential suffering (Karp, 1994; Ratcliffe, 2014).

We acknowledge that human distress arises in a dialectical relation between nature and nurture (Kleinman, 1985 p.11). Low mood can become a pragmatic concern in mundane interactions which may lead to medicalization, given the availability and dominance of medical categories. In interactions, low mood can be considered pathological depression, common sadness or another condition entirely.

By focusing on life outside the clinical realm, we can more easily observe whether broader cultural concerns enter the medicaliza- tion of low mood and whether depression labelling and treatment become a cultural code. From a cultural perspective, one would expect a gradual fit between culture and diagnosis since disease categories can become ways of experiencing illness and, over time, “we learned approved ways of being ill“ (Kleinman et al., 1978, p. 252). In Western countries, suffering might be a partial result of low mood undercutting the norms of active citizenship and a respon- sible and happy life (Petersen, 2011; Philip, 2009; Rose, 2007).

In the case studied in this paper, mental health institutions are rife with uncertainties. There is uncertainty about the meaning and validity of depression diagnosis and treatment (Fox, 1957; Kokanovic et al., 2013). Depression diagnosis and treatment have been critically assessed publicly, and the medicalization critique, which is half a century old, might have become a common cultural repertoire by now. By focusing on everyday interactions in which low mood is relevant, this research illuminates the pragmatic concerns that may drive medicalization, given the ontological un- certainty surrounding depression labelling.

Our analysis is based on 316 observations in the Netherlands. We encountered mourning after loss, joblessness, stress, violence, marginalization, substance abuse, somatic conditions, broken dreams and doubts about life choices in relation to low mood. People report themselves or others feeling a lack of energy, being unable to get out of bed for days, anxiety, avoiding contact, distrust, uselessness, suicidal thinking, lack of sleep, panic, lack of an appetite, hypersensitivity and tension or anger. Sometimes, only a light reference to one of these phenomena is made, while on other occasions, long spells of suffering were reported.

We searched for communicative references to low mood and analyzed if and how medical terms and treatments were implied. While the use of the word “depression” might point to medicali- zation, it is also used to refer to bad weather or a lazy morning. Therefore, a large part of this research involved the repeated interpretation of the meaning of words in context. As shown below, this is also what people do themselves: attending to interactions illuminates the shifting uncertainties about (medical) categories in everyday life.

3. Medicalization between institutions and pragmatic concerns

In response to the global increase in many (mental) health di- agnoses, diversity in prevalence statistics, a lack of biological ex- planations and the mega-marketing of pharmaceuticals, social science scholars have developed theories of ‘medicalization’ (Conrad and Schneider, 1992), ‘bio-medicalization’ (Clarke et al., 2003) and ‘pharmaceuticalization’ (Abraham, 2010). Medicaliza- tion describes a process by which formerly non-medical problems come to be defined and treated as medical problems. Studies have focused on the power of medical professionals, biomedical re- searchers, pharmaceutical firms, policy-makers and activists to turn normal behavior into medical conditions.

Research has provided us with manifold insights into the social construction of health conditions and has often highlighted the downsides of medicalization: the regulation of deviant behavior (Scheff, 1970) and unwanted effects, such as stigma (Goffman, 1961), rising costs (Conrad et al., 2010) and damaging treatment (Illich et al., 1975). Classic medicalization studies often followed an institutional approach and were largely based on top-down models of social control. In response, scholars influenced by Foucault pointed to the diversity in clinical practice and the productive side of medical authority (Hacking, 2007; Rose, 2007). Nevertheless, Lupton identified a major deficiency in Foucauldian approaches to medicalization, which “neglect examination of the ways hegemonic medical discourses and practices are variously taken up, negotiated or transformed by members of the lay populace” in everyday life (Lupton, 1997 p.94). Existing bottom-up studies have reported that patients inform their doctors about how to interpret symptoms, that activists influence medical practices (Brown et al., 2004; Epstein, 1996) and that knowledge of diseases spreads through local networks (Liu et al., 2010). People seem ‘eager for medicali- zation’ (Becker and Nachtigall, 1992), claiming institutionally un- recognized conditions (De Graaff and Br€oer, 2012; Dumit, 2006). Pharmaceuticals in everyday life have been the subject of recent research (Bell and Figert, 2012; Coveney et al., 2012; Graf et al., 2014; Hardon et al., 2013). Studies like these show that, while ‘classic’ medicalization sought to separate the working population from the sick, healthy people are now additionally encouraged to maximize their physical and emotional well-being through self- medication.

Nevertheless, most studies conducted to date have focused on patients, professionals and treatments as the starting point for anal- ysis and theorizing. In these cases, medicalization has already taken place to a significant degree. Retrospective interviews offer some access to the process leading up to the diagnosis or help seeking. However, those who do not seek diagnosis or treatment are over- looked, which can easily lead to overestimating and misconstruing medicalization. Recent analyses of self-help books (Barker, 2014; Philip, 2009), for example, suggest addressing medicalization in everyday life, but they do not address the actual uptake of self-help registers. Medicalization must take root in commonly held assump- tions among people who have not yet been exposed to doctors, diagnosis or treatments, and studies in this respect are lacking.

Therefore, we propose shifting the analytic lens to everyday life and common interactions and then ask if and how medical cate- gories and treatments are relevant. In this way, we can address the extent to which medicalization has affected life outside clinical encounters, include the experience of low mood, scrutinize inter- actional concerns about low mood and avoid top-down concep- tions of medicalization, which enables us to scrutinize non- medicalization.

Everyday life refers to people's mundane beliefs and practices (De Certau, 1984; Schutz, 1932; Scott, 2009). Following an

C. Br€oer, B. Besseling / Social Science & Medicine 183 (2017) 28e3630

interactionist perspective (Blumer, 1969; Mead and Morris, 1934), this study focuses on concerns and conflicts and people's attempts to solve them. With regard to health, this approach has long been applied fruitfully to chronic conditions, suffering, identity (Bury, 1982; Charmaz, 1983), everyday health experience (Saltonstall, 1993) and medicalization (Rossol, 2001). Lock and Kaufert (1998) proposed studying pragmatic concerns in relation to medicaliza- tion and emphasized the strategic use of medicalization. We build on their research and widen the conceptual lens to include any problem-solving attempt, whether strategic or not. It is assumed here that humans are inclined to solve the problems they experi- ence and that adverse (mental health) experiences cannot be reduced to institutional arrangements.

From a conflict perspective, the dominance or hegemony of a medical register is never uncontested nor absolute and might even produce countervailing tendencies. Therefore, we interrogate the breadth of medicalization and its content (Brown et al., 2015) and include cases where medicalization does not occur or is contested. We specify people's relation to medical diagnosis and treatment according to the ‘resonance model’ developed earlier (Br€oer, 2008; Br€oer and Heerings, 2013). We attend to 1) consonance, where everyday understandings and practices are fully aligned with medical ones, 2) dissonance, where people in everyday life oppose or rework medical understandings and practices, 3) autonomy, where no medicalization takes root in everyday life, and 4) feed- back, where everyday understandings and practices shape clinical ones (De Graaff, 2016). Medicalization is thus considered the outcome of the dialectical relation between mental health in- stitutions and everyday life concerns. We expect that people's concerns and pragmatic attempts at problem solving are an understudied micro triggers of medicalization.

4. Methods

The observational material for our analysis was gathered in the Netherlands. According to survey data and patient registrations, depression is reported yearly by approximately 8e10% of Dutch people, while approximately 5e6% of are treated medically, usually by a general physician administering anti-depressants, slightly more than 1% are treated in specialized care and approximately 0.03% are hospitalized for depression (Verkenning, 2013; Verweij and Houben-van Herten, 2013).

Studying interactions at home, at work or in public, we ask when, where, by whom and how low mood is experienced, expressed and dealt with both by the sad person and those in his/ her immediate social environment outside clinical settings or before diagnosis and treatment. We search for instances of the medicali- zation of low mood, the modification of or resistance to depression diagnosis or treatment and instances where depression diagnosis and treatment are irrelevant. Additionally, we investigate which interactional concerns are at stake in relation to low mood, how these are concerns dealt with, which norms are enacted, and whether these factors push towards medicalization?

We collected data through focused ethnography (Hammersley and Atkinson, 2007; Knoblauch, 2005) using purpose-built obser- vational logs. The instrument has been tested repeatedly by the first author and builds on ‘mass observation’ (Danforth and Navarro, 2001; Willcock, 1943), ‘self-recording’ (Wheeler and Reis, 1991) and experience sampling (Csikszentmihalyi and Larson,1987). Both authors participated in the collection and analysis of the data. The first author trained sociology students, including the second author, who participated in a course about “citizenship and health” at the University of Amsterdam. In five consecutive years, the first authors assigned data collection and interpretation tasks, and students received an observational form, instruction and multiple feedback

sessions to fine-tune observation and interpretation. Observers were instructed not to ask about or trigger the ex-

pressions of low mood. Observations centered on students’ own lives and the moments when they encountered low mood in themselves or others on the street, at home, in schools and at work in any kind of interaction. We included observations that referred to low mood broadly defined. Any experience or communicative mentioning of feeling down, sad, depressed or somber was included. We also included the media reports that students encountered; media data are included in the analysis in regard to medicalization in general. Students were asked to keep a notebook and take notes as soon as possible after each occurrence. Students reported on their own lives and consented to sharing these reports with fellow students and researchers. Observations were shared on a secure university platform and could be uploaded anonymously.

Since the student-observers were part of the interactions, they were asked to collect contextual information, which facilitated interpretation. Dozens of observation were furthermore discussed in class, which added to the depth of this contextualization. Stu- dents were trained to use their awareness of the situation and their own emotions for the log and for the analysis. The classroom dis- cussions were meant to build an “interpretive zone” (Wasser and Bresler, 1996) in which differing perspectives on the material could be elaborated and cross checked.

From 2009 to 2013, 137 students posted 479 observations, of which we randomly selected 316 according to time constraints for coding, which lasted about six months full-time. Sampling went far beyond the point of theoretical saturation to arrive at descriptive statistics. We constructed 168 codes, 1457 quotations and 5503 codings. In 60% of the observations, low mood related to women, and in the other 40%, it related to men. The average age of the students was 23.9, with a range from 20 to 43 (SD: 4.4). Since the students encountered a range of other people, the average age of depressed/sad people was 29.5, with a range from 6 to 82 (SD: 14). Below, we report the age, if available. In approximately 10% of the interactions, we found reference to a person being diagnosed with depression, and on a few occasions, we came across a professional. Since the observers were students, observations do not represent the general public. However, in the observations, non-students and a variety of social situations come to the fore.

Data were analyzed in two ways: deductively, qualitative the- matic analysis guidelines were followed (focusing on the explicit mentioning of low mood and medical categories and treatments) (Hsieh and Shannon, 2005; Mayring, 2000), and we coded for a number of standard variables (gender, age, and place). Inductively, we noted what participants considered relevant in relation to low mood, particularly interactional concerns and norms (Horton- Salway, 2001). The analysis was assisted by Atlas.ti software.

During the formal analysis, neither author initially attended to any particular aspect and therefore coded openly in combination with extensive memo-writing and collaborative analysis. A wider group of colleagues participated in two coding sessions to assess the reliability and validity of emerging codes. In the course of the analysis, theory-driven concepts were fine-tuned and applied to the material. The collaborative interpretation of observations revealed that the analysts’ uncertainty about the meaning of the depression label resembled the uncertainties about low mood and depression people experience in daily life.

An ethical approval statement was not required due to the ethics procedure of our research school and university during the time of data collection. Nonetheless, this research has been presented to and discussed with the ethics board of the research school. Following their suggestions, the data collection procedure was changed to offer students the possibility to contribute data anon- ymously. Al students included in the study consented in

Fig. 1. Low mood and resonance of medicalization as percentages of codings (n¼335).

C. Br€oer, B. Besseling / Social Science & Medicine 183 (2017) 28e36 31

participation, data sharing and publication.

5. Low moods

As said, we use the term low mood as an etic category to encompass emic terms. The Dutch equivalent was not used in the observed interactions. People spoke of low mood in a number of ways: as depressed or “depri”, a Dutch shorthand; having no energy or enjoyment; not feeling well; being down, apathetic, dispas- sionate, doubtful, or lonely; not getting outside or feeling closed off; or reporting tension, misery, burden, obligation, pressure and meaninglessness. In particular, people used the following meta- phors: experiencing a “dip”; being in a “well”; going through a difficult phase; not feeling comfortable in one's skin and a “snap in the brain”. These registers appear to be based on spatial imagery (high-low mood) and reasoning in terms of phases, processes, burden and misery. People also made sense of mood with bodily imageries (tension and discomfort in one's skin). References to “the brain” or neurochemical models (Rose, 2003) were infrequent, appearing in 13 cases (4%), of which 10 came from media coverage. A wider biomedical “explanatory model” (Kleinman et al., 1978) e including genetics e appeared in approximately 30% of references to explanations. In 64% of the explanations, people employed a psychosocial model (Kangas, 2001; Westerbeek and Mutsaers, 2008) and related low mood to major disruptive life events such as divorce, unemployment, loss, disease, abuse and violence and mentioned stress, familial conflicts, doubts about life choices or monotonous work. Next to medical treatment or counseling, many people referred to talking, moving to a different house, dressing up, or participating in sports or leisure activities as remedies. In our sample, people experienced, mentioned or encountered low mood most often at home (36%), at restaurants and caf�es (19%), at work or school/university (15%), in mediated conversations (12%) and on the street (10%).

Asking if and how medical categories and treatments resonate when low mood occurs, peoples everyday interactions contained four variants:

1. Medicalization: people approach low mood consonant with the categorization of medical depression.

2. De-medicalization and normalization of low mood: people take a dissonant stance towards medical depression categorization and try to normalize low mood against medicalizing tendencies.

3. Non-medicalization: people take an autonomous stance to- wards medical depression categorization and do not relate low mood to depression at all.

4. Normalization of depression labelling and treatment: people seeking doctors' advice or medical treatment is presented as normal.

As a first entry into these variants, we present their numerical occurrence in the sample in Fig.1. While we are fully aware that the sample does not represent a population, the numbers are part of the question if medicalization is dominant in the sample. Domi- nance includes both a high frequency of medical categorizations and centrality in the construction of meaning.

Following this overview, we present two variants in greater detail: medicalization and de-medicalization. Non-medicalization is not confined to apparently mild cases of low mood and occurred in relation to prolonged suffering too. Normalization of treatment occurred in few cases. People sought doctors’ advice and treatment, which is consonant with medicalization. They also uncoupled treatment from disease and considered it to be similar to dieting or vitamin taking. This finding echoes recent research on pharma- ceuticalization (Fox and Ward, 2008).

5.1. “There was just something not right in my head”: medicalization

The first example of medicalization came from a coffee bar conversation witnessed by our student-observer. Two old friends inquired about each other's well-being and reflected on the time when Female (F) was depressed.

P41, small caf�e, M(ale) and F(emale), ages approximately 35

F: I am lot better recently.

M: Oh, luckily, but it's nice to hear!

F: I still sometimes think back to the time I told you that I was depressed and I took the antidepressants. When I had told you that, it was also a relief that you knew why I sometimes felt so sad and useless and did not want to do things. There was just something not right in my head. I had the idea that the pieces for you were also put into place.

M: Yes, that is correct, yes. I was glad I knew what was wrong with you.

F: I was also pleased. Especially that there was a drug that I could take. I really felt better because of it.

This is an example of the medicalization of low mood: when F felt sad, she sought medical attention, which, in retrospect, empowered her and alleviated her suffering. M confirmed her description of the past. We can also observe the pragmatic rele- vance of medicalization: the diagnosis eased past relational ten- sions. F stressed that she was “relieved” that M received an explanation for the sadness. Medicalization also enabled M and F to structure the current interaction and to address current and past feelings.

This example is consonant with (bio)-medicalization, as F referred to “something not right in my head” and drugs. Here, medicalization occurred in the past and was confirmed in the present. Clear-cut examples such as this most often involved retrospection and successful treatment. People pragmatically point

C. Br€oer, B. Besseling / Social Science & Medicine 183 (2017) 28e3632

to successful treatment to validate medicalization. The second example is a subtle form of incipient medicalization

and contains uncertainty.

P27 M(ale)1, student-observer, age 22, M(ale)2, age 21, in the subway

M1: Hey, why did you disappear so fast suddenly, at that party?

M2: Yes, I do not know …

M1: …

M2: I actually felt like shit. I did not feel like partying at all. And then there was also that issue in the street.

M1: Okay …

M2: … and as I stood there with my beer while I was not in the mood. Then, I went home. […] At home, I actually felt like shit. I did not sleep so well, and the next day, I stayed home all day. Did not feel well … really a bit depressed.

A pragmatic concern triggered this interaction: M1 could not make sense of M2's conduct and asked him about it. M1 remained open to M2's answers, replying with silence or “okay”. M2 gradu- ally offered longer replies and more detail. In the last response, he paused twice and summed up his feelings as “depressed”. We interpret this as incipient medicalization and see potential conso- nance with a depression diagnosis in that M2 seems to lists symptoms he then calls depression. M2 pragmatically used depression to arrive at a conclusion that brought sense and coherence to his account of troubling interactions and low mood. M2 might also have interpreted the open response of M1 as a lack of support. In this light, M2 medicalized his mood to achieve legitimacy.

This situation is a form of pragmatic problem solving because the term “depressed” almost literally answers a question. This ex- change also exemplifies that the category of depression is employed and made relevant far beyond clinical or professional encounters. It has become so common that it enables reflexive “self-labeling” (Thoits, 1985). However, this is not equivalent to a diagnosis. The label was used hesitantly: the response “really a bit depressed” constructs “depression” as somewhere between com- mon sadness and clinical depression, which we interpret as un- certainty about the label.

5.2. “Case of the winter blues?”: de-medicalization

While the above examples are consonant with medical catego- rizations, we now turn to dissonant relations. The first example of de-medicalization is a conversation between partners that occurred when the M(ale) came home from playing sports. They talked about a female friend.

P226, F(emale), 23 and partner, M(ale) 30, at home

F: And all is well with J. [sports friend]?

M: Yes, but it's not going so well with S. [ wife of J. ]

F: Oh?

M: Yes, she is not feeling so good, a little glum.

F: Case of the winter blues?

M: Well, you know that S. is always a bit gloomy. Now running (turning) to the doctor. [pause]

But if she would do something more, pick up a hobby and not only a part-time job and hanging out at home a lot, I think she will feel much better in her skin.

This part of the conversation again began with a question about someone's mood. F scrutinized why an absent friend, S, was “not feeling so good”: first with a probe (“Oh?”) then with an interpre- tation (“winter blues”). In response, M emphasized that S's feelings had existed for a while and that she was under medical treatment. M thereby assigned S an identity that explained her low mood. The reference to “the doctor” points to the medicalization of her iden- tity. Then, M paused, questioned S's identity (“But”) and re- interpreted S. M then de-medicalized S's situation, pointing to a lack of activity and motivation and suggesting that S should lead a more active life to drive out gloom.

M thus took a dissonant position toward medical treatment. His remarks were built on an active citizenship discourse instead of medicalization (Petersen, 2011). The low mood itself was not questioned, but the medical solution was. S was made responsible for tackling her low mood. This reinstated an active citizenship norm, but it did not separate the normal and the pathological through medicalization. This example shows that low mood, when treated by a doctor, is not a straightforward example of medicali- zation but is instead doubted. Medical treatment itself does not legitimize the depression label.

In the next example, the patient de-medicalizes low mood: the female student simultaneously seeks reassurance from her family doctor and criticizes him/her.

P247, University Library, F(emale)1, 28, student, F(emale)2, student

F1: Yeah, you still know that I told you a while back that I was at the doctor and I said I didn't feel comfortable in my own skin and didn't feel like doing things anymore.

F2: Yea …

F1: And that when the doctor, without actually asking further said, yes, perhaps you can try this (anti-depressant).

F2: Oh yeah, I remember, yes, you said …

F1: Yeah, I found it so bizarre that it was just prescribed so indiscriminately. And I'm already, it's not that I'm eh … quick to want to use medications. I just wanted to tell my story or something.

F2: Yeah, bizarre, huh …

The student sought help from her family doctor but resisted the pharmaceuticalization of her needs. This marks a dissonant relation (partly accepting and partly rejecting medicalization). The refer- ence, “I just wanted to tell my story” indicates an attempt to normalize low mood.

This dissonant position of de-medicalizing a doctor is in line with the critique of the over-prescription of anti-depressants. Cri- tiques of medicalization are common in media framings of low mood, including newspaper headlines such as “Numerous side ef- fects of antidepressants” and a talk show that ridiculed “just another telephone number” for depression diagnosis.

6. Pragmatic concerns in interactions

In almost all of the interactions, we were able to inductively discern one or more of the following pragmatic concerns:

C. Br€oer, B. Besseling / Social Science & Medicine 183 (2017) 28e36 33

1. Sense-making: interactions in which the question arises of how to make sense of your own or another person's low mood.

2. Responsibility: interactions in which the question arises of who has to do what in cases of low mood.

3. Rule breaking: interactions in which happiness and active personhood as norms are violated.

4. Identity: interactions in which low mood is related to the question of the identity of the sad person.

5. Relation: interactions in which the question arises of how to address a conflict in a relationship.

Again, as a first approach to these concerns, we present descriptive statistics that show that none of the concerns is dominant. Moreover, in our sample, all concerns co-occur with medicalization, de-medicalization and non-medicalization, which indicates that none of these concerns engenders medicalization. However, the strength and content of correlations vary, and there is a concern that almost always co-occurs with medicalization: rela- tional conflict. Below, we report on sense-making and relational concerns in detail, while the other concerns are briefly mentioned. (see Fig. 2).

6.1. Making sense of low mood

The questions “how are you?”, “how am I?” or “how is (s)he?” were fundamental to many interactions and triggered sense- making once low mood was on the table. Sense-making is often but not always tied to medicalization: of the 165 instances of sense- making, 102 coincided with medicalization, and in 31 instances, there was no medicalization, while in 32 cases, there was de- medicalization. Therefore, sense-making does not appear to lead up to medicalization.

Depression diagnosis and treatment often aided in making sense of mood, particularly in retrospect (as seen in P41 above). In other cases, invoking a clinical definition merely shifted concerns about meaning. In the example below, we see that invoking depression helps M but not F in making sense of the troubles of a third person, D.

Fig. 2. Types of concerns as percentages of all interactional concerns (n¼457).

P200, At home, F(emale) student-observer (age unknown) and M(other) 63

F: Hey, how nice, a postcard from D (our former cleaning lady). Do you ever speak to her?

M: That's nice that she sent a card. It's going pretty good with her now, since the past few days, she may return home.

F: Oh, what was the matter with her?

M: She was really depressed for a while. She checked herself into a center where she was taken care of internally.

F: Oh, what for? You couldn't tell that D was depressed. Well. She marched cheerfully around our house with a duster.

M: Yeah, but she said she was really all tied up in knots with herself. It's good that she asked for help.

In this exchange, M framed the troubles of her cleaner as depression, which, for M, addressed the pragmatic concern of making sense of the cleaning lady's troubles. However, for F, this produced a different concern: F had to adjust her image of the cleaning lady. While the depression label pragmatically seemed to make sense of the troubles of the cleaner, it also induced questions about her social identity.

The next example shows how diagnosis and treatment raise even more concerns. Here, two friends tried to make sense of the well-being of A, a friend of F1. F1 knew that A had been threatened physically while living abroad.

P190: Two female students, one is a student-observer (age unknown)

F1: By the way, A's psychologist has suggested that she should start with anti-depressants.

F2: Anti-depressants? Why?

F1: He says she is depressed.

F2: What do you think?

F1: I do not know, I think everything she's been through is just very difficult stuff, and yes, she may be depressed. I do not know, she seems uncomfortable in her skin, but it seems logical.

F1 and F2 tried to make sense of the diagnosis and treatment of A. F1 did not embrace the psychologist's definitions and treat- ment of A's troubles, but did not reject the diagnosis, either, which presented her with the question of how to make sense of the diagnosis and treatment. In the last reply, F1 did not reach a conclusion. In this sense, medicalization has not solved the pragmatic concern of making sense but complicated it instead. Treatment with anti-depressants itself did not establish cer- tainty. Cases such as this particularly undermine straightfor- ward assumptions about the medicalization of everyday life: even taking medication does not constitute an all-out embrace of the medical model of low mood. The material includes cases in which routine medication use was even seen as a normal part of everybody's lives, not a pathology.

7. Relational conflicts

In 78 instances, people expressed concerns about their relations with spouses, friends or colleagues in terms of low mood, and 65 of these went hand-in-hand with medicalization. The majority of the relational concerns were more or less pronounced conflicts (48),

C. Br€oer, B. Besseling / Social Science & Medicine 183 (2017) 28e3634

which almost always related to medicalization (47/48). We main- tain that these conflicts were leading up to medicalization.

In the first example, a young woman visited her friend at home and described an encounter with another friend:

P92, 2 Females, age 23 and 25, at home

F1: You know S, right?

F2: Yes, the blonde, I even got her on Hyves (Facebook variant, CB).

F1: She is really depressed! You know.

F2: Why?

F1: Well just, she's so negative about everything.

F2: About what?

F1: She dislikes everything and thinks it's no fun. It's just not nice anymore.

F2: Oh, yeah …

This brief exchange revolves around the pragmatic problem of how to address a friend's low mood, which we interpret as a mild but common relational conflict. F1 medicalized this conflict by referring to S as “really depressed”, which F2 first questioned (“why”, “about what”) and then confirmed (“Oh, yeah”). The con- flict was resolved by blaming the “depressed” friend. The depres- sion label helped F1 to distance herself from her then-former friend.

In our material, we see different ways of expressing the idea that a relation in which low mood as a concern is “not nice anymore”. Relations were strained when friends or family did not want to join festive events or pleasurable activities and when get-togethers were not joyful (28 of 78 times). In these cases, a norm of happi- ness was reinforced when significant others distanced themselves through medicalization.

In the next example, alongside distancing, we witness empathy. Two students talked about a fellow student in their student housing who stopped going to class, withdrew from student life and “looks bad”, according to them. They wondered how to address it and spoke about the possibility of him being depressed prior to this exchange.

P17, Male 24, Female, 24, meeting of student home seniors

M: I just do not feel like doing it anymore. I pulled him out of his room during the floor-dinner, well, reluctantly.

F: Yes dear, but you have to see it as a disease. He does not sit all day in his room to bully you.

M: No, I understand that, but I (pull my hand off) don't bother anymore.

Both M and F acknowledged that there is a relational conflict (“bully you”) and assumed that the fellow student was depressed, but they responded differently: M stopped trying to activate him, while F, a psychology student, exculpated the fellow student and foregrounded the “disease”. The depression label helped them both to adjust their approach and ease the relational tension. Implicitly, we could assume that M also expected the fellow student to act on his situation. We found a norm of being active and responsible underlying half of the relational concerns.

While the two former examples concern absent persons’ low moods, the next one demonstrates how “depressed” persons utilize the depression label to ease pragmatic conflicts. The following is

part of an email conversation between a teacher and a student. The student had left the room during class, and the teacher emailed her about it as follows:

P93, Female student, 25, Teacher 42, email from teacher

Good day Sara,

How is it going today?

Sincerely (cordially), Tom

Hey Tom

It is at least better than yesterday. I found it very dismal to walk away from the seminar. In recent weeks, I have ignored that I am actually still sort of 'recovering’, thus, I have again fallen into old traps, and the result is that the well-known burnout / depres- sion symptoms somewhat resurface. I am going to take better care of myself this time around, and hopefully I'll feel stronger quickly.

Greetings, Sara

In her response, the student foregrounded the discomfort, possibly shame, she experienced when she walked away from class. She apportioned blame on herself (“I have ignored”, “I am going to take better care of myself”) and medicalized the symptoms (“burnout/depression”), resolving the conflict and allowing her to return to class again. The teacher responded by saying, “no problem that you walked away”, “take your time” and “adjust your plan- ning”, thereby using both the medical and responsibility registers that Sara offered.

8. Medicalization and protracted uncertainty

Medicalization research often focuses on (proto)-patients, pro- fessionals and pills and thereby neglects to ask if and how medi- calization affects the non-clinical realm. Undoubtedly, a medicalized approach to low mood is institutionalized in the US (Horwitz and Wakefield, 2007) and beyond, but we do not know whether this has affected the way people deal with low mood in everyday life. In this study, using innovative methods, we looked at generic low mood in everyday life interactions in the Netherlands to determine if and how low mood is medicalized and which interactional concerns and cultural norms underpin everyday medicalization.

Outside the clinical realm, among friends, in public, at work or in school, we observed how people deal with low mood ranging from short-lived gloom to decades of suffering. Medicalized approaches to low mood resonate widely in the sample. The way low mood is addressed in common interaction is often consonant with depres- sion diagnostics and treatment, but we also found many in- teractions in which low mood was not medicalized. In relation to medicalization, these interactions were autonomous. We found dissonance, too: interactions in which medicalization was con- tested. This contestation, however, was also structured by medical categories and treatment and added to the dominance of medicalization.

In this case, sense-making is largely based on psychodynamic and social explanatory models (Kleinman et al.,1978) of depression, as Kangas and others have found (2001), while “Biomedicalization” (Clarke et al., 2003) and “Brain talk” are far less important than Rose (2007) suggests. Moreover, low mood is seldom pharmaceuti- calized: the availability and use of anti-depression medication is of limited relevance to everyday life, and pills are particularly controversial.

C. Br€oer, B. Besseling / Social Science & Medicine 183 (2017) 28e36 35

Therefore, we nuance Horwitz and Wakefield (2007) claim that a medicalized approach to low mood is so common that it has replaced normal sadness in discourse and experience. Low mood can be experienced without reference to clinical depression, even though the clinical approach is widely shared in the sample we studied. Moreover, while we see a cultural shift in that a medical- ized idiom (Nichter, 2010) is often common and taken for granted, it is also evident that the medicalized depression idiom itself is rife with uncertainty. People struggle to make sense of low mood and the depression label in everyday interactions, similarly to the way analysts attempt to make sense of this material. The way uncer- tainty surrounding low mood is dealt with in everyday life is similar to how it is approached by professionals and patients (Fox, 1957; Kokanovic et al., 2013; Rafalovich, 2005; Timmermans and Buchbinder, 2012; Weiner and Martin, 2008). The category of depression is furthermore reshaped in everyday life. The word “depression” has become normal and can be used to the extent that “depression” has lost psycho-medical connotations.

Looking at the interactions in greater detail, we consistently find that low mood interrupts the flow of interactions and triggers sense-making, (re)definitions of responsibilities and concerns about identities. These concerns are often unresolved by using depression labeling. In cases of current suffering, the use of the depression label often raises new concerns for the sufferer and his/ her intimate relations (“what is depression?”, “How should we deal with a depressed person?”). Treating low mood as clinical depression in daily life does not disambiguate responsibilities.

Norms of active citizenship, self-reliance and happiness are widely shared and lead to the problematizing of low mood but not necessarily to its medicalization. Similarly, everyday concerns and sense-making do not always encourage medicalization. However, interactional conflicts appear to be consistently tied to medicali- zation. Annoyance, irritation or doubts about friendship or collegial contact trigger medicalization, particularly by the non-depressed person. Approaching low mood as a medical problem settles these relational conflicts: The “depression” label and treatment enable either distancing or empathy and provide coherence and legitimacy, which is the classic case of micro-controlling deviance.

We interpret the large-scale, though non-encompassing, employment of depression diagnosis and treatment, the attempts at de-medicalization, the diversity in explanatory models and prolonged uncertainties as outcomes of the particular institution- alization of depression in care practices. Since contemporary care practices e in the Netherlands and beyond e are diverse and conflictual in themselves, people have a rather broad array of in- terpretations at their disposal. In the Netherlands and elsewhere, de-medicalization is even part of mental health policy. However, institutionalized diversity as a whole resonates dominantly in everyday life (see: Br€oer and Heerings, 2013 for a similar configu- ration regarding ADHD). Depression diagnosis and treatment, through their pragmatic use, dialectically mediate between everyday attempts to make sense of low mood and institutions marked by uncertainty.

Author contribution

Christian Br€oer: lead author, designed study and methods, collected and analyzed data.

Broos Besseling collected and analyzed data, co-authored parts of the text. Gail Zuckerwise and AJE edited the text. We benefitted from comments and collegial support by Patrick R. Brown and members of the Political Sociology group at the University of Amsterdam. This research was performed without external funding.

References

Abraham, J., 2010. Pharmaceuticalization of society in context: theoretical, empir- ical and health dimensions. Sociology 44 (4), 603e622.

Barker, K.K., 2014. Mindfulness meditation: do-it-yourself medicalization of every moment. Soc. Sci. Med. 106, 168e176.

Becker, G., Nachtigall, R.D., 1992. Eager for medicalization - the social production of infertility as a disease. Sociol. Health Illn. 14 (4), 456e471.

Bell, A.V., 2016. The margins of medicalization: diversity and context through the case of infertility. Soc. Sci. Med. 156, 39e46.

Bell, S.E., Figert, A.E., 2012. Medicalization and pharmaceuticalization at the in- tersections: looking backward, sideways and forward. Soc. Sci. Med. 75 (5), 775e783.

Blumer, H., 1969. Symbolic Interactionism: Perspective and Method. Prentice-Hall, Englewood Cliffs.

Br€oer, C., 2008. Private trouble, policy issue, how policy discourses shape our experience of aircraft sound. Crit. Policy Stud. 2 (2), 93e117.

Br€oer, C., Heerings, M., 2013. Neurobiology in public and private discourse: the case of adults with ADHD. Sociol. Health Illn. 35 (1), 49e65. http://dx.doi.org/ 10.1111/j.1467-9566.2012.01477.x.

Brown, P., De Graaf, S., Hillen, M., Smets, E., van Laarhoven, H., 2015. The inter- weaving of pharmaceutical and medical expectations as dynamics of micro- pharmaceuticalisation: advanced-stage cancer patients' hope in medicines alongside trust in professionals. Soc. Sci. Med. 131, 313e321.

Brown, P., Zavestoski, S., McCormick, S., Mayer, B., Morello-Frosch, R., Altman, R.G., 2004. Embodied health movements: new approaches to social movements in health. Sociol. Health Illn. 26 (1), 50e80.

Bury, M., 1982. Chronic illness as biographical disruption. Sociol. Health Illn. 4 (2), 167e182.

Charmaz, K., 1983. Loss of self: a fundamental form of suffering in the chronically ill. Sociol. Health Illn. 5 (2), 168e195.

Clarke, A.E., Shim, J.K., Mamo, L., Fosket, J.R., Fishman, J.R., 2003. Biomedicalization: technoscientific transformations of health, illness, and US biomedicine. Am. Sociol. Rev. 68 (2), 161e194.

Conrad, P., Mackie, T., Mehrotra, A., 2010. Estimating the costs of medicalization. Soc. Sci. Med. 70 (12), 1943e1947.

Conrad, P., Schneider, J.W., 1992. Deviance and Medicalization : from Badness to Sickness : with a New Afterword by the Authors (Expanded ed.). Temple Uni- versity Press, Philadelphia.

Conrad, P., 2007. The Medicalization of Society: On the Transformation of Human Conditions into Treatable Disorders. Johns Hopkins University Press, Baltimore.

Crawford, R., 1980. Healthism and the medicalization of everyday life. Int. J. Health Serv. 10 (3), 365e388.

Coveney, C., Gabe, J., Williams, S., 2012. Boosting brainpower? from the medical- isation of cognition to the pharmaceuticalisation of routine mental life. Salute E Societ�a.

Csikszentmihalyi, M., Larson, R., 1987. Validity and reliability of the experience- sampling method. J. Nerv. Ment. Dis. 175 (9), 526e536.

Danforth, S., Navarro, V., 2001. Hyper talk: sampling the social construction of ADHD in everyday language. Anthropol. Educ. Q. 32 (2), 167e190.

De Certau, M., 1984. The Practice of Everyday Life. University of California Press, Berkeley.

De Graaff, B.M., 2016. Should I Be Worried? Citizens' Experiences and the Risk Politics of Cell Site Deployment. (Doctoral dissertation). University of Amsterdam.

De Graaff, B., Br€oer, C., 2012. ‘We are the canary in a coal mine’, establishing a disease category and a new health risk. Health Risk Soc. 14.2, 129e147.

Dumit, J., 2006. Illnesses you have to fight to get: facts as forces in uncertain, emergent illnesses. Soc. Sci. Med. 62 (3), 577e590.

Epstein, S., 1996. Impure Science : AIDS, Activism, and the Politics of Knowledge. University of California Press, Berkeley.

Fox, N.J., Ward, K.J., 2008. Pharma in the bedroom... and the kitchen.... the phar- maceuticalisation of daily life. Sociol. Health Illn. 30 (6), 856e868.

Fox, R.C., 1957. Training for uncertainty. Student-Physician 207e241. Goffman, E., 1961. Asylums; Essays on the Social Situation of Mental Patients and

Other Inmates, [first ed. Anchor Books, Garden City, N.Y. Graf, W.D., Miller, G., Nagel, S.K., 2014. Addressing the problem of ADHD medication

as neuroenhancements. Expert Rev. Neurother. 0, 1e13. Hacking, I., 2007. Kinds of people: moving targets. Proc. Br. Acad. 151, 285e318. Hammersley, M., Atkinson, P., 2007. Ethnography: Principles in Practice. Routledge. Hardon, A., Idrus, N.I., Hymans, T.D., 2013. Chemical sexualities: the use of phar-

maceutical and cosmetic products by youth in south sulawesi, Indonesia. Reprod. Health Matters 21 (41), 214e224.

Horton-Salway, M., 2001. The construction of ME: the discursive action model. Discourse as Data A Guide Analysis 147e188.

Horwitz, A.V., Wakefield, J.C., 2007. The Loss of Sadness. Oxford University Press, New York.

Hsieh, H.F., Shannon, S.E., 2005. Three approaches to qualitative content analysis. Qual. Health Res. 15 (9), 1277e1288. http://dx.doi.org/10.1177/ 1049732305276687.

Illich, I., Cochrane, A., Williams, R., 1975. Medical Nemesis. Australian Broadcasting Commission, Science Programmes Unit.

Kangas, I., 2001. Making sense of depression: perceptions of melancholia in lay narratives. Health 5 (1), 76e92.

C. Br€oer, B. Besseling / Social Science & Medicine 183 (2017) 28e3636

Karp, D.A., 1994. Living with depression: illness and identity turning points. Qual. Health Res. 4 (1), 6e30.

Kirmayer, L.J., 2001. Cultural variations in the clinical presentation of depression and anxiety: implications for diagnosis and treatment. J. Clin. Psychiatry 62, 22e30.

Kleinman, A., 1985. Culture and Depression: Studies in the Anthropology and Cross- cultural Psychiatry of Affect and Disorder. Univ of California Press.

Kleinman, A., Eisenberg, L., Good, B., 1978. Culture, illness, and care: clinical lessons from anthropologic and cross-cultural research. Ann. Intern. Med. 88 (2), 251e258.

Knoblauch, H., 2005. Focused ethnography. Forum Qual. Soc. Res. 6 (3). Kokanovic, R., Bendelow, G., Philip, B., 2013. Depression: the ambivalence of diag-

nosis. Sociol. Health Illn. 35 (3), 377e390. Liu, K., King, M., Bearman, P.S., 2010. Social influence and the autism epidemic. AJS;

Am. J. Sociol. 115 (5), 1387. Lock, M., Kaufert, P.A., 1998. Pragmatic Women and Body Politics. Cambridge Uni-

versity Press. Lupton, D., 1997. Foucault and the medicalisation critique. Foucault, Health Med.

94e110. Mayring, P., 2000. Qualitative content analysis. Forum Qual. Soc. Res. 1 (2). McPherson, S., Armstrong, D., 2009. Negotiating ‘depression’in primary care: a

qualitative study. Soc. Sci. Med. 69 (8), 1137e1143. Mead, G.H., Morris, C.W., 1934. Mind, Self & Society : from the Standpoint of a Social

Behaviorist. University of Chicago Press, Chicago. Nichter, M., 2010. Idioms of distress revisited. Cult. Med. Psychiatry 34 (2), 401e416. Petersen, A., 2011. Authentic self-realization and depression. Int. Sociol. 26 (1),

5e24. Philip, B., 2009. Analysing the politics of self-help books on depression. J. Sociol. 45

(2), 151e168. Rafalovich, A., 2005. Exploring clinician uncertainty in the diagnosis and treatment

of attention deficit hyperactivity disorder. Sociol. Health Illn. 27 (3), 305e323. Ratcliffe, M., 2014. The phenomenology of depression and the nature of empathy.

Med. Health Care Philos. 17 (2), 269e280.

Rose, N., 2003. Neurochemical selves. Society 41 (1), 46e59. Rose, N., 2007. The Politics of Life Itself : Biomedicine, Power, and Subjectivity in the

Twenty-first Century. Princeton University Press, Princeton, NJ. Rossol, J., 2001. The medicalization of deviance as an interactive achievement: the

construction of compulsive gambling. Symb. Interact. 24 (3), 315e341. Saltonstall, R., 1993. Healthy bodies, social bodies: men's and women's concepts and

practices of health in everyday life. Soc. Sci. Med. 36 (1), 7e14. Scheff, T.J., 1970. Being Mentally Ill: a Sociological Theory (Transaction Books). Schutz, A., 1932. Der sinnhafte aufbau der sozialen welt; eine einleitung in die

verstehende soziologie. J. Springer, Wien,. Scott, S., 2009. Making Sense of Everyday Life (Polity). Szasz, T.S., 2007. The Medicalization of Everyday Life: Selected Essays. Syracuse

University Press. Thoits, P.A., 1985. Self-labeling processes in mental illness: the role of emotional

deviance. Am. J. Sociol. 221e249. Timmermans, S., Buchbinder, M., 2012. Saving Babies?: the Consequences of

Newborn Genetic Screening. University of Chicago Press. Torres, J.M., 2014. Medicalizing to demedicalize: lactation consultants and the (de)

medicalization of breastfeeding. Soc. Sci. Med. 100, 159e166. Verkenning, V.T., 2013. Nationaal Kompas Volksgezondheid, vol. 9. RIVM, Bilthoven. Verweij, G., Houben-van Herten, M., 2013. Depressiviteit en antidepressiva in

Nederland. Wasser, J.D., Bresler, L., 1996. Working in the interpretive zone: conceptualizing

collaboration in qualitative research teams. Educ. Res. 5e15. Weiner, K., Martin, P., 2008. A genetic future for coronary heart disease? Sociol.

Health Illn. 30 (3), 380e395. Westerbeek, J., Mutsaers, K., 2008. Depression narratives: how the self became a

problem. Literature Med. 27 (1), 25e55. Wheeler, L., Reis, H.T., 1991. Self-recording of everyday life events: origins, types,

and uses. J. Personality 59 (3), 339e354. Willcock, H., 1943. Mass-observation. Am. J. Sociol. 445e456. Williams, S.J., Calnan, M., 1996. The 'limits' of medicalization?: Modern medicine

and the lay populace in 'late' modernity. Soc. Sci. Med. 42 (12), 1609e1620.